To Helmy H.

BALIGH
Teacher and Friend

R.M.B.
B.O.
STRATEGIe ORGANIZATIONAL
DIAGNOSIS AND DESIGN
Developing Theory for Application

Second Edition

Richard M. Burton
The Fuqua School of Business
Duke University
Durham, North Carolina

B0rge Obel
Department of Management
Odense University
Odense, Denmark

With contributions by

Starling D. Hunter III

Sloan School of Management
MIT
Cambridge, Massachusetts

Mikael S0ndergaard and Dorthe D0jbak

Department of Management
Odense University
Odense, Denmark

~.

"
KLUWER ACADEMIC PUBLISHERS
BostonIDordrechtILondon
Distributors for North, Central and South America:
Kluwer Academic Publishers
101 Philip Drive
Assinippi Park
Norwell, Massachusetts 02061 USA
Telephone (781) 871-6600
Fax (781) 871-6528
E-Mail <[email protected]>

Distributors for all other countries:

Kluwer Academic Publishers Group
Distribution Centre
Post Office Box 322
3300 AH Dordrecht, THE NETHERLANDS
Telephone 31 78 6392 392
Fax 31 786546474
E-Mail <[email protected]>

" Electronic Services <http://www.wkap.nl>

Library of Congress Cataloging-in-Publication Data

Burton, Richard M.
Strategic organizational diagnosis and design : developing theory for
application 1 Richard M. Burton, B0rge abel : with contributions by Starling
D. Hunter III, Mikael S011dergaard, and Dorthe D0jbak. -- 2nd ed.
p. cm.
Includes bibliographical references and index.
ISBN-13: 978-0-7923-8247-8 e-ISBN-13: 978-1-4684-0021-2
DOl: 10.1007/978-1-4684-0021-2
1. Organization. 2. Corporate culture. 3. Psychology, Industrial.
I. abel, B0rge. II. Hunter, Starling. III. S0ndergaard, Mikael.
IV. D0jback, Dorthe. V. Title.
HD31.B8524 1998
302.3'5--dc21 98-28277
CIP

Copyright © 1998 by Kluwer Academic Publishers. Second Printing 1999.

All rights reserved. No part of this publication may be reproduced, stored in a

retrieval system or transmitted in any form or by any means, mechanical, photo-
copying, recording, or otherwise, without the prior written permission of the
publisher, Kluwer Academic Publishers, 101 Philip Drive, Assinippi Park, Norwell,
Massachusetts 02061

Printed on acid-free paper.

STRATEGIe ORGANIZATIONAL
DIAGNOSIS AND DESIGN
Developing Theory for Application

Second Edition
Table of Contents

Preface .................................................... xv

1. Introduction ......................................... 1

1.1 Introduction ........................................... 1

1.2 What is an Organization? ................................ 2
1.3 An Information Processing View of Organizations ............ 4
1.3.1 Neo-information Processing as a Basis for
Organizational Design . . . . . . . . . . . . . . . . . . . . . . . . . . .. 7
1.4 Strategic Organizational Design . . . . . . . . . . . . . . . . . . . . . . . . . .. 9
1.5 The Basic Features .................................... 11
1.5.1 Fit Criteria for Designing the Knowledge Base ....... 15
1.5.1.1 Contingency Fit ........................ 16
1.5.1.2 Situation Fit ........................... 17
1.5.1.3 Design Parameter Fit ................... 17
1.5.1.4 Total Design Fit ........................ 18
1.6 Creating a Knowledge Base from the Literature ............ 19
1.6.1 Composing the Knowledge Base ................... 24
1.7 The Organizational Consultant for Designing an Organization 31
1.8 Validating Theory for Application ........................ 32
1.8.1 The Foundation Expert System ................... 34
1.8.2 Validation and Continued Model Development ....... 35
1.8.3 A Process of Validation .......................... 36
1.9. Summary............................................ 37

2. Describing the Organization for Design Purposes 41

2.1 Introduction .......................................... 41

2.1.1 The Scandinavian Airline System (SAS) ............ 42
2.2 Organizational Configurations ........................... 45
2.2.1 Simple Configuration ............................ 46
2.2.2 Functional Configuration ........................ 48
viii Table of Contents

2.2.2.1 A Functional Organization ............... 49

2.2.3 Divisional Configuration . . . . . . . . . . . . . . . . . . . . . . . .. 53
2.2.3.1 A Divisional Organization. . . . . . . . . . . . . . .. 55
2.2.4 Matrix Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . .. 59
2.2.4.1 A Matrix Organization .................. 61
2.2.5 Ad Hoc Configuration ........................... 64
2.2.6 Bureaucracies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 65
2.2.7 International Configurations ..................... 66
2.3 Organizational Complexity . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 68
2.3.1 Horizontal Differentiation. . . . . . . . . . . . . . . . . . . . . . .. 69
2.3.2 Vertical Differentiation .......................... 71
2.3.3 Spatial Differentiation ......................... " 72
2.4 Formalization ....................................... " 73
2.5 Centralization. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 75
2.6 Coordination and Control ............................... 76
2.7 The Organizational Structure of SAS ..................... 77
2.8 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 80

3. Management and Leadership Style . ............... 83

3.1 Introduction .......................................... 83

3.1.1 Management and Structure ...................... 84
3.2 Literature Review ..................................... 86
3.3 Management and Leadership Style as a Contingency ........ 92
3.3.1 Definition of Management and Leadership Style:
Microinvolvement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 92
3.3.2 Management and Leadership Style as a
Contingency Factor ............................. 99
3.4 Managing the Management ............................ 104
3.5 SAS: The Management Style of Carlzon and Stenberg. . . . . .. 105
3.6 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 108

4. Organizational Climate ........................... 111

4.1 Introduction ......................................... 111

4.2 Climate and Culture .................................. 112
4.3 Literature Review on Climate. . . . . . . . . . . . . . . . . . . . . . . . . .. 117
4.5 Measuring and Categorizing Climate. . . . . . . . . . . . . . . . . . . .. 120
4.6 Climate as a Contingency .............................. 124
4.6.1 Describing a Group Climate ..................... 125
4.6.2 The Group Climate Effects on Structure ........... 126
Table of Contents IX

4.6.3 Describing the Developmental Climate ............ 128

4.6.4 Developmental Climate Effect on Structure ........ 130
4.6.5 Describing the Internal Process Climate ........... 131
4.6.6 Internal Process Climate Effects on Structure ...... 133
4.6.7 Describing the Rational Goal Climate ............. 134
4.6.8 Rational Goal Climate Effects on Structure ........ 136
4.7 Managing the Climate ................................ , 137
4.8 The Climate at SAS ......... . . . . . . . . . . . . . . . . . . . . . . . . .. 138
4.9 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 140

5. Size and Skill Capabilities .. .. . .. .. .. . . . . . . . . . . . ... 145

5.1 Introduction ......................................... 145

5.2 A Literature Review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 146
5.2.1 Size: A Variety of Perceptions .................... 146
5.2.2 The Measurement of Size ....................... 147
5.2.3 Size as Imperative ............................. 148
5.2.4 An Information-Processing Perspective on Size ..... 149
5.3 The Effects of Size and Skill Capability on Organizational
Structure ........................................... 151
5.3.1 Measuring Size and Skill Capability for
Design Purposes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 151
5.3.2 Size as a Contingency .......................... 154
5.3.2.1 Size Effects on Complexity .............. 155
5.3.2.2 Size Effects on Centralization ........... 156
5.3.2.3 Size Effects on Formalization ............ 157
5.3.2.4 Size Effects on Configuration . . . . . . . . . . .. 158
5.4 Managing Size ....................................... 160
5.5 The Size of SAS ...................................... 161
5.6 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 162

6. The Environment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 165

6.1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 165

6.2 Background and the Literature ......................... 167
6.2.1 Measures ofthe Environment .................... 167
6.2.2 The Environment-Structure Relationship . . . . . . . . .. 171
6.3 Describing the Environment: Equivocality, Uncertainty,
Complexity, and Hostility .............................. 174
6.4 Environment as a Contingency Factor . . . . . . . . . . . . . . . . . . .. 180
x Table of Contents

6.5 Environmental Effects on Configuration, Coordination,

Media Richness, and Incentives ......................... 190
6.6 Operationalizations of the Environmental Measures ........ 200
6.7 Managing the Environment ............................ 202
6.8 The SAS Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 204
6.9 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 207

7. Technology ......................................... 211

7.1 Introduction ......................................... 211

7.2 A Literature Review ................................... 212
7.3 Technology as a Contingency ........................... 220
7.3.1 Measuring Technology .......................... 220
7.3.2 Technology as Contingency Factor ................ 224
7.3.2.1 Technology Effects on Formalization ...... 225
7.3.2.2 Technology Effects on Centralization ..... 226
7.3.2.3 Technology effects on Organizational
Complexity . . . . . . . . . . . . . . . . . . . . . . . . . .. 227
7.3.2.4 Technology Effects on Configuration ...... 229
7.3.2.5 Technology Effects on Coordination and
Control Mechanisms ................... 232
7.4 Managing Technology ................................. 234
7.5 The Technology ofSAS ................................ 236
7.6 Summary ........................................... 238

8. Strategy ............................................ 241

8.1 Introduction ......................................... 241

8.2 A Literature Review. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 241
8.2.1 Structure Follows Strategy ...................... 241
8.2.2 The Counter Proposition: Strategy Follows Structure 243
8.2.3 Fit: Strategy and Structure . . . . . . . . . . . . . . . . . . . . .. 245
8.3 Definition of Strategy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 248
8.4 The Two-Stage Model: Description, Strategy, Structure ..... 254
8.5 The Propositions for Strategy Description . . . . . . . . . . . . . . . .. 255
8.5.1 Describing a Prospector ......................... 257
8.5.2 Prospector Effects on Structure .................. 259
8.5.3 Describing an Analyzer Without Innovation ....... 260
8.5.4 Analyzer Without Innovation Effects on Structure ... 262
8.5.5 Describing Analyzer with Innovation .............. 263
8.5.6 Analyzer with Innovation Effects on Structure ...... 265
Table of Contents Xl

8.5.7 Describing a Defender .......................... 266

8.5.8 Defender Effects on Structure .................... 267
8.5.9 Reactor Strategy .............................. 269
8.6 International Dimensions .............................. 270
8.7 Choosing the Right Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . .. 271
8.8 The Strategy of SAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 272
8.9 Summary ........................................... 276

9. Organizational Design: A Synthesis .............. 281

9.1 Introduction ......................................... 281

9.2 Diagnosis and Design .................................. 282
9.3 Contingency Fit ...................................... 286
9.3.1 The Simple Configuration ....................... 286
9.3.2 The Functional Configuration .................... 288
9.3.3 Divisional Configuration ........................ 290
9.3.4 Matrix Configuration ........................... 291
9.3.5 Ad Hoc Configuration .......................... 292
9.3.6 The Professional Bureaucracy ................... 294
9.3.7 The Machine Bureaucracy ....................... 295
9.3.8 Organizational Complexity ...................... 296
9.3.9 Formalization ................................. 299
9.3.10 Centralization ................................ 303
9.3.11 Coordination and Control ....................... 307
9.3.12 Media Richness and Incentives ................... 308
9.4 Situation Fits and Misfits .............................. 308
9.5 Design Parameter Fit ................................. 319
9.6 Total Design Fit ...................................... 320
9.7 Designing the Organization ............................ 325
9.7.1 Design is Lifecycle Management: Evolution
and Revolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 327
9.7.2 Design is Exploration and Exploitation ............ 332
9.7.3 Design is Learning ............................. 335
9.8 SAS ................................................ 338
9.9 SAS and the Lifecycle ................................. 341
9.10 Summary ........................................... 342

10. Using the Organizational Consultant ............ 347

10.1 Introduction ......................................... 347

10.2 Practical Organizational Diagnosis and Design ............ 349
XlI Table of Contents

10.3 Alignment ofthe Understanding ofthe Organization's

Situation ............................................ 351
10.4 What Can We Learn By Doing? ......................... 351
10.5 Cases............................................... 354
10.6 Organizational Consultant Setup . . . . . . . . . . . . . . . . . . . . . . .. 355

11. Using Publicly-Available Information for Stra-

tegic Organizational Diagnosis and Design . ... 357

11.1 Introduction ......................................... 357

11.2 Data Sources, Collection and Analysis . . . . . . . . . . . . . . . . . . .. 357
11.2.1 Description of Data Sources ..................... 357
11.2.2 Content Analysis Approach. . . . . . . . . . . . . . . . . . . . .. 360
11.3 Company Profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 362
11.4 Discussion of Answers to OrgCon Questions . . . . . . . . . . . . . .. 364
11.4.1 Environment .................................. 364
11.4.2 Strategy ..................................... 366
11.4.3 Technology................................... 370
11.4.4 Product Diversity . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 373
11.4.5 Climate ...................................... 375
11.4.6 Management Preferences ....................... 377
11.4.7 Size......................................... 379
11.4.8 Age and Ownership ............................ 380
11.4.9 Current Configuration. . . . . . . . . . . . . . . . . . . . . . . . .. 380
11.4.10 Current Complexity ............................ 380
11.4.11 Current Formalization .......................... 383
11.4.12 Current Centralization ......................... 385
11.5 Total Organizational Analysis .......................... 386
11.5.1 Size ......................................... 386
11.5.2 Climate...................................... 386
11.5.3 Management Style . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 387
11.5.4 Strategy ..................................... 388
11.5.5 Current Organizational Characteristics ........... 388
11.5.6 Situation Misfits .............................. 389
11.5.7 Organizational Consultant Recommendations ...... 389
11.5.8 Organizational Configurations ............... . . .. 389
11.5.9 Organizational Characteristics ................... 390
11.5.10 Organizational Misfits .......................... 392
11.5.11 More Detailed Recommendations ................. 392
11.6 Detailed Discussion of Organizational and Situational Misfits 392
Table of Contents Xlll

12. Oticon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 399

12.1 Introduction ......................................... 399

12.2 Case 1: The Top Management Group That Got Fired
... by Itself .......................................... 399
12.2.1 The Demant Era ............................... 400
12.2.2 Growth and Expansion in The Seventies ........... 401
12.2.3 Technology Changes in the Eighties .............. 402
12.2.4 The Difference Between Behind The Ear
and In The Ear Aids . . . . . . . . . . . . . . . . . . . . . . . . . . .. 403
12.2.5 External Influences Leading to Internal Weaknesses 404
12.2.6 Need for Changes .............................. 406
12.2.7 One Must Take the Good with the Bad ............ 407
12.2.8 Kolind's Style ................................. 407
12.2.9 Reductions, Cut Backs and the New Oticon ........ 409
12.2.10 Questions for Consideration ..................... 409
12.3 Case 2: The Spaghetti That Became Revolutionary ......... 410
12.3.1 What Made Kolind Think of Spaghetti? ............ 410
12.3.2 Holly, Ivy and Unthinkable Thoughts ............. 411
12.3.3 Realizing the Unthinkable ...................... 412
12.3.4 Implementation ofthe Undefinable ............... 413
12.3.5 The Physical and Psychic Influences .............. 414
12.3.6 The Computer Network and the Paperless Office .... 415
12.3.7 Spaghetti Sometimes Gets Sticky ................. 416
12.3.8 Spaghetti in Action ............................ 417
12.3.9 Questions for Consideration ..................... 418
12.4 Case3: What Happens ifthe Spaghetti Gets Cold? .......... 419
12.4.1 Kolind's Management Style ..................... 419
12.4.2 How to Get Value for Values ..................... 420
12.4.3 Interaction and Integration ...................... 421
12.4.4 Thisted's Role in the Spaghetti ................... 422
12.4.5 Spaghetti for All Nationalities ................... 423
12.4.6 How to Become Part ofthe Spaghetti ............. 424
12.4.7 More Control than Meets the Eye ................ 425
12.4.8 The Transparent Office ......................... 426
12.4.9 Authority Versus Group Mentality ................ 427
12.4.10 Enthusiastic Salaries ........................... 427
12.4.11 It Takes More than Success to Prevent Failure ..... 428
12.4.12 Questions for Consideration ..................... 429
12.5 Case 4: Scouts are More Attractive ...................... 430
12.5.1 Questions for Consideration ..................... 432
xiv Table of Contents

Appendix 1 ............................................... 433

Appendix 2 ............................................... 434
Appendix 3 ............................................... 435

13. Sample Cases and Exercises ..................... 439

13.1 Introduction ......................................... 439

13.2 ABB Electromechanical Meters ......................... 440
13.2.1 Background ................................... 440
13.2.2 The Electromechanical Meter Organization ........ 441
13.2.3 Manufacturing ................................ 442
13.2.4 Design Engineering ............................ 443
13.2.5 Marketing .................................... 443
13.2.6 Intergroup Communications ..................... 443
13.2.7 Managerial Responsibilities ..................... 444
13.2.8 Questions for Consideration ..................... 444
13.3 Duke University Press . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 445
13.3.1 Organizational Description ...................... 445
13.3.2 Questions for Consideration ..................... 449
13.4 GTE Government Systems Mobile Subscriber
Equipment Division . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 449
13.4.1 Background and Issues ......................... 449
13.4.2 Analysis ..................................... 451
13.4.3 Questions for Consideration ..................... 452
13.5 Bluestone Group, Inc. ................................. 452
13.5.1 Background ................................... 453
13.5.2 Part I . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 455
13.5.2.1 Current Organization .................. 455
13.5.3 Part II ............................ . . . . . . . . . .. 456
13.5.3.1 Goals and Objectives ofthe Dominant
Coalition . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 457
13.5.3.2 Environment ......................... 457
13.5.3.3 Work-flow and Technologies Used ........ 458
13.5.3.4 Existing Organizational structure . . . . . . .. 459
13.5.4 Questions for Consideration ..................... 460

Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 463

Preface to the Second Edition

We will not repeat our preface discussion from the first edition. Here we
only add some new comments:

• a note to the reader and user,

• changes in the book and the Organizational Consultant (OrgCon),
and,
• our thanks to the many individuals who have contributed critically to
this venture, read and reviewed the book, contributed chapters and
cases, and similarly used and critiqued the OrgCon.

For the reader, there are two very contrasting approaches to reading this
book and learning about organizational design. The more traditional ap-
proach is to read the book, and then use the OrgCon on cases and applica-
tions. The second approach is to begin with the OrgCon software and only
examine the book as you find it helpful. Which approach is better? It is your
choice, not ours. In our experience, students in organizational design prefer
to start with the OrgCon and a case, rather than with the book itself.
Readers who have more background in organization theory and design
usually examine the book first. We have tried to write the book so that it
can serve both as a reference and an integrated presentation.
We made a number of changes in this edition. We have revised the
OrgCon and put it in a Windows frame. It is much friendlier and you can
spend your time focusing on the cases and learning about organizational
design. The OrgCon is contained on the accompanying CD-ROM-here you
will also find a number of data input files for cases.
In the book itself, we have kept the main ideas and the basic approach.
But, there are significant changes. We have added a new chapter on climate
and incorporated climate into the multidimensional contingency theory.
Climate helps determine what the organization should be, and how it
should be designed. The addition of climate to the model makes it more
nearly complete and adds considerably to our understanding of the organi-
zation and its design.
The synthesis Chapter 8 on diagnosis and design in the first edition has
been totally rewritten as Chapter 9 here. We have eliminated a good deal
of the earlier redundancy and spelled out more clearly what we mean by
design. It is not an easy concept. Design is both a product and a process.
xvi Preface to the Second Edition

Design is organizationallifecycle management and dynamic. Design can be

viewed as March's concepts of exploration and exploitation. And finally,
design is organizational learning: it is a never ending process of experience,
error and refinement.
We have added capsules on new organizational forms and a number of
other topics. They are snippets which are related and relevant, but may not
fit t1:le flow and the order ofthe book. Norman Davies, in his EUROPE: A
HISTORY, 1996, called them "capsules." His capsules: cross the boundaries
of the chapters, illustrate curiosities, whimsies, and inconsequential side-
streams, glimpses of "new methods, new disciplines, and new fields" of
recent research. Our capsules are similar in spirit: the research support is
weak, most of support is conjectural or assertive; a popular new word as
delayering or concept as hypercompetition; or small case stories. We hope
the reader will find them interesting.
The Windows 95 version of OrgCon has been a multi-person, multi-
location project. In Odense, Jan Mikkelsen has been our fundamental
programmer. Roxanne Zolin, Stanford University, planned and managed
the project as well as adding critical programming elements. In Omaha,
Miriam Zolin developed the help-part. Peter Obel has been our expert for
all the things that go beep in the night as well as fixing problems when
others gave up. The computer team, Bjame Nielsen and Jan Pedersen at
Odense have as usual provided invaluable 24 hours a day support.
Our friends have made this venture fun; even when we could not laugh,
they did for us. We have been very fortunate to have their support and help.
Starling D. Hunter III, Mikael S~ndergaard, and Dorthe D~jbak permitted
us to include research chapters; we are very pleased. Starling in Chapter
11 demonstrates how to use publicly available information to create the
input data for an OrgCon analysis, and then the implication ofthe results.
Mikael and Dorthe in Chapter 12 developed a very interesting case: Oticon,
the spaghetti organization. This book is a research book. It is not an easy
read. Yet, a few colleagues have used it as a textbook and provided very
valuable feedback. We would like to thank Ray Levitt, Torkild J~rgensen,
Gerry DeSanctis, Bo Eriksen, Carl Jones, and Henrik Bendix. Several
individuals have read and critiqued our effort: Pat Keith, Stan Rifkin, Ray
Levitt, Jan Thomsen, Starling Hunter, Dorthe D~jbak, Torbj~m Knudsen,
and Jeanette Larsen. Arie Lewin, Duke University, is our special critic; he
has the ability to coach and make the end result better. We are very
grateful.
The Fuqua School of Business, Duke University and The School of
Business and Economics, Odense University, our home institutions, have
provided much support. Part of the revision was written at Stanford
University, where Obel was a Research Scholar at Cife and Scancor,
Stanford University. Ray Levitt has been a most gracious host. We also like
Preface to the Second Edition XVll

to thank Jim March and the many researchers at Scancor for the many
question they asked, which forced us to rethink and sharpen our argu-
ments. We express our thanks to Stanford University. Our work has been
generously supported by The Danish Social Science Research CounciL
At Odense University, Mona Andersen made disarray into wonderful
text. She made the impossible seem normal for this edition as she did for
the first one.
At Kluwer, Zack Rolnick has made it all possible. The Kluwer team has
responded at critical times to make it happen.
As ever, Helmy Baligh is our anchor who keeps us focused on what is
important.

We would like to thank them all.

Rich Burton and Bl<1rge Obel

Preface to the First Edition1

Organizational design is a normative science with the goal of prescribing

how an organization should be structured in order to function effectively
and efficiently. Organizational theory is a positive science that states our
understanding about how the world operates and contrasts that under-
standing with a view of how the world could possibly operate. It provides
the theoretical underpinnings for organizational design. In this book, we
attempt to construct an approach for diagnosing and designing organiza-
tions built on a knowledge base of organizational theory.
Organizational design is a young field that incorporates many concepts
and approaches. In organizational design literature to date, there seems to
have been only two ways of doing things in this field-either to be so general
and so simple that the various interpretations do not yield practical design
implications, or to be so detailed and specific that generalization to other
situations is almost impossible. We attempt here to strike a balance-and
offer an approach that is applicable to a broad range of situations.
In our view, organizational theory exists as a large body of related
languages, definitions, hypotheses, analyses, and conclusions. Our know-
ledge is vast, diverse, somewhat inconsistent, and generally unconnected.
Yet there is an underlying core of knowledge that can be used for analytical
purposes. Creating this balanced approach requires that the knowledge be
distilled and augmented to produce a set of clear and consistent design
rules that can be used to recommend what the organization's design should
be.
In organizational theory we have attempted to find simplicity. The
search for a dominant contingency or imperative has led to a parade of
paradigms which examine elements oftechnology, size, environmental, and
strategy paradigms. Moreover, little effort has been made to remove the
simplicity of these paradigms and put the various pieces together in a
reasonable fashion. Schoonhoven (1981) lists a number of problems with
the contingency view of organizational design, including lack of clarity, lack
of understanding about interactions, and a lack of acknowledgment of the
functional forms of the interactions. Others (Donaldson, 1982, 1987;
Schreyogg, 1980; Pennings, 1987) have debated whether a contingency view

1. The chapters and sections mentioned refer to the first edition. Several
chapters have been deleted and rearranged in the second edition.
Preface to the First Edition XIX

was appropriate or not. The difficulties come from many sources. In many
studies the distinction between description and prescription is not at all
clear. In many cases a silent shift is made from the description to the
prescription. Indeed, we must incorporate knowledge that comes from
descriptive studies. But we want to prescribe. We want to recommend what
the organizational design should be.
Designing an organization is no simple task. From Mintzberg's five
classes of structure, Mintzberg (1979), one can generate over a million
design alternatives. How does one search such a design space? To illustrate,
let us assume that an organizational structure can be defined as functional,
divisional, or matrix and as centralized or not and formalized or not. There
are then 3x2x2 =12 possible designs from which to choose. The number of
choices grows nonlinearly as the number of organizational dimensions
approaches useful realistic proportions. Let us add the choice that the
organization could have high specialization or not. The number of choices
is now 2x12 = 24. In a contingency theory of organization, these design
alternatives would need to be evaluated for all possible conditions. In the
simple case often contingency dimensions and only two values allowed for
each, we would have to evaluate twenty-four classes of structures under
1,024 different conditions. When the dimensions and number of variables
approaches a useful size for analytical purposes, the choice and evaluation
of organizational design alternatives becomes an enormous problem. There
is no way students, teachers, consultants, and managers can handle such
a complex choice situation without some order and systematic approach.
This book is accompanied by Organizational Consultant-a computer-
based expert system that operationalizes the theoretical propositions
developed in this book. Organizational Consultant is a program that applies
the underlying theory to deal with millions of design possibilities in a
systematic and comprehensive manner to diagnose an organization and
offer recommendations for its design, configuration, complexity, formaliza-
tion, centralization, media richness, committees, meetings, and liaison
relations.
The book is divided into two parts. Part 1 consisting of the first 8
chapters is the theory part, where the multiple contingency model is
developed. One special feature of Part 1 is its detailed analysis of Scandina-
vian Airlines Systems (SAS) from 1950 to today. SAS has undergone a
number of changes in its situation and design during this period, and we
analyze those changes to provide insight into organizational design. Other
business cases are included to illustrate various concepts and to relate the
theory to practice.
Part 2 is a manual/case part consisting of Chapters 9-13. Part 2
describes how the Organizational Consultant operates and how it can be
used to analyze cases.
xx Preface to the First Edition

This part also contains seven cases specifically developed to be used with
Organizational Consultant. For Part 2, Nancy Keeshan is a coauthor.
Although Part 1, Part 2, and Organizational Consultant can be con-
sidered as a unit, each is self-contained. For instance, Organizational
Consultant can also be used with other textbooks in organizational theory
such as Robbins (1990), Daft, (1992), and Mintzberg, (1984).
Organizational Consultant is a teaching tool in itself. Explanation and
help functions contain many parts of the theory. Extensive use ofOrganiza-
tional Consultant can teach the student many issues of the theory.
The book and software have been developed over some years. We have
used both the book and software in many courses both at The Fuqua School
of Business and Odense University. We would like to thank our students
for valuable comments and critique. Many students developed cases from
their own organization as part of the courses and we were allowed to
include some ofthese cases in Part 2 of this book. We would like to thank
Steve Hulme, Louis Grilli, David Thomason, Kathy Brodeur, Barbara
Johnson, and Kelly Leovic. Ray Levitt at Stanford also used both the book
and Organizational Consultant in his course. Ray's many comments helped
improve the final version. Additionally, he convinced his students Hyun-
Chan Cho, Hajime Kitarnura, and Tin-Hsuan Wu to prepare a case for Part
2. We are very grateful for his effort. Many individuals have worked with
us as teaching assistants, programmers, and research assistants. We would
like to thank Berit Jensen, Claus Bo JlIlrgensen, Mogens Kjrer Harregaard,
Kim Madsen, Mads Haugard, Bo Hegedal and Sf/Jren Hjortkjrer. Our
colleagues Caroll Stephens, Arie Lewin, Tom Naylor, Mikael SlIlndergaard,
Bjarne Nielsen, and Henrik Bendix offered comments, critique, and help in
various ways. A number of executives allowed us to test Organizational
Consultant in their organization. We are grateful to Don Namm, Glenn
Weingarth, Don Melick, Benny Mortensen, Hans Jf/Jrgen Hansen, Jf/Jrn
Henrik Petersen, Sven-Erik Petersen, and Arne Fredens. We would also
like to thank the reviewers and staff from Kluwer. Their comments and
help is highly appreciated. Mona Andersen has worked with us in all these
years. She has typed and organized several versions as well as prepared
this final one. It has been a pleasure to work with Mona.
We would to thank the Danish Social Science Research Council, Duke
University, and Odense University for the support they have provided to
make this research possible.
And finally, and most importantly, Helmy H. Baligh has been our fellow
traveler on many ventures. He is coauthor on a number of base papers for
this book and has been our most supportive and most critical colleague-
thus the dedication.

Richard Burton and Bf/Jrge Obel

CHAPTER 1

Introduction

1.1 Introduction
On Saturday, January 30, 1993, the Herald Sun, the local newspaper in
Durham, North Carolina, reported that IBM was facing problems. IBM had
in the previous week announced a cut in dividends and planned to replace
its chairman, John Akers. IBM's profit went from a $6 billion profit in 1990
to $2.8 billion and $4.9 billion losses in 1991 and 1992 respectively. Peter
Lieu, a computer analyst at Furman Selz, was quoted as saying that "John
Akers inherited a mess and the mess is a highly centralized organization
with virtually no delegation of responsibility." The problems facing IBM
were compared to the situation at AT&T, which, after losing $1.23 billion
in 1988, turned profitable in 1990 with a reported 1992 fourth-quarter
profit of $1 billion. The success of AT&T was attributed to a decentralized
management style introduced by the late James Olson and continued under
the new leadership by Robert Allen. Allen cut staffby 5 percent and wrote
off $6.7 billion in old analog technology. He diversified and brought in a
new management. Meanwhile; IBM was struggling with its old manage-
ment style and was accused of "failing to 'obsolete' its own products quickly
enough." IBM's problems had arisen because of a new competitive situation
in all of its markets-a change away from the use of mainframe computers
and toward workstations and networks. Additionally, competition in the
personal computer market had changed with declining growth and
confusion about which operating system would take the lead. Further, IBM
also "had some bad luck in the form of the global recession, which cut into
international profits." Today, IBM has a new CEO, who continues to
decentralize and downsize with an emphasis on workstations, PCs and
network services.
This story points to a number of issues that are important for business
success; situations change. Competition, technology, and economic
conditions change over time, and firms have to adjust to these changes.
Adjustments depend on the management and its style, which is the basis
for selecting a strategy that will lead the firm to success in its new
conditions. Finally, the firm needs an organization and an organizational
structure that will enable it to carry out its selected strategy. The proper
fit between the firm's environment, management style, climate, size,
2 Chapter 1. Introduction

technology, strategy, and structure is a necessary condition for a business

success. However, it is not sufficient nor a guarantee for success.
IBM and AT&T are two very large corporations, but the same argu-
ments can be made for small and medium-size firms as well. Small local
companies in Europe now face new and stiff competition due to the single
economic market in the European Union (EU). New information technology
changes the way business is carried out in many small and medium- size
corporations. AT&T and IBM struggled to find a way to fight back. It has
not been easy for them to find the appropriate response, even with large
staffs of experts and numerous consultants. Small and medium-size firms
may have even greater difficulties finding the proper fit.
This book summarizes our knowledge on how to design efficient
organizations to meet changing conditions. The focus of the book is on
strategic organizational design. This includes the composition of organiza-
tional units, reporting relationships among units, and other structural
relationships. It also includes the design of organizationwide methods,
procedures, and work technologies (Nadler and Tushman, 1988), as well as
the organization's incentive system.

1.2 What is an Organization?

Organizations are everywhere and central to our lives. We are part of many
organizations-where we live, where we work, where we shop, and where we
do many other activities. Scandinavian Airlines System (SAS) is an
organization, as is General Motors, Odense University, The Fuqua School
of Business, and a two-person pizza shop. Organizations are such a part of
our lives that we seldom reflect on exactly what we mean when we speak
of an organization.
So what is an organization? Many definitions can be found in the
literature. Here are some:

• "Organizations are social units (or human groupings) deliberately

constructed and reconstructed to seek specific goals" (Etzioni, 1964, p.
3).
• "The purpose of organizations is to exploit the fact that many (virtually
all) decisions require the participation of many individuals for their
effectiveness ... Organizations are a means of achieving the benefit of
collective action in the situations in which the price system fails"
(Arrow, 1974, p. 33).
• "Organizations are social entities that are goal-directed, deliberately
structured activity systems with an identifiable boundary" (Daft, 1992,
p.7).
1.2 What is an Organization? 3

• "An organization is a consciously coordinated social entity, with a

relatively identifiable boundary, which functions on a relatively
continuously basis to achieve a common goal or a set of goals" (Robbins,
1990, p. 4).

The definitions have much in common and explain the concept of what an
organization is.
An organization is a social entity. It exists for and is made up of
individuals. Corporations, small businesses, political parties, and religious
institutions are examples. An organization is more than a simple collection
of people, such as shoppers at a supermarket who are there for their own
individual purposes. An organization has goals and it exists for a purpose.
This purpose is shared, for the most part, by its members. Individuals fulfill
their own goals by participating in the organization. Organizations have
activities: they sell groceries, make automobiles, sell computers, elect
councils, and help the poor. Organizations have boundaries: some individu-
als or activities are inside, and others are outside. The outside is called the
environment, which is an important consideration for the design of the
inside. Finally, an organization is deliberately constructed: it is an artificial
entity, it is designed.
Organizations are designed to do something-undertake activities to
accomplish goals. Additionally, members of an organization benefit by being
members. They gain from cooperation because individuals have different
talents and individuals' efficiency usually improves with specialization
(Arrow, 1974). Organizations emerge when the cooperation and specializa-
tion cannot be obtained through a market and price system but activities
have to occur within an organization. A fundamental issue in designing an
organization is to group small activities together so that goals are realized
or conversely, to take a large task and break it into smaller tasks. In
manufacturing an automobile, a large task, thousands of smaller tasks
must be grouped to accomplish the larger task. The organizational design
problem is both how to put the smaller tasks together and how to take a
large task and break it down into appropriate smaller tasks. Whichever
view is adopted, the smaller activities must be coordinated to accomplish
the larger task and to realize the organization's goal. Coordination is
putting the pieces together. Without coordination, we do not have an
organization but only a collection of separate activities. Coordination, then,
introduces the need for information and information exchange in order to
reduce the uncertainty associated with the tasks; otherwise the required
coordination cannot be obtained. Generally, we use information processing
as an integrating concept in the design of organizations (Tushman and
Nadler, 1988).
4 Chapter 1. Introduction

As is discussed in the next section, various criteria may be applied to

help us select among possible designs. Three criteria are paramount in
organizational design-effectiveness, efficiency, and viability-an organiza-
tion should be designed to meet these criteria:

• Effectiveness: An organization is effective if it realizes its purpose and

accomplishes its goals.
• Efficiency: An organization is efficient if it utilizes the least amount of
resources necessary to obtain its products or services.
• Viability: An organization is viable ifit exists over a long period oftime.

Effectiveness is contrasted with efficiency. Effectiveness is doing the right

thing; efficiency is doing it right. Usually, effectiveness does not incorporate
efficiency: that is, an organization can accomplish its goals but be quite
inefficient in its use of resources. An efficient organization uses few
resources but may not accomplish its goal well. We want to design
organizations that are both effective and efficient, as both are likely to be
important for viability or long-term survivability. However, an organization
can survive for many years and not be known to be particularly effective or
efficient. The U.S. government is relatively long-lived but is not known for
effectiveness in many of its activities and certainly is not known for its
efficiency. These criteria represent different concepts that are desirable and
provide general guidance in selecting appropriate organizational configura-
tions and organizational properties.

1.3 An Information-Processing View of

Organizations
We live in a world where information, information exchange, and informa-
tion processing are of paramount importance. Companies are linked to their
customers and suppliers by numerous information networks. Telephone,
fax, and electronic mail are modern ways of communicating. The informa-
tion revolution (Business Week, June 13, 1994) is changing the world. The
development of cellular phones, multimedia systems, and interactive
television systems affects the way people and organizations operate and
work (ORIMS Today, June 1994, pp. 20-27). The Internet is growing; it
doubles every few months.
An organization processes information in order to coordinate and control
its activities. By processing information, it observes what is happening,
analyzes and makes choices about what to do, and communicates the above
to its members. Bounded rationality (March and Simon, 1958, Chapter 6)
1.3 An Information-Processing View of Organizations 5

is the underlying concept which makes organization necessary and informa-

tion processing is the mechanism by which organizations can obtain
coordination of their activities. Burton and abel (1994) summarize this
perspective with a focus on management science contributions to organiza-
tional studies.
Information processing is a way to view organizations and their designs.
Information "channels can be created or abandoned, and their capacities
and the types of signals to be transmitted over them are subject to choice,
a choice based on a comparison of benefits and costs" (Arrow, 1974). Both
information systems and individuals possess a capacity to process informa-
tion, but "this capacity is not, however, unlimited and the scarcity of
information-handling ability is an essential feature for the understanding
of both individual and organizational behavior" (Arrow, 1974). Innovations
in information technology change both the organization's demand for
information processing and its capacity for processing information.
In a small lumber firm, the CEO coordinated activities by reading the
mail in the morning when orders arrived. He processed the orders and
issued directions for the day. He was the central figure coordinating
activities in an environment in which orders and specifications have to be
in writing. He balanced the risk and centralized the information and
decision making in order for his firm to be effective and efficient. The
introduction of the fax machine changed how this firm operated. When
orders came in at any given time, information demand increased. The
demand for quick response grew because customers wanted offers
immediately by fax. Because modern fax systems allow customers to send
requests to more than one firm, competition is enhanced and becomes
fierce. The lumber firm's old centralized decision procedure no longer
worked; decisions had to be delegated, and rules for handling orders and
balancing the risk had to be developed. Changing the information flow
changed the organization.
Galbraith (1973,1974), in a seminal work, presented the organizational
design problem as an information processing problem: "the greater the
uncertainty of the task, the greater the amount of information that has to
be processed between decision makers" (1974, p. 28). The task uncertainty
can arise from the technology and environment (Thompson, 1967, p. 12), as
well as other sources. Various design strategies are developed. Following
March and Simon (1958), the mechanistic model increases information
processing capability to obtain integration of interdependent activities by
coordination by rules or programs, by hierarchical information processing,
and by coordination by targets or goals. If the information processing
demand comes from many routine and predictable tasks, then formalization
in form of rules and programs can increase the number oftasks that can be
handled.
6 Chapter 1. Introduction

When there are uncertainties associated with the tasks, then informa-
tion processing is referred up the hierarchy to a level where an overall
perspective exists. This is the traditional exception-based hierarchical
decision making. Such hierarchical decision making can handle only a
limited amount of uncertainty. Ifthe uncertainty exceeds the capacity of
the hierarchy, then targets or goals have to be set for the various tasks,
making them somewhat independent. The coordination has moved from a
procedure orientation to a results orientation. Organizations can either
reduce their need for information processing or increase their capacity to
process information (Galbraith, 1974, p. 30). The need for information
processing can be reduced by increasing slack resources: just-in-time (JIT)
inventory requires precise coordination; buffer inventory is an alternative.
Buffer inventory resources replace the need to process the information
required for JIT. Information processing needs can be reduced by creating
self-contained tasks: for example, a two product firm can create two self-
contained single-product divisions that need not communicate. Of course,
this strategy of reducing the need for information processing may incur
high opportunity costs from noncoordination. Buffer inventory is expensive,
and JIT eliminates that cost but may increase other costs. JIT increases the
likelihood of not having raw materials available when they are needed,
which incurs an opportunity loss. Single-product divisions may ignore inter-
dependencies in production or marketing, which may be costly in lost
opportunities. Thus, reducing information needs must be balanced with the
returns to coordinating the activities.
An alternative approach for the organization is to increase its capacity
to process information. In a hierarchical organization, the hierarchical
processing of information can be increased by investment in a vertical
information system. The demand for information processing capacity that
arises from uncertainty frees the organization to be able to react to
unforeseen events. In a highly specialized organization, it may be difficult
to have all specialized groups react in a coordinated manner. An informa-
tion system may increase the speed and amount of information that can be
exchanged. The introduction of satellites, information computer networks,
the internet and integrated CAD-CAM systems can increase the informa-
tion processing capacity ofthe organization. The type offeedback, amount
of information, and media richness may require face-to-face communica-
tions. Information processing capacity also can be increased by creating
lateral relations. Direct contact, liaison roles, task forces and permanent
teams are examples of strategies that will increase information processing
capacity.
The development of new information technology requires a revisite of
Galbraith's strategies. Interactive information networks, multimedia
systems, and generally the speed and amount of information that can be
1.3 An Information·Processing View of Organizations 7

processed should be taken into account. When Galbraith developed his

strategies, a liaison role was required; now a multimedia interactive
information system may be the answer.
Following Arrow (1974, p. 38), we argue that the definition of informa-
tion is qualitative. The basic design problem is to create an organizational
design that matches the demand for information processing with the
information processing capacity. To find an efficient organizational design
is an issue of the cost of information and information channels.
SAS (a case study discussed throughout this book) had a major change
in its information processing requirements in 1975. After the formation of
SAS and the creation oflATA (International Airline Transport Association),
the environment was relatively stable in terms of prices and demand.
Technology evolved steadily, except for the introduction of the commercial
jet engine. There were few surprises. SAS relied heavily on coordination by
rules and programs within a well-defined hierarchical authority and control
structure.
With the oil crisis of 1975 and dissolution of lATA, it was a new world
of increased uncertainty. SAS did not adjust until 1981, with increased
information processing and later with the creation of more self-contained
business divisions.
Many argue that both the information processing demands and the
capacity have increased recently but the cost of information processing has
decreased. There has been a reduction in slack resources, perhaps a slight
increase in self-contained units, a large investment in information systems,
and a large increase in lateral relations, which has led to "leaner and
meaner" organizations: less inventory, less equipment, and fewer employ-
ees, particularly middle managers. Those who remain use information
much more quickly and efficiently.
The information processing view of organization underlies many ofthe
arguments in this book. It gives managerial substance and relevance to the
contingency theory propositions in organizational design.

1.3.1 Neo-information processing as a basis for

Organizational design.
The information processing view of organization is well established, and
has been the basis for further development in the development of organiza-
tional models (Carley and Prietula, 1994). The development of the Virtual
Design Team (VDT) (Jin and Levitt, 1996) is an example where individual
actors in a project organization were modeled as information processing
units. The actors have different information processing skills and informa-
8 Chapter 1. Introduction

tion channels as well as different communication media. Organizational

behavior can be simulated to see the effect of deployment of organizational
resources, team building, and organizational structure. The original VDT
approach modeled large but relatively routine projects, and has recently
been extended to model semi-routine fast-paced project organizations
(Thomsen, 1998). In these models information processing is described over
time based on the individual's information processing capacity. Addition-
ally, the fit between the information processing capacity and information
processing demand over time is measured. This is micro view of the
information processing approach to organizational design for a specific set
of organizations.
In this book we take a general macro view ofthe information processing
approach to organizational design-a neo-information processing view . Here
the information processing demand on the organization is compared over
time to the information processing capacity ofthe organization. Addition-
ally, we posit that differences between the information processing demand
and capacity may arise due to more general phenomena than uncertainty
as presented by Galbraith(1973, 1974). Differences between the information
processing demand and capacity may arise due to uncertainty, but they
may also arise due to strategic choices. If an organization decides to change
its overall policy for meeting customer orders from a one week guaranteed
delivery to delivery within 48 hours, the information processing demand is
changed dramatically. To meet such a demand may require changes in the
information structure, organizational structure, and human capacity may
be needed. An advanced information processing system may be necessary,
formalization of procedures may be installed, limitations in the type of
orders accepted may be decided, and more delegation to better trained
individuals may be installed.
In the life cycle approach to organizational development, the information
processing demand changes as the organization passes through the various
phases of its development. Proper development of the information proces-
sing capacity over time is thus necessary for the organization to stay both
effective and efficient.
To develop a neo-information processing model of organization, we posit
that the basic work of an organization can be seen as information process-
ing: observing, transmitting, analyzing, deciding and taking action for
implementation. Information is the nerve system of organization and that
is our focus. We assume:

• individuals are rationally bounded (March and Simon, 1958),

• organizational work is partitioned, i.e., there is a division oflabor,
• information is costly to gather, transmit, store and analyze (Arrow,
1974),
1.4 Strategic Organizational Design 9

• information activities are not performed with perfect reliability,

• information is then scarce and imperfect,
• and thus, coordination is problematic,
and,
• individuals have self interest as well as organizational concerns.

These assumptions are the basis for an information processing view of

organization and are incorporated in the March and Simon's (1958) classic
work on organization, Galbraith's (1974) organizational design framework,
Williamson's (1975) transactions economics cost paradigm, and Tushman
and Nadler's (1988) information processing model of organization, to name
a few. Information processing provides an interpretative framework for
describing the world around us, the essence of positive science. Further, the
mechanisms of organization for design purposes are information processing
choices.
The information processing view is the basis of our thinking about
organization. For example, the observation that a large organization tends
to be decentralized can logically be supported from the information
processing view. From an information processing view, we argue that large
organizations need to decentralize in order to deal with the information
processing demands and mitigate for possible top management information
overload. Decentralization may also be decided from motivational reasons.
Better motivated employees have higher information processing capacity.
We integrate such views in our model by introducing organizational climate
dimensions.
The neo-information processing view is a frame to interpret research
findings whether they were generated from this basis or not. Thus, we shall
consistently interpret the positive science research literature on the
organization through information processing lenses. We then utilize the
research literature to support and enhance the information processing
model of organization for design purposes. Our goal is not to simply restate
and review the literature as it exists; our goal is to interpret the vast
knowledge base of organization theory for our own purposes: organizational
diagnosis and design utilizing an information processing view of organiza-
tion. This is the neo-information processing model of organizational design.

1.4 Strategic Organizational Design

The foundation for strategic organizational design is what the organization
does at the organizational level and not on a task level. The specific tasks
depend on organizational structure (Burton and Obel, 1984) and will thus
be a result ofthe design.
10 Chapter 1. Introduction

The goals and mission ofthe organization are the basis for the specifica-
tion of what the organization should do. If the goal is to run a profitable
gourmet restaurant, then its overall style, and the type offood and wine it
serves are what the organization does. The specific tasks by which the food
is prepared and served depend on the structure. From the goals, missions
and the particular environment a strategy has to be found. The strategy
could be to have a wide variety of different choices on the menu. But variety
also could be obtained by having only a narrow choice each night and
changing the choice every night or every week. The strategy affects the way
the restaurant is run, which affects structure, decision making, standard-
ization, and so on. Some specificity about what the organization does is the
basis for the design, but the level of specificity depends on the particular
organization. The next step in the design of an organization is to specify the
values for factors that effect the choice of the right organizational structure
and then have decision rules that can be applied.
Generally, the point of departure for the strategic organizational design
is the goals and mission of the organizational unit. The organizational unit
may be an independent owner-run organization, a corporation, or a
subsidiary or department. In each case, the design process begins by asking
what are the goals of the organization. From there on, the contextual
situation can be specified. The level of detail depends on the organizational
unit. If the organizational unit is a corporation the measures will be more
aggregate than if the organizational unit is a department. The boundary of
the organization depends upon the organizational unit specified. The
boundary for the corporation is the environment outside the organization:
the markets, the government, the customers, etc. The boundary for a
department is its environment and what is outside the department, which
will include other departments inside the organization. The organizational
unit, the environment and the boundary go together. Changing anyone
changes the others. We focus on the design ofthe organization within its
environment and boundary.
The strategic organizational design specifies the organizational
groupings-the units in the organization that have to be designed. Addition-
ally, the relationships between these units have to be designed. The organi-
zationwide information, measurement, control, and incentive systems are
part of the strategic organizational design.
Our main thesis is that good fit is necessary. The organization should
have a strategic organizational structure that matches its particular
situation. A good fit means better performance (Donaldson, 1987). Contin-
gencies may specify Ii structure, or a given structure may lead the
organization to find a situation that fits its organizational structure. For
example, as we discuss in more detail in Chapter 6, an organization may
choose a flexible structure that fits a turbulent environment, or it may
1.5 The Basic Features 11

manipulate its environment to reduce the turbulence. We also have seen a

number of smaller companies that deliberately decided not to grow because
growth would require a more elaborate decentralized structure. In those
cases, the CEO did not want to lose a position of influence and therefore
controlled the situation so that growth would not occur.
The strategic organizational design is a result as well as a process. The
contextual situation changes either because of exogenous factors or because
of deliberate decisions by the organization. Such changes in the contextual
situation may result in a need for a new strategic organizational design.
When an organization, for example, has growth as a goal, a continuous
change in the contextual situation is part of the goal.
This book develops a comprehensive theory of strategic organizational
theory and design. Our basic view is that an organization is an information-
processing entity; our model is an synthesis of many partial theories of
organizational design.
It is not our purpose in this book to review the various approaches to
organizational design and criticize the various views. Many excellent books
do that job (Scott, 1992; Hall, 1991). Our purpose is to develop an informa-
tion processing approach to organizational design which incorporates the
research and experiential knowledge from organization theory and design.
We have argued in various ways and sought strong support for our
statements. The resulting knowledge base is the basis for how we scrutinize
the organizational theory literature, analyze organizations and help
students learn about the fundamentals of organizational design.
We did not limit our efforts to developing a theory for strategic
organizational diagnosis and design: we also want it to be useful. We have
implemented the theory as a computer-based expert software system: the
Organizational Consultant. 1 Organizational Consultant is a decision-
support tool that can help managers, students and consultants diagnose
organizational problems and recommend organizational changes. Organiza-
tional Consultant has been used at many companies, universities, and
business schools.

1.5 The Basic Features

Organizational theory is a positive science that focuses on understanding
organizations. It is a multidisciplinary science, in which separate dis-
ciplines have created their own distinct questions, hypotheses, methodolo-
gies, and conclusions. These various views do not necessarily fit together
into a comprehensive and consistent view, nor do they necessarily offer re-
commendations on how organizations should be designed.
12 Chapter 1. Introduction

Organizational design, in contrast, is a normative science that focuses

on creating an organization to obtain given goals. Design "is concerned with
how things ought to be, with devising structures to attain goals" (Simon,
1981, p. 133). Organizational design depends on organizational theory: its
prescriptive purpose complements the descriptive function of the positive
theory. An organization is an information processing entity and the
organizational design has to match the demand for information processing
with the information processing capacity of the organization. The model
developed in this book takes its point of departure in an information
processing organizational design framework and augments it with what we
know from organizational theory. It tries to evaluate the normative
implications of descriptive analyses from organization theory.
To make decisions, the organization has to transmit and process
information. For example, when an order is received in a unit production,
information has to be transmitted to the department that purchases the
raw materials, to the production department, and to the accounting and
finance groups. Scheduling decisions have to be made so that production
can run efficiently and customer demands can be met. The organization has
to develop an organization with an information system that can meet the
demand for information processing. When the organization produces
standard products and orders come in regularly, planning may reduce the
information processing demand. However, orders that come in irregularly
and are very different need another organizational structure and informa-
tion system.

Figure 1.1. Organizational Context

1.5 The Basic Features 13

The basis for organizational design is the goals and mission. The goals and
mission determine the boundary of the organization relative to its
environment. Goals and mission influence the choice oftechnology as well
as size. These choices depend on the strategy. The strategy usually has the
closest link to the goals and mission. These relationships are summarized
in Figure 1.1.
The first step in the organizational design process is to establish the
contextual basis from the organizations' goals and mission. From this basis
the multiple contingency relationships have to be established. The multiple
contingency model states that the organizational structure depends on
multiple dimensions in the contextual situation. It was obvious from the
story about mM and AT&T that environment, management style,
technology, size, and strategy play an important role for the choice of a new
organizational structure.
The literature abounds with theoretical developments and empirical
evidence of the relationship between the contextual factors and the
organizational structure. However, in most cases the relationship has been
established between variation in one contextual variable holding the other
contextual variables constant. This provides only a very limited model of
the relationship between the contextual variables and the organizational
structure. In this book we develop a multiple contingency model construct-
ed from the bits and pieces in the literature.

Navigating the 7 C's of Organizational Design.2

Seven is a magical number: there are seven wonders of the world, seven days of the week,
memory chunks of seven plus or minus two and in management the seven S's. The later
suggests that there are seven elements of a successful organization which are integrated and
fit together with strategy, structures, systems, style, staff, skills and superordinate goals.
Seven also works for the organizational design framework: The 7 C's. They incorporate the
ideas of a consistent organizational design into a list of the important concepts: Configuration,
Complexity, Compliance, Centralization, Coordination, Communications, and Compensation.
Organizational design is to navigate the organizational ship through the 7 C's to maintain its
seaworthiness. Some of the C's exist in the model presented in Figure 1.3: configuration,
complexity, centralization; the others are C words which are similar in meaning to the traditional
words: compliance for formalization, coordination for meetings, liaison roles, rules, etc.,
communications for media richness and finally, compensation for incentives. The 7 C's are a
convenient way to help remember the comprehensiveness of what it takes to specify what an
organizational design includes. Navigation is necessary to maintain fit through the many
changes given by the changes in management style, climate, size, environment, technology
and strategy.

The contingency concepts and relations are summarized in Figure 1.2. The
contingency factors are on the left, and the design recommendation
possibilities are on the right. An effective and efficient organizational
design provides a good fit between the contingency factors and properties
and the structure of the organization. Additionally, we look at the multi-
14 Chapter 1. Introduction

dimensional contingency models in a dynamic perspective, where the

contextual factors and the organizational structure have to be aligned as
changes in one or the other happen over time.

The Contingency Factors for Properties and Structural

Configuration of the
Organizational Structure Organization

Structural
Configuration

Management Style Simple, functional, divisional, machine-

bureaucracy, matrix, Professional
bureacracy, and adhoc.
Climate

Size/Ownership

Environment
===> I Properties I
Complexity and differentiation
Technology Formalization
Centralization
Strategy Span of control
Rules
Procedures
Professionalization
Meetings
Reports
Communications
Media richness
Incentives
Fit Criteria:

Effectiveness
Efficiency
Viability

Figure 1.2. The Multi-contingency Model of Organizational Theory

The contingency theory of organization is the basis for the know ledge base
in our model. That know ledge base is gleaned from many literature sources.
The knowledge from the literature has been translated into meaningful
if-then rules and composed into a set of consistent rules. For example the
rule that states "If the organization is large, then decentralization should
be high" is a consistent rule. Overall, there are approximately 450 rules in
1.5 The Basic Features 15

the knowledge base, gleaned from a broad spectrum of the contingency

theory literature. As depicted in Figure 1.2, these if-then rules relate the
contingency factors of size, technology, strategy, environment, climate,
ownership, and management preferences to the organizational design
recommendations.
The definition of terms must be clear such that the user can respond to
the questions and implement the recommendations. The if-then rules in the
knowledge base should be a clear translation of our knowledge. The rules
must also provide an integration of the many partial theories. We have
chosen to use the structural properties shown in Figure 1.2 and then
translate the recommendations given in the literature into these concepts
using the neo-information processing perspective as a means in the
translation. This provides a basis for the development of a multiple-
contingency modeL We have followed similar attempts to combine knowl-
edge from the literature, such as Daft (1992) and Robbins (1990). In section
1.6 this process is discussed in more detaiL

1.5.1 Fit Criteria for Designing the Knowledge

Base
Fit is an organizing concept for the creation and development of the
knowledge base. Fit suggests a synthesis and integration of concepts and
ideas to create definitions. The challenge is to create a knowledge base
system that utilizes known theory for a given situation to suggest appropri-
ate organizational design recommendations. To meet all of these goals, the
knowledge base must fit together across a number of dimensions.
Here, we present four fit criteria that are necessary in order to obtain a
useable system (Khandwalla, 1973; Drazin and Van de Ven, 1985; Miller,
1992). In Figure 1.3, the contingency theory-organizational design model
is presented pictorially. This picture is a convenient means to illustrate the
four fit criteria needed for a useable organizational design consultant.
Those criteria are

• Contingency fit.
• Situation fit.
• Design parameter fit.
• Total design fit.
16 Chapter 1. Introduction

1.5.1.1 Contingency Fit

Each if-then contingency proposition must fit with the neo-information

processing framework and also be consistent with the contingency theory
and represent our knowledge well.

The Contingency Factors for Properties and Structural

Configuration of the
Organizational Structure Organization

Structural
Configuration

S Management Style
I Simple, functional, divisional, D
T machine-bureaucracy, matrix, E
U Climate professional bureaucracy, and adhoc S
A I
T CONTINGENCY FIT G

>
Size/Ownership N
I Properties
o
N Environment P
A
F Complexity and differentiation R
Technology Formalization A
I
T
Centralization M
Span of control E
Strategy Rules T
Procedures E
Professionalization R
Meetings
Reports F
Communications I
Media richness T
Incentives
Fit Criteria:

Effectiveness
Efficiency
Viability

Figure 1.3. Organizational Design Fit

A rule that states "If the organization is large, then the structure should be
centralized" does not fit what we know and should not be a part of the
knowledge base. Further, these design propositions must fit and be
consistent for each contingency across all design parameters. This is labeled
1.5 The Basic Features 17
contingency fit in Figure 1.3. The contingency fit criterion can largely be
achieved through careful attention to the contingency theory literature and
translation of that knowledge into appropriate if-then statements. Most
empirical research has emphasized how to obtain contingency fit for one or
more contingency factors. The studies reviewed in Chapters 3 through 8
show that contingency fit is important for the performance of the firm.

1.5.1.2 Situation Fit

The situation fit requires that the design situation is internally consistent.
That is, the situational facts, which give answers to the "if' questions, must
make sense. For example, an equivocal environment and a routine
technology do not fit. There is no recommended design for this situation,
and it is identified as a situational misfit. Something must be changed.
Here, the routine technology may be changed to a less routine technology.
This misfit says that we cannot recommend an organizational design that
can keep a routine technology viable in the face of an equivocal environ-
ment. The literature on strategy is abundant with research on this issue
(Quinn, Mintzberg, and James, 1988).
Situation misfits have to be dealt with. Alignment has to be established.
Situation misfits may appear due to changes in the environment and thus
be exogenous to the organization. Misfits may also appear because of
management decisions. A manager may change the size and create misfits
which then have to be adjusted to.
The control of situational fits and misfits is a key to organizational
success. In the organizational life cycle model where the organization
evolves from the small start-up organization to the multi-market-multi-
product organization, the organization has to pass through many faces of
misfits. Only those organizations that manage to create the proper misfits
and then resolve these will be successful.

1.5.1.3 Design Parameter Fit

For each design parameter, the set of if-then propositions that lead to a
design recommendation must fit and be in balance. For example, a design
recommendation that the organization should be decentralized can be
driven by a number of contingencies. Management style, climate, size,
environment, technology, and strategy all may suggest decentralization.
And the recommendation is strong. However, the more likely situation is
that there are design propositions that suggest decentralization and others
that suggest centralization. Here, the design propositions must be in proper
18 Chapter 1. Introduction

relative balance to obtain an appropriate recommendation. For example,

the management style may lead to centralization and the organizational
size to decentralization. The issue is to weigh the more salient recommen-
dation. The two statements must be in balance and fit our knowledge. The
certainty factor (d) approach3 helps access the design parameter fit. Ifthe
recommended design parameter has a high certainty factor, it is likely that
there is a design parameter fit. Also, the cfs on all possible recommenda-
tions can be inspected. If there are many different recommendations all
with high cfs, then the possibility for a design parameter misfit exists. The
organizational theory literature itself includes hypotheses on such possible
misfits. For example, ifformalization is low, then incentives should not be
based on process, appliance to standards, and procedures. Khandwalla
(1973) showed that design parameter fit would lead to higher performance.
The concept of design parameter fit is associated with the notion of
equifinality (Doty, Glick and Huber, 1993). The contingency factors may
recommend multiple acceptable properties ofthe organizational structure
but not all combinations are acceptable. For example, it may be recom-
mended that centralization should be high or low and that formalization
also should be high or low. However, only two ofthe four possible combina-
tions may be acceptable. Most likely only an organizational structure with
a high centralization and a low formalization or an organizational structure
with a low centralization and a high formalization will be acceptable.

1.5.1.4 Total Design Fit

Total fit is the most demanding of all. It assumes that the contingency fit,
the situation fit, and the design parameter fit criteria have all been met.
And now, the total fit criterion requires that the design recommendations
fit together internally and, more important, fit the actual situation. That
is, the recommendations must be useable. Miller (1992) discusses the
problems of simultaneously obtaining contingency fit and design parameter
fit. The problems he encounters are partially due to different definitions of
concepts in the various contingency paradigms. A problem that is solved in
the model developed in this book. However, even with a situational fit and
a contingency fit, design parameter fit may not be present. This is due to
the fact that each contingency relationship may lead to more than one
design recommendation. The right set or combination then has to be chosen
to obtain total design fit. Note also that total design fit is unobtainable if
a serious situational misfit exists. Ifthe organization has put itself in a bad
position, then no appropriate organizational design exists.
1.6 Creating a Knowledge Base from the Literature 19

1.6 Creating a Knowledge Base from the

Literature
The organizational design model follows the "if-then" format of production
rules in expert systems. Since most knowledge in organizational theory and
design can be stated as simple if-then propositions, it seemed natural to
adopt such a representation. This book is an expert system in itself; the
rules developed are put in a form that makes it relatively easy to analyze
and diagnose organizational design problems.
To illustrate, the procedure and model the next section take the well-
known models of Duncan (1979) and Perrow (1967) and restate them as
rule-based expert systems. The issues of translation and composition are
discussed, and a contingency theory rule-based knowledge base is briefly
presented. Several partial theories ofthe contingency imperatives of size,
strategy, technology, and so on are given. A comprehensive treatment of the
various paradigms and their composition is given in the next chapters. The
knotty problem of composing the several partial theories into a consistent
and comprehensive knowledge base within an information processing
framework is analyzed. Unfortunately, the literature is rather silent on how
to put the pieces together. Finally, we discuss the implementation of our
rules in the Organizational Consultant expert system.
Duncan (1979) and Perrow (1967) use different organizational theory
models that can be translated into if-then rules. First, we demonstrate the
translation of each into a knowledge base. Second, we consider how to put
the separate knowledge bases together into a combined knowledge base.
This illustrates the complexity of composing a knowledge base from a large,
diverse, and sometimes inconsistent organizational theory literature. We
begin by stating Duncan's (1979) model of environmental contingencies as
a rule-based knowledge base.
Duncan's knowledge base can be stated as the following six if-then rules:

• If environmental complexity is simple

and environmental change is static,
then organizational structure is functional.
• If environmental complexity is simple
and environmental change is dynamic,
then organizational structure is mixed functional.
• If environmental complexity is complex
and environmental segmentation is yes
and environmental change is static,
then organizational structure is decentralized.
• If environmental complexity is complex
20 Chapter 1. Introduction

and environmental segmentation is yes

and environmental change is dynamic,
then organizational structure is mixed decentralized.
• If environment complexity is complex
and environmental segmentation is not
and environmental change is static,
then organizational structure is functional.
• If environmental complexity is complex
and environmental segmentation is not
and environmental change is dynamic,
then organizational structure is mixed functional.

Duncan himself presents the equivalent of the above rule-based knowledge

base in the form of a decision tree rather than if-then statements. This
knowledge base is in fact a small expert system. Consider the goal of this
expert system; it is to recommend an organizational design that is stated
as either functional, mixed functional, decentralized, or mixed decentral-
ized. These are the possible answers to complete the statement "then the
organizational structure is" as given in the six rules above. The recom-
mended structure is thus dependent or contingent on the environment, as
stated in the "if' part of each rule. Duncan describes the environment along
three dimensions-complexity, change, and segmentation. Environmental
complexity is either simple or complex, environmental change is either
static or dynamic, and environmental segmentation is either segmentable
or not. An organization can find itself in 2x2x2 =8 possible environments.
The knowledge base is the set of six if-then rules that links the environmen-
tal condition with the recommended organizational design, which is
contingent on the organization's environment.
To illustrate Duncan's expert system, consider an airline company in the
1960s with a static and simple environment. Using the if-then rules in the
knowledge base, we find that the recommended structure is functional.
Today the environment is complex, probably not segmentable, and certainly
dynamic. The recommended structure is now a mixed functional struc-
ture-functional with lateral activities. But if the environment were
segmentable, then the structure should be mixed decentralized- decentral-
ized with lateral activities. The recommended structure is dependent on the
environment and is relatively sensitive to shifts in the environment; as the
environment changes, the recommended structure changes as well.
We now turn to a Perrow's (1967) model of technology contingencies.
Like Duncan's model, it can also be translated into a knowledge base of
if-then rules. The goal is to recommend an organizational design from the
choices: routine, engineering, craft, or nonroutine. The organization's
technology has two dimensions: task variability, which can be routine or
1.6 Creating a Knowledge Base (rom the Literature 21

high, and problem analyzability, which can be ill defined or analyzable.

There are then 2x2 = 4 possible technologies, which are the facts for the
expert system.
Perrow's knowledge base can be stated as the following four statements:

• If task variability is routine and

problem analyzability is ill defined,
then organizational structure is craft.
• If task variability is high and
problem analyzability is ill-defined,
then organizational structure is nonroutine.
• If task variability is routine and
problem analyzability is analyzable,
then organizational structure is routine.
• If task variability is high and
problem analyzability is analyzable,
then organizational structure is engineering.

Perrow developed this know ledge base by describing and analyzing the real
world. Here, these propositions serve as normative statements as to how
organizations ought to be structured in order to become efficient. This
approach takes the knowledge of positive science and uses it in a normative
system that recommends how organizations should be structured. To take
the step from a descriptive form to a normative form presumes a causal
relation and further it is a desired relation that will yield a better-
performing organization. To justify such steps, we have tried to make sure
that the relation is found generally in the literature and not restricted to
single study and that an argument for the causal relationship could be
found using the information processing view of organization.
Above, Duncan's (1979) and Perrow's (1967) contingency models were
translated into separate knowledge bases. In Duncan's system, the
organizational design is dependent on the organization's environment; in
Perrow's system, the organizational design is dependent on the technology.
Each system is self-contained and independent. Yet the reader must
somehow feel that each one is incomplete and limited. Can the two systems
be put together so that the composed system is more complete and more
practical?
Both the environment and the technology are important contingencies
in design (Daft, 1992; Robbins, 1990). But the process of putting the two
systems together is not at all obvious. Each approach has its own vocabu-
lary and definitions, and the relations among the definitions are not clear.
How do Duncan's functional and decentralized structures combine with
Perrow's routine, engineering, craft, and nonroutine structures? Are
22 Chapter 1. Introduction

Duncan's four organizational designs simply different names for Perrow's

four designs? For example, is a functional organization the same as a
routine organization, or are the two sets of recommendations independent?
Creating a comprehensive knowledge base is a design problem itself; it
is a synthesis of the knowledge from many different sources. How do we put
the pieces together so that the resulting design is a mechanism that meets
its purpose?
Taken separately, Duncan and Perrow's systems individually are
logically complete. Consider the nature ofthe Duncan system. There are six
if-then statements. Each is stated without equivocation. For example,
consider this:

• If environmental complexity is simple

and environmental change is static,
then organizational structure is functional.

This statement depends on two antecedents: environmental complexity and

environmental change. Either the environmental complexity is simple, or
it is not. And either the environmental change is static, or it is not. Ifboth
antecedents are factually met, then the organizational structure is
functional or, interpreted normatively, should be functional. There are no
maybe's, probabilities, or equivocalities of any kind. If the two conditions
are met, then the result is certain. If either condition is not met, then the
statement does not apply; it is not "fired."
Duncan's six statements are identifiable; each applies or it does not, and
it applies with certainty. Further, the six antecedents statements are
mutually exclusive, so no two statements can be fired in a given situation.
There cannot be a conflicting design recommendation. And further,
assuming any environment can be categorized into one of the six probable
environments, the Duncan system is comprehensive.
In brief, Duncan assigns all possible environments to six mutually
exclusive categories, each of which has a certain design recommendation.
Duncan says nothing about technology, size, managerial style or strategy.
It is implicit that these contingent factors can take on any value. That is,
Duncan's design recommendations do not depend on whether the size is
small or large, the technology is divisible or not, the managerial style is
hands-on or not, or the strategy is prospector or not. These contingencies
are not relevant. Despite the reasonableness of Duncan's statement, there
is much evidence that his system is normatively incomplete.
Similarly for Perrow, technology is categorized into one of our four
mutually exclusive categories, each of which has a certain design recom-
mendation. There is no consideration of size, strategy, managerial
preference, or Duncan's consideration of environment. Perrow's approach,
1.6 Creating a Knowledge Base from the Literature 23

too, also must be incomplete. Both Duncan and Perrow might argue that
design was not their purpose. Each was describing the world: testing
hypotheses in the best of positive science tradition. But our goal is to go
beyond positive science and to use its knowledge to create a practical aid for
designing organizations.
The synthesis of Duncan's and Perrow's knowledge bases can be simple
or enormously complex. First, the simple approach is to consider the two
systems as independent. That is, for any organization, the environment can
be categorized according to Duncan and the technology according to Perrow.
Then, a design recommendation could be a simple combination such as
Duncan's recommendation of a functional organization plus Perrow's
routine organization. But do these fit together? What does it mean? We also
could have a functional organization and a craft organization. There are
6x4 =24 possible design outcomes-each of which must be realizable. AB for
each system separately, these recommended designs are stated with
complete certainty.
The second approach involves dependency and integration. One kind of
dependency could be definitional. Are Duncan's organizational designs
simply different words for Perrow's concepts, or are they different aspects
of organization, as assumed above? Let us assume they are different
aspects of organization but are not independent. Consider again Duncan's
if-then proposition:

• If environmental complexity is simple

and environmental change is static,
then organizational structure is functional.

Let us assume that this statement is compatible with Perrow's routine

organization but not compatible with a nonroutine organization. One
approach is to simply add a third antecedent to Duncan's statement; this
approach becomes unwieldy very quickly. A more fruitful approach is to
recognize that Duncan's statement is not true universally for other
conditions that may exist. That is, we hold the statement with a certainty
less than total. But by how much?
Unfortunately, the literature does not give excellent guidance, and there
is considerable incompleteness and ambiguity about the strength of these
dependencies. Our knowledge base relies heavily on the positive science
literature and empirical studies that usually involve hypothesis testing:
analysis of variance models, multiregression models, and so on. The
explained variance is usually not high, and this is a major clue that the
models explain only a small amount of the phenomena under study.
Referring to our fit criteria each model states a contingency fit for one
contingency variable.
24 Chapter 1. Introduction

1.6.1 Composing the Knowledge Base

The construction of a knowledge base for an expert system is a statement
of our knowledge about organizational theory. The above rule statements
are clear and can be applied one by one, but they are yet to be composed
into a consistent and coherent set of rule statements. Returning to the
above discussion of the Duncan and Perrow knowledge bases, each one is
individually clear, but can we put them together into a consistent and
coherent system to obtain design parameter fit? It is not obvious how to
compose the two independent systems into an integrated expert system,
even if we have resolved the problem of concept definitions. The composi-
tion of the rules is essential to the development of the knowledge base.
There are a number ofways to put the pieces together. In the area of expert
systems, for example, fuzzy logic, Bayesian theory, and certainty factors
(cfs) have been used to combine facts from a number of production rules
(Harmon and King, 1983, O'Leary, 1996).
In our development we have used the certainty factor principle, which
worked well in our Organizational Consultant expert system. This is not to
say that other systems would not function well. We have presented only the
production rules in pure form in Chapters 3 through 8-that is, without
associating certainty factors to the rules. In Chapter 9 we discuss how to
combine the rules both formally and informally.
We now illustrate the use of the certainty factor principle to compose the
knowledge base. We capture its essence with a few sample rules. The rules
are meant to illustrate how we translated the knowledge from the
literature into if-then rules for the knowledge base. We begin with size as
a contingency factor and then consider other contingencies. In Chapters 3
through 8 we state more generally all the propositions that have formed the
basis for our rules.
It is generally accepted that the size of the organization affects its
structure. The literature is replete with support (Blau and Schoenherr,
1971; Pugh, Hickson, Hinings, and Turner, 1969) and counter argument
(Aldrich, 1972). We have taken this idea and translated it into rules that
state that the size of the organization should affect its structure. To
illustrate, one rule in our knowledge base is:

• If size is large,
then formalization should be high (cf20).

The size factor is then a fact. An organization is large, or it is not. If it is

large, then the recommendation is that the formalization of the organiza-
tion should be high with written and well-defined jobs, regulations, and
work standardization and relatively less freedom on the job for employees.
1.6 Creating a Knowledge Base from the Literature 25

The certainty factor of 20 is a qualifier in the statement (certainty factor

measures the degree of belief one has in the statement and can range from
-100 to 100). A cf20 implies that the statement has a relatively weak effect
but cannot be ignored. A stronger statement would increase the certainty
factor, and cf100 would be total certainty. Negative certainty factors reflect
disbelief. Hall, Hass, and Johnson (1967) argue that size is important but
is not the only determinant of structure; hence, we choose a relatively weak
effect. Size affects not only formalization but also organizational complex-
ity, centralization, configuration, and so on, which are incorporated in the
knowledge base.
The technology is another determinant of the appropriate structure. As
discussed earlier, Perrow (1967) investigated the relation between
technology and structure. We include the effect of the technology on the
structure. To illustrate, one rule is given:

• If technology is routine,
then organizational complexity should be low (cf20).

The organizational complexity is the degree of horizontal or vertical

differentiation in the organization. Robbins (1990) indicates that there is
empirical evidence but it is qualified. Hence, the certainty factor is 20. The
goal is to recommend the appropriate level of complexity for the organiza-
tion. If technology is routine, then the organizational complexity is recom-
mended to be low.
The organization's strategy is another determinant of organizational
design. In 1962, Chandler stated his now famous proposition that
"structure follows strategy." In our knowledge base, we use Miles and
Snow's (1978) hypotheses about strategy and structure. The organization's
strategy can be categorized as: reactor, defender, prospector, or analyzer.
A prospector's "domain is usually broad and in a continuous state of
development" (Miles and Snow, 1978, p. 56). To illustrate, one rule from the
knowledge base is:

• If the strategy is prospector,

then centralization should be low (cf20).

The rationale for the rule statement is that the prospector who has a large
number of diverse activities requires a decentralized structure; if not,
organizational activity tends to slow down because of decision bottlenecks.
The statement is qualified since the strategy is not the only determinant of
the structure; hence, the certainty factor is relatively low, as in previous
examples.
26 Chapter 1. Introduction

The environment is an additional determinant of organizational design.

Duncan (1972) provided early empirical evidence. For our knowledge base,
one rule is:

• Ifthe environmental uncertainty is low,

then the centralization should be high (cf 20).

Duncan's (1979) model is stated differently but provides support. Robbins

(1990) argues that centralization is possible when the environment is
stable, since there is time to process requisite information, but again, it is
a qualified statement.
Environmental hostility (Robbins, 1990) is also a determinant of the
structure. Greater hostility requires greater centralization of the organiza-
tion. To illustrate, one rule is:

• If the environmental hostility is extreme,

then the centralization should be high (cf 40).

We are more certain that an extremely hostile environment calls for a

unified effort; hence, the certainty factor is 40.
In the contingency framework, is management style important? Child
(1972) argues that the environment and technology leave some discretion
for managers; they have some choices concerning the organizational design.
We include rules that relate management's preference for microinvolve-
ment in decision making as a determinant of the structure:

• If management's preference for microinvolvement is high,

then centralization should be high (cf 40).

That is, a greater desire for microinvolvement by management indicates

greater centralization. A certainty factor of 40 indicates some qualification,
but is a strong recommendation.
In the above rules, we have included the influence of size, technology,
strategy, environment, and management style on the structure: formaliza-
tion, centralization, and so on. Many of the other rule statements are
similar in form and related in content to those given.
Compound rules, which have two or more contingent conditions
required, are also part of the knowledge base. This, in part, takes care of
some interaction effects CSchoonhoven, 1981; Pennings, 1987). To illustrate,
one rule requires both high organizational complexity and nonroutine tech-
nology:

• If the organizational complexity is high

1.6 Creating a Knowledge BaBe from the Literature 27

and the technology is not routine,

then the horizontal differentiation should be high (cf 60).

If both conditions are met, then horizontal differentiation should be high

with considerable confidence-a certainty factor of 60. (The Duncan and
Perrow knowledge bases, as represented earlier, contain only compound
rule statements.) The interaction effect may also be modeled by relation-
ships between various structural dimensions.
In our knowledge base, we also include rules that caution the user but
do not give a suggested structure. This is a statement about a situational
misfit. To illustrate, one rule is:

• If the strategy is prospector

and the technology is routine,
then this may cause problems.

Intuitively, one can judge that a prospector strategy and a routine

technology are not compatible. The expert system leaves open any
recommendation for resolution of this situation. It demonstrates that the
system is limited in its recommendations. These rules are representative
of the 350-plus rules in the knowledge base. The rules have been translated
from the literature into a set of if-then statements. The if-then rules are
statements that specify when contingency fits are obtained for each
individual contingency factor. We now turn to the composition ofthese rules
where the contingency factor influences have to be balanced and the conflict
in influences resolved. The certainty factors are used to compare each
statement and summarize our understanding and knowledge. The resulting
recommendation states how design parameter fit can be obtained and the
strengths of the recommendation. A high certainty factor specifies that
design parameter fit has been obtained. A low certainty factor specifies that
there may be design parameter misfits. Consider two knowledge-base
statements:

• If size is large,
then decentralization is high (cf 30).
• If the strategy is prospector,
then decentralization is high (cf 20).

Now let us assume that both antecedents are true-that is, the size is "large"
and the strategy is "prospector," then the rule for combining the conclusion
on the structure is given by:

• Decentralization should be high (cf 44),

28 Chapter 1. Introduction

where the calculation follows the rules from the MYCIN concept4 (M4,
1991). Here, we have 30 + (100 - 30)x20/100 =30 + 14 = 44.
The resulting conclusion on decentralization is stronger than either
statement would be alone. These two rules concerning size and strategy
have a positive relation with decentralization. Each contingency factor adds
to the conclusion, but neither contingency factor is sufficient to make a
certain conclusion or recommendation. Greater confirmatory information
leads to stronger certainty of the recommendations. This is consistent with
the concept that additional confirmatory information has positive marginal
effect.
We also may have knowledge that tells when a relationship is not
positive. For example, one rule might state, "If structural complexity is low,
then decentralization should not be high." We can incorporate the negation
by using a negative certainty factor:

• If structural complexity is low,

then decentralization should be high (-cf 30).

By combining all three statements, we find that decentralization is high (cf

20). Thus, positive and negative effects can be appropriately included. The
negative certainty factors are used to obtain design parameter fit and thus
sort multiple recommendations.
In this way, we are combining the various contingency theories into a
unified and consistent statement of our knowledge. The literature is rather
silent on the appropriate combinations (although there is considerable
evidence for the separate contingency factors to determine the appropriate
structure). Validation ofthe knowledge base requires a continuing revision
through application in real situations and cases, and we discuss this issue
later in this chapter.
It is important here to note that the sequence in which the information
is collected and applied does not change the results. If we first know that
decentralization should be high (cf 20) and then later learn from another
rule that decentralization should be high (cf30), then the calculation is as
follows:

20 + (100 - 20)x30/100 = 20 + 24 = 44.

The result is the same, independent ofthe order. It is, however, interesting
to note that the marginal contribution to results does depend on the
sequence. In the first situation the rule that stated that decentralization
should be high (cf30) contributed with cf30. It dropped to 24 in the second
case. This can be interpreted as a decreasing marginal utility of informa-
tion. If you do not know anything about the organizational structure, then
1.6 Creating a Knowledge Base from the Literature 29

just getting some infonnation is ofhigh value. But ifyou already know from
other contingencies that the centralization should be high with a certainty
factor of 80, then the most likely recommendation will be high centraliza-
tion no matter how much more you learn. Therefore, the importance of the
extra confinnatory information has a lower value in this case.
From an infonnation processing point of view, each contingency factor
demands a degree of information processing capacity. The environment
may demand a high infonnation processing capacity, while technology
demands low infonnation processing capacity. The result is not necessarily
a medium demand for information processing. There is a requirement for
a high infonnation processing capacity related to the environment and a
low capacity related to technology. Medium infonnation processing capacity
in both areas may be both too little and too much. The certainty factor
principle will find the dominate demand for infonnation processing but also
summarize demand in directions other than the dominant one. The expert
system approach also lets you explain why a particular result was
composed, giving the full background for evaluation. The certainty factor
principle has proved to be an efficient tool for analysis and design. A trace
of rules that are fired to provide a recommendation will show where the
demand for infonnation processing capacity originates. In fact the expert
system can be designed without any use of certainty factors. The expert
system can then find the rules that apply to the particular input and then
sort these in groups that lead to similar conclusions. The final balancing
can then be done by the individual based on his or her knowledge or
experience or view of the particular situation. In our expert system,
Organizational Consultant, we have, however, used certainty factors, and
that has proved useful and beneficial.
Our model can be used without complete infonnation about the
situation. But the more facts that are known, the more likely it is that the
knowledge base will provide a recommendation that has a high certainty
factor. A high certainty factor indicates that many contingencies pull in the
same direction and that little conflict about the demand for information
processing exists. A low certainty factor can have multiple causes. A low
certainty factor may come from lack of facts or from facts that give
conflicting recommendations. The lack offacts can either be contingencies
where you have no facts or contingencies where you are unsure about the
value. You may think that the environment is uncertain, but you may want
to discount the effect of this infonnation by assigning a low certainty factor
to the premise in the rule. A certainty factor assigned to a fact in a rule is
given in parenthesis as shown in the rule below:

• If the environmental uncertainty is low (cf20),

then centralization should be high (cf20).
30 Chapter 1. Introduction

The above rule will result in a recommendation that centralization should

be high (cf (20x20)/100 =4).
In most cases complete information is not available. As shown above,
there is diminishing marginal value of additional information. In many
cases it may be too costly to obtain complete information, even if it were
possible.
In summarizing, contingency theory suggests that an appropriate
organizational design is contingent or dependent on such factors as size,
strategy, technology, environment, and leadership (Penning, 1992). There
is a large literature (e.g., Daft, 1992; Robbins, 1990) that supports this
view . Yet the composition and integration into a comprehensive contingency
theory of organization is still ad hoc. In our model such a view can be
incorporated by assigning a certainty factor to a fact. In Organizational
Consultant this is part of the input to the expert system. The certainty
factor value associated with the fact then changes the strength of that
contingency factor. Ifwe assume that size is large with a certainty factor
50, the result from the above statement about the relationship between size
and decentralization would change from a recommendation that decentral-
ization should be high (cf 30) to a recommendation that decentralization
should be high (cf 15 (30x50/100)). The certainty factor associated with a
fact can therefore either represent a belief about the importance of the
contingency factor, or it can represent that the there are uncertainties
about the true value of the contingency factor and therefore the strength is
diminished.
We now turn to the strength of these design rules. A larger certainty
factor indicates a stronger belief in the design recommendation; a smaller
certainty factor indicates a weaker recommendation. A negative certainty
factor represents disbelief. A certainty factor of 100 or -100 is a recommen-
dation and zero is no recommendation. There are many factors that go into
assigning a certainty factor. A certainty factor of 60 or greater is a very
strong statement. Between 30 and 60, the recommendation is a strong
recommendation but not mandatory and is not sufficient by itself to be a
strong recommendation. Thus, other contingencies must further support to
obtain a strong recommendation. Above 80, the recommendation is almost
certain (Harmon and King, 1985).
The statement of the if-then rules and the composition of these is the
specification of the way relationships are modeled and as shown in Figure
1.2. The certainty factor principle is used to specify the strength of the
statement and the composition of the rules. A high certainty factor
represents a fit-both a contingency fit and a design parameter fit.
1.7 The Organizational Consultant for Designing an Organization 31

1.7 The Organizational Consultant for

Designing an Organization
So far, we have discussed only the creation of the rule-based knowledge
base that represents our knowledge of the theory of organizational design.
We have implemented the above model in the Organizational Consultant5
expert system, which diagnoses and designs organizations. Here we want
to use our knowledge about organizations to specify appropriate organiza-
tional structures and properties for given organizational situations, as
shown in Figure 1.2. So Organizational Consultant wants to determine
values for the structure and properties. The user has to enter the facts
about the organization's situation-that is, answers for the contingent
circumstances.
Organizational Consultant analyzes the current organizational structure
using the many facts related to the functioning of the organization. The
structure is then described in terms of the configuration and its properties.
Some inputs such as strategy, size and management style are determined
by independent expert systems. Based on the input, the system recom-
mends the configuration and structural properties that give the best fit
with the specified situation. The situation itself is analyzed and possible
situational misfits are given. Finally, the current and prescribed organiza-
tional structure are compared, and possible changes recommended. The
system allows the user to change input values and rerun the consultation,
thereby providing a way to perform sensitivity analyses.
Analyzing organizations involves four steps:

1. Specify the organization's situation in terms ofthe particular organiza-

tion analyzed.

2. Transform the specification under step 1 into general concepts usable for
design purpose.

3. Provide general recommendations based on the specifications under step

2.

4. Translate the general recommendations to specific recommendations for

the particular organization.

This book is about the knowledge we have to move from step 2 to step 3. It
shows how general knowledge from a number of sources can help guide the
particular organization in its choice of a proper organizational design.
32 Chapter 1. Introduction

The expert system provides decision support in moving from step 1 to

step 4. Decision support for the other steps can also be created, but are not
discussed in this book. The expert system does not make any decisions. It
provides a methodology and a structure in moving from step 1 to step 4
using the existing knowledge on organizational theory intelligently. The use
of Organizational Consultant to solve cases is discussed in Part 2 of this
book. The recommendation from Organizational Consultant has proved
useful as will be discussed in the next section.

1.8 Validating Theory for Application6

The validation process used on Organizational Consultant relies on
information obtained from cases, consultation with executives, dialogue
with experts, and executive M.B.A. student assignments. The classic work
on the validity issue in social science is Campbell and Stanley (1963). Cook
and Campbell (1976) later developed four concepts of validity: internal
validity, statistical conclusion validity, external validity, and construct
validity. The validation of an expert system involves the same general
principles of validation, with some differences in detail. The issue is well
discussed in a number of works (Blanning, 1984; Berry and Hart, 1993;
Preece, 1990).
The principles have been stated in various ways. Feldman and Arnold
(1983) define validity in terms of content, construct, and criterion-related
validity. They need answers to these questions: "Does it make sense to a
group of experts, is it measuring the underlying characteristics, and is it
related to real-world intent?" McGraw and Harbison-Briggs (1989) present
a variety of verification and validity techniques. Hayes-Roth, Waterman,
and Lenat (1983) consider validation to be a major concern in developing
expert systems.
The expert system designs organizations that fit well-defined sets of
circumstances. It is a normative model that shows what a segment of the
world ought to be like. The validation process of such a system should be
expected to differ from that of a positive model. For example, the correct-
ness of the predictions of the positive model would be replaced by the
correctness of the "predictions" of the expert system-whether the design
recommendations are reasonable and helpful. Validating the expert system
involves showing it to be correct, consistent, comprehensive, relatively
complete, and operational. The goal of the validation is to inspect whether
the model provides designs that are efficient, effective, and viable (Figure
1.2).
1.B Validating Theory for Application 33

Table 1.1. Comparing Validation Approaches

Feldman and Cook and Campbell

Arnold (1983) (1976) O'Leary (1988)
Content Internal validity A priori

Analyze the knowledge base for

Statistical validity • Accuracy,
• Completeness,
• Base weights.

Construct External validity In situ

Construct validity Test the inference engine.

Criterion-related Analyze the condition-decision

for decision quality.

Analyze the condition-decision

to determine if these are the
right answers for right reasons.

O'Leary (1988) suggests that there are six possibilities for analyzing an
expert system: "Analyze the knowledge base for accuracy, analyze the
knowledge base for completeness, analyze the knowledge-base weights, test
the inference engine, analyze the condition-decision matches for decision
quality, and analyze the condition-decision matches to determine whether
the right answers were found for the right reasons" (p. 75). No one of these
six possibilities is sufficient to validate the expert system. Yet these
questions do address the fundamental validity issues and are related to the
social science validity operationalizations (albeit in a somewhat different
form). In Table 1.1, these validation criteria are juxtaposed with others
developed by Feldman and Arnold (1983) and Cook and Campbell (1976).
Each validation approach begins with the model itself and its content and
then considers its relevance within the context of the problem. McGrath,
Martin and Kulka (1982) suggest a triangulation approach: cases, experts,
and students. Assuring the validity of Organizational Consultant by using
the results of its application in cases and real-world situations is analogous
to Davis' (1984) list of expert behaviors: solve the problem, explain the
result, learn, restructure the knowledge, determine relevance, and degrade
gracefully. This is learning by doing.
Expert system validation can be considered as a process or a prod-
uct-that is, how to do validation and the results you obtain. There are at
least three ways to discuss validation: an application of validation criteria
34 Chapter 1. Introduction

(e.g., O'Leary's to the expert system), the product of validation (i.e., what
changes were made to the expert system), and the process of validation.
Our interest in the validation process concerns what to do, why do it, and
what can be learned.

1.8.1 The Foundation Expert System

Validation of an expert system is an ongoing, never-ending process that is
realized through application and experience. A knowledge-base expert
system is validated, and hopefully improved, the same wayan individual
becomes an expert: through practice and experience. The object is therefore
to develop a validation process that specifies what to do, how to learn, and
what can be learned. The process of validating Organizational Consultant
described below addressed all three aspects. Throughout the process, the
expert system continues to offer better organizational design solutions. But
however much we improve it, it remains an imperfect expert system that
requires a skilled and experienced user to apply it in a reasonable fashion.
The model embodied in the system is the result of the synthesis of
partial theories in the literature, subsequent theorizing, and the transfor-
mation of positive models into normative ones, using an information
processing framework. In earlier research (Baligh and Burton, 1981;
Burton and Obel, 1984), it was suggested that decision rules (that is, if-then
conditional imperative statements) could be developed and applied to
recommend appropriate organizational designs. The tasks of both develop-
ing the requisite number of rules and applying them are involved and
laborious. The use of a "logic machine" makes the application much easier
and helps in the process of modifying the set of rules first developed by an
expert. Knowledge-base expert systems supply the logic machine needed.
Thus, a knowledge base for organizational design could be systematically
developed within a number of different logic machines or shells.
The model began with a set of rules we had developed earlier in
production rule form (Baligh, Burton and Obel, 1987). To these we added
partial theories from the literature. Sometimes these were normative;
sometimes we had to derive normative forms from positive or descriptive
ones.
1.B Validating Theory for Application 35

1.8.2Validation and Continued Model Develop-

ment
The first working expert system was the result of literature models and
new theorizing only. Once running, the system produced designs that were
used to modify and improve it. The validation process we describe is based
on the work of O'Leary (1988). O'Leary had six recommendations: (1)
analyze the knowledge base for accuracy, (2) analyze the knowledge base
for completeness, (3) analyze the knowledge-base weights, (4) test the
inference engine, (5) analyze the condition-decision for decision quality, and
(6) analyze the condition-decision matches to determine whether the right
answer was found for the right reasons. We call the first three a priori
criteria and last three in situ tests.
The a priori criteria demand that the contingency-design relations (See
Table 1.1) be stated as accurately as possible in the production rule
knowledge-base format. Briefly stated, the basic theory and its rules were
developed by the authors independently, who then compared their rules.
The bases of their efforts were literature models and their own theoretic
work and models. The result was a set of prescriptive if-then rules mapping
facts into design variables. These rules were reviewed for accuracy,
consistency, and completeness by the authors. The creation of design rules
is relatively straightforward. Crafting a consistent and coherent set ofrules
is an ongoing and difficult job. O'Leary's third criterion, analyzing the
knowledge-base weights, is a formidable task. There are two ways to
consider the base weights, or certainty factors, attached to the if-then rules.
These factors may refer to the relative strength of the various contingency
statements made in the literature and the authors' models. Consider these
two statements: "If size is large, then decentralization is high" and "If the
strategy is analyzer, then decentralization is medium." The issue is to
assign certainty factors to each of these rules to reflect their relative
strengths or importance to the goals of the organization, both separately
and collectively. The knowledge-base weight may refer to the strengths of
the arguments that support the rule-that is, the quality and correctness of
the rule. It is this latter component that is the object of an in situ test.
(Testing the inference engine was not a major issue.) These issues were
considered using cases, working with executives and students.
O'Leary's criteria were not applied in strict order; we moved generally
from a priori c-onsideration to in situ ones. The modifications were to adjust
certainty factors, add rules, and drop rules (Figure 1.4). We reduced
O'Leary's six criteria to a two step process of (1) "basic theory" or a priori
careful specification of the expert system, and (2) in situ experience and
modification from use.
36 Chapter 1. Introduction

START HERE

apriori : First step

create prototype

Modify if·then statements • aset of if·then statements I'.ith cis from the
ad~st cfs organization theory literature

Critique the prototype for possible changes

in situ: Next steps

• Develop test cases

• Test with executives
• Test cases with experts
Modify if·then statements • Students test in our organization
Adjust cis

Observations and moafications

• Finetuning for internal consistency and domain applicability
• Significant changes for gross errors

Figure 1.4. Validation Process

1.8.3 A Process of Validation

The validation process used in Organizational Consultant relies on
information obtained from cases, consultation with executives, dialogue
with experts, and executive M.B.A. student assignments. The order of
testing was similar to that previously described but involved all three
sources in no special order.
In the process of validation, we make small and large changes in
production rules, in the ease of use, and in the domain of applicability of
Organizational Consultant. Through application, issues are raised, but
resolution depends on theory and its appropriate incorporation into
Organizational Consultant. It also forced us to reconsider the theory, thus
necessitating further theory development. The validation process included
three different types of tests: test cases, executives, and executive MBA
students. More information about the actual validation can be found in
1.9 Summary 37

Baligh, Burton, and Obel (1996). The final result ofthe validation process
was both the theoretical development presented in Part 1 of this book and
the actual Organizational Consultant expert system.

1.9 Summary
An information processing view of organization provides the basis for
developing a normative approach for organizational design and the creation
of a knowledge-base expert system to help us both learn about organiza-
tional design and make recommendations about what the organizational
design should be. For the normative approach, we have utilized the positive
science organization contingency theory as a source of knowledge within the
design framework. The model is pictured in Figure 1.2 with the contingent
factors on the left and the design elements on the right.
Four concepts of fit were introduced to construct the model and design
organizations as shown in Figure 1.3. The model has been validated and
implemented as the expert system Organizational Consultant, which has
proven to be a useful tool in teaching and consulting.

Notes
1. Organizational Consultant runs on any PC supporting Windows 95 and
is a CD companion to this book.

2. Starling Hunter suggested the 7 C's.

3. The certainty factor principle will be discussed in a subsequent section.

4. Note CF =CF1 + (100 - CF1)xCF2/100.

5. More information about Organizational Consultant and the way it
operates is included in Part 2 of this book and the accompanying CD.

6. This section is based on Baligh, Burton, and Obel (1996).

38 Chapter 1. Introduction

References
Aldrich, Howard E. 1972. "Technology and Organization Structure: A Reexami-
nation of the Findings of the Aston Group." Administrative Science Quarterly,
17(1),26-43.
Arrow, Kenneth J. 1994. The Limits of Organization. New York: W.W. Norton.
Baligh, Helmy H., Richard M. Burton, and BliJrge Obel. 1987. "Design of Organiza-
tional Structures: An Expert System Method." In Jean-Louis Roos (Ed.),
Economics and Artificial Intelligence. Oxford: Pergamon Press, 177-182.
Baligh, Helmy H., Richard M. Burton, and BliJrge Obel. 1996. "Organizational
Consultant: Creating a Useable Theory for Organizational Design." Management
Science, 42(12), 1648-1662.
Berry, D.C., and AE. Hart. 1993. "Evaluating Expert Systems." Expert Systems,
7(4), 199-207.
Blau, Peter M., and Richard A Schoenherr. 1971. The Structure of Organizations.
New York: Basic Books.
Blanning, R.W. 1984. "Knowledge Acquisition and System Validation in Expert
Systems for Management." Human Systems Management, 4, 280-285.
Burton, Richard M., and BliJrge Obel. 1984. Designing Efficient Organizations:
Modelling and Experimentation. Amsterdam: North Holland.
Burton, Richard M., and BliJrge Obel. 1994. "Organization Studies: An Information
Processing Perspective." In Saul Gass and Carl Harris (Eds.), The Encyclopedia
of Operations Research and Management Science. Boston: Kluwer.
Campbell, D.T., and J.C. Stanley. 1963. Experimental and Quasi-Experimental
Designs for Research. Chicago: Rand-McNally.
Carley, K.M. and M.J. Prietula (eds). 1994. Computational Organization Theory,
Lawrence Erlbaum Associates.
Child, John. 1972. "Organization, Structure, Environment and Performance: The
Role of Strategic Choice." Academy of Management Review, 561-575.
Cook, T. D., and D. T. Campbell. 1976. "The Design and Conduct of Quasi-
Experiments and True Experiments in Field Settings." In M. D. Dunnette (Ed.),
Handbook ofIndustrial and Organizational Psychology. Chicago: Rand-McNally.
Daft, Richard L. 1992. Organization Theory and Design (4th ed.). St. Paul, MN:
West (2nd ed., 1986, 3rd ed., 1989).
Davis, R. 1984. "Amplifying Expertise with Expert Systems." In P .H. Winston K. A
Prendergast (Eds.), The A. 1. Business. Cambridge, MA: MIT Press.
Donaldson, Lex. 1987. "Strategy and Structural Adjustment to Regain Fit and
Performance: In Defense of Contingency Theory." Journal of Management
Studies, 24(1), 1-24.
Doty, D. Harold, W. H. Glick, and G. P. Huber. 1993. "Fit, Equifinality, and
Organizational Effectiveness: A Test ofTwo Configurational Theories." Academy
of Management Journal, 38(6), 1198-1250.
Drazin, R., and A H. Van de Ven. 1985. "Alternative Forms of Fit in Contingency
Theory." Administrative Science Quarterly, 30, 514-539.
Duncan, Robert B. 1972. "Characteristics of Organizational Environments and
Perceived Environmental Uncertainty." Administrative Science Quarterly, 17(3),
313-327.
References 39

Duncan, Robert B. 1979. "What Is the Right Organization Structure?" Organi-

zational Dynamics, Winter, 59-79.
Feldman, D.C., and H.J. Arnold. 1983. Managing Individual and Group Behavior
in Organizations. New York: McGraw-Hill.
Galbraith, Jay R 1973. Designing Complex Organizations. Reading, MA: Addison-
Wesley.
Galbraith, Jay R 1974. "Organization Design: An Information Processing View."
Interfaces, 4(3), 28-36.
Hall, Richard M. 1991. Organizations, Structures, Process, Outcomes. Englewood
Cliffs, NJ: Prentice-Hall.
Hall, Richard M., J. Eugene Hass, and Norman Johnson. 1967. "Organization Size,
Complexity, and Formalization." American Sociological Review, 32(6), 903-912.
Harmon, P., and D. King. 1983. Expert Systems: Artificial Intelligence in Business.
New York: Wiley.
Harmon, P., and D. King. 1985. Expert Systems: Artificial Intelligence in Business.
New York: Wiley.
Hayes-Roth, F., D. Waterman, and D. Lenat. 1983. Building Expert Systems.
Reading, MA: Addison-Wesley.
Jin Y. and RE. Levitt. 1996."The Virtual Design Team: A Computational Model of
Project Organizations." Computational and Mathematical Organization Theory,
2 (3),171-196
Khandwalla, P. 1973. "Viable and Effective Organizational Designs of Firms."
Academy of Management Journal, 16, 481-495.
March, James G. And Herbert A. Simon. 1958. Organizations. New York: John
Wiley & Sons, Inc. 262 p.
McGrath, J. E., J. Martin, and R .A. Kulka. 1982. Judgement Calls in Research.
Beverly Hills, CA: Sage.
McGraw, KL., and K Harbison-Briggs. 1989. Knowledge Acquisition: Principles and
Guidelines. Englewood Cliffs, NJ: Prentice-Hall.
Miles, Raymond E., and Charles C. Snow. 1978. Organizational Strategy, Structure,
and Process. New York: McGraw-Hill.
Miller, Danny. 1992. "Environmental Fit Versus Internal Fit." Organizational
Science, 3(2), 159-178.
M4. 1991. User's Guide. Palo Alto, CA: Cimflex Technowledge.
Nadler, David, and Michael L. Tushman. 1988. Strategic Organization Design:
Concepts, Tools and Processes. Glenview, IL: Scott-Foresman.
O'Leary, D. E. 1988. "Methods of Validating Expert Systems." Interfaces. 18, 72-79.
O'Leary, D. E. 1996. "Verification of Uncertain Knowledge-based Systems."
Management Science, 42(12),1663-1675.
Pennings, J.M. 1987. "Structural Contingency Theory: A Multivariate Test."
Organizational Studies. 8(3) 223-240.
Perrow, Charles. 1967. "A Framework for Comparative Analysis of Organizations."
American Sociological Review, 32, 194-208.
Preece, A.D. 1990. "Towards a Methodology for Evaluating Expert Systems." Expert
Systems, 7(4), 215-223.
Pugh, D.s., D.J. Hickson, C.R Hinings, and C. Turner. 1969. "The Context of
Organization Structures." Administrative Science Quarterly, 14,91-114.
40 Chapter 1. Introduction

Quinn, James Brian, Henry Mintzberg, and Robert M. James. 1988. The Strategy
Process: Concepts, Contexts, and Cases. Englewood Cliffs, N.J: Prentice-Hall.
Robbins, Stephen P. 1990. Organization Theory: Structure, Design, and Applica-
tions. Englewood Cliffs, NJ: Prentice-Hall. (1st ed., 1983, 2nd ed., 1987).
Schoonhoven, Claudia Bird. 1981. "Problems with Contingency Theory: Testing
Assumptions Hidden within the Language of Contingency Theory." Administra-
tive Science Quarterly, 26,349-377.
Scott, W. Richard. 1992. Organizations, Rational, Natural and Open Systems. (3rd
ed.). Englewood Cliffs, NJ: Prentice-Hall.
Simon, Herbert A. 1981. The Science of the Artificial. Cambridge, MA: MIT Press.
Thomsen, Jan. 1998. The Virtual team Alliance(VTAJ: Modeling the Effects of Goal
Incongruence in Semi-routine, Fast-paced Project Organizations, Ph.D.
Dissertation, Stanford University.
CHAPrER2

Describing the Organization for De-

sign Purposes

2.1 Introduction
When you examine an organization that you do not know, you can look
at it in a number of ways. You may gather information about what the
organization actually does-the goods and services it provides. You may
be told who the boss is and who makes the decisions. You may be shown
an organizational chart. Many companies also have explicitly stated
objectives that drive their strategy. This information tells you about the
design of the organization, the way the organization is put together, who
does what, and who talks with whom.
We want to understand how the organization uses information to
make decisions, coordinate its activities, and implement what it wants
to do. The organizational configuration, or structural configuration,
provides a beginning description to understand how the organization
uses information. The information view is a tool for both describing and
designing the organization
Organizational design begins with a description of the organization.
This description cannot be arbitrary, but must facilitate the design pro-
cess itself. Concepts should be clear, precise, observable and measur-
able. Management must be able to implement the recommended organi-
zational design. We employ an information processing framework as our
fundamental view; an organization is viewed as the way it deals with
information. An organizational design is the specification of configura-
tion, complexity, formalization, centralization, and coordination and
control which helps determine how the organization will process its
information. We begin with the Scandinavian Airline System (SAS), a
case we will examine throughout the book.

New Forms and Organizational Description

Recently, there has emerged great interest in new organizational forms. These new forms
are intriguing as they nominally deviate from the well-known functional, divisional, matrix,
and bureaucratic configurations. Yet, we want to examine what is new, and what is
modified from what we already know. We live in a fast paced world with new information
technologies and quick response customer environments. The pace of corporate life has
increased and the pace of change has quickened. But the pace of change does not
necessarily mean that existing fundamental concepts are no longer applicable. Information
42 Chapter 2. Describing the Organization for Design Purposes

processing is now faster and less expensive; does this change the underlying basis for
organization. We argue that new forms do require new research, but we also want to argue
that what we have learned from years of research and experience remains relevant.
A fundamental question here is whether the new forms require a new vocabulary for
describing the organization, or whether the traditional organizational descriptors included
in this chapter still apply. That is, can we describe the new forms in terms of configuration,
organizational complexity, formalization, coordination mechanisms, incentives, etc. or do
we need a new set of descriptors. We suggest that the traditional terms still apply and
further we can utilize the extant organizational research in our understanding of new
forms.
New forms may be new values or a new combination of values for the organizational
descriptors herein. An older form might be a bureaucratic configuration accompanied by
high formalization, high complexity and procedural based incentives. A new flexible form
may have an ad hoc configuration with low formalization, low complexity and result based
incentives. The flexible form does not then require new conceptual descriptors; the flexible
form takes on new values for the well known organizational descriptors. That is, a new
vocabulary is not the issue, our research and understanding of organizations should
address the new combination of organizational values that we have not researched and
understood well. Indeed, we do want to understand the flexible organization, what it is and
the situations where it is appropriate to use.
The information processing view of organization then suggests that the new forms of
organization involve the same management issues as older forms of organization. There
is more similarity than difference between new and old forms. Information remains the
means by which the organization is coordinated and activities fit together. The information
processing perspective can then be used to analyze new organizational forms. Generally,
a new organizational design should balance the demand for information processing and
the capacity for information processing. Thus, the view and propositions presented in this
book can be used to evaluate new as well as old organizational forms, using the traditional
organizational descriptors.
In Chapter 6, the issue of flexibility is discussed in the capsule on hypercompetition. In
Chapter 8 strategy and new organizational form are examined. We return to new forms
and explore some conditions when a new form would be appropriate in Chapter 9.

2.1.1 The Scandinavian Airline System (SAS)l

The Scandinavian Airline System (SAS) is used as an illustration in
each chapter in this book beginning with an introduction to the company
and its history. In the final section of this chapter we show the various
organizational structures that SAS has used in its ongoing challenge
with changes in competition, high operating costs, and strategic focus.
The three Scandinavian countries-Norway, Sweden and Denmark-
are culturally close and have a strong tradition of cooperation and trade
SAS was the result of the joint Scandinavian cooperation on air
traffic. The time for its creation in 1951 was well chosen. The market
situation invited cooperation. The airline business was at that time one
of the most regulated businesses. The 1944 Chicago convention passed
international rules for civil aviation, stating that each country had
complete sovereignty over its own air territory and could make sole
2.1 Introduction 43

decisions about rights to fly to, from, and over its territory. This meant
that each country's government was responsible for granting concessions
and issuing traffic permits. With the SAS agreement, the three Scandi-
navian airlines gained a stronger negotiating position in bilateral
agreements. The relatively poor negotiating position of the national
companies resulted from Scandinavia's sparsely populated area and the
limited interest that foreign airline companies had in this market.
Scandinavian countries as a whole did not have much to offer when they
applied for traffic rights abroad. Copenhagen's position on the continent
was the only trump card SAS had (Buraas, 1972, p. 160), but it was
enough to give SAS access to the American market and later to the
Japanese market.
SAS's air routes were, therefore, based on concessions they had been
granted by local governments and on agreements that had been reached
with foreign governments; this meant limited possibilities of expanding
flying to new markets. SAS's freedom of choice was further limited by
international cooperation among the airline companies. The Interna-
tional Air Transport Association (lATA), which was the air companies'
own organization, was established in 1946 as a supplement to the
bilateral network of agreements. lATA's main task was to coordinate
ticket prices on a global scale. Through membership in lATA, SAS could
consider the prices fixed during the period between two lATA meetings
(Buraas, 1972, p. 170). With members all over the world, lATA had so
much power that even companies outside lATA followed the prices fixed
by lATA. To establish the price, each company calculated its lowest
acceptable price. Then the proposal with the largest support became the
price, but each country had the right of veto (Buraas, 1972, p. 171).
When agreement was reached, the price proposals were presented to the
governments of each country for confirmation. If agreement could not be
reached, it meant that one airline could offer tickets cheaper, which
meant that a government-supported price war could occur. No govern-
ment either could or would accept this outcome. The guarantee offixed
prices was an advantage as well as a disadvantage for the companies.
The advantage was that each company could calculate on a fixed price
but in return could not use price competition to gain market share.
Those cooperating in lATA focused on keeping prices high. Airline
companies needed a concession to fly, and with a market in constant
growth, there was room for everybody. This meant that with fixed
prices, management simply needed to keep planes in the air as much as
possible to maximize the total profit. SAS thrived on technical compe-
tence: pilots, technicians, and system managers were its heroes
(Hofstede, 1991, p. 196). Unavoidably, these groups influenced decisions
about buying planes. Consequently, during the 1950s and 1960s SAS
44 Chapter 2. Describing the Organization for Design Purposes

offered modern equipment, which was frequently replaced. During these

years, management became bureaucratic and was dominated by the
technical competence of the employees, and with a growing market and
strong governmental protection, a bureaucratic management was an
appropriate organizational structure.
This was SAS from its establishment in February 1951 until the mid
1970s. Then the oil crisis started. When the Arab oil producers suddenly
turned off the taps, they struck a blow at the entire airline business. The
prognosis was that known oil reserves would last only until the end of
the century. SAS, therefore, had to consider whether it could survive. In
the short run, the world economy was so severely affected that the
market for air traffic changed fundamentally. The turbulence continued
with deregulation, first in the United States and then later in Europe.
Prior to 1975, SAS operated in a stable environment. Since 1951, the
market grew constantly and predictably. Participation in lATA meant
that the national companies had not been forced to compete. But the oil
crisis changed everything. Oil prices that had followed normal price
increases now rose explosively. SAS operating costs increased sharply,
but revenues did not. SAS found itself in a completely new competitive
situation after 1975 (Obel, 1986).
With a radically changed environment, a substantially worse econo-
my, and a stagnant market, SAS had to start thinking along new lines.
But it was difficult. During the good years SAS had built a large,
bureaucratic organization that could not easily be turned around. Cost
consciousness was not enough, and the board of directors recognized the
need for a results oriented company. To realize these goals, study groups
were established to create a profit without firing employees. But
eventually, it was realized that the SAS management was too conserva-
tive to support new ways ofthinking.
It was obvious that SAS needed restructuring to be able to function
in the new environment. During fall 1980, the first steps were taken
toward a larger organizational change. The SAS board of directors
decided to divide SAS into three divisions-that is, corporate manage-
ment, the airline company SAS, and SAS's subsidiaries. When Jan
Carlzon left the Swedish Linjeflyg and became president of the airline
company SAS, a new era began. He restructured SAS's bureaucratic
organization to fit the new economic environment and created a
completely new objective for SAS. Carlzon changed SAS from being
product oriented into being market oriented. Decentralization became
a key word in the daily life of employees. In the new market environ-
ment each airline company had to fight harder to win the customer's
favor. Carlzon's goal was for customers to choose SAS because SAS was
the best alternative.
2.2 Organizational Configurations 45

The new SAS was different from the old SAS in many respects. Many
new issues were now relevant for the success of the company, and
changes were more frequent, which required SAS to deal with more
information. Handling the information processing requirements success-
fully was essential. Later in this chapter and throughout the book, we
return to SAS and discuss the new SAS in more detail.

2.2 Organizational Configurations

In describing an organization, many people begin with the organiza-
tional configuration. The configuration is represented most frequently
as an organizational chart. The configuration specifies the general
principle for dividing work, breaking tasks into subtasks and coordinat-
ing activities. There are a large number of organizational configurations,
and new ones emerge from time to time. The following basic configura-
tions are the most common:

• Simple: A flat hierarchy and a singular head for control and decision
making.
• Functional: Unit grouping by functional specialization (production,
marketing, finance, human resources, and so on).
• Divisional: Self-contained unit groupings into somewhat autonomous
units coordinated by a headquarters unit (product, customer, or
geographical grouping, including multinational).
• Matrix: A structure that assigns specialists from functional depart-
ments to work on one or more interdisciplinary teams that are led by
project leaders. Permanent product teams are also possible. A dual
hierarchy (such as function and projects) manages the same activities
and individuals at the same time. A three-dimensional matrix of
product, function, and country is common in multinational organiza-
tions.
• Machine bureaucracy: Highly routine operating tasks, very formal-
ized rules and regulations, tasks grouped into functional depart-
ments, centralized authority, decision making follows the chain of
command, and an elaborate administrative structure with sharp
distinction between line and staff.
• Professional bureaucracy: Highly skilled professionals, high complexi-
ty, decentralization, and internal professional standards.
• Ad hoc: High horizontal differentiation, low vertical differentiation,
low formalization, decentralization, and great flexibility and
responsiveness.
46 Chapter 2. Describing the Organization for Design Purposes

In the choice of a configuration, or more generally a structure, we apply

our criteria of efficiency, effectiveness, and viability. To meet these
goals, the organizational structure must fit internally and also fit or
match the external environment.
As discussed above, an efficient organization is one that uses its
resources well to produce goods or services. Efficiency is increased with
fewer resources utilized or more products created. Effectiveness is
measured by how well the organization produces goals or services that
are valued-that is, the organization's outputs must be desired by
someone, either a customer or the public at large. Efficiency is an
internal measure, and effectiveness is an external measure. Efficiency
is doing the thing right; effectiveness is doing the right thing. In the
total quality management literature, quality relates to choosing the
right strategy to fulfil the goals and the right organization to implement
the strategy. The fundamental organizational design problem is to
structure the organization to be efficient, effective, and viable.
Organizational fit is a measure of how well the structure facilitates
coordinated activities-both internally and externally. Misfits are easy
to illustrate. If the one production unit makes 100 automobile bodies
and another produces fifty engines, there is a misfit of activities; and
there is a misfit of organizational structure if the structure does not
facilitate the needed coordination. Similarly, there is an environmental
misfit when 100 automobiles are produced when only fifty can be sold.
Fit is our fundamental design criterion. The organization must fit
together to obtain coordinated activities; the organization must fit its
environment and strategy so that its activities make it viable in the
larger world. An organization that fits internally will be efficient; one
that fits externally will be effective: In Chapter 1 the various fit criteria
were presented.
The specification of a configuration does not prescribe in detail how
the organization should be designed. There are many ways to create a
divisional or a functional structure. It has to be decided how many
functions or how many divisions and on which basis the functions and
divisions should be based. These issues are discussed in later capsules.
Let us review each configuration more closely.

2.2.1 Simple Configuration

The simple configuration consists of a top manager and individuals.
There may be little functional specialization and no well-defined
departmental structure with departmental heads. Decision making,
coordination, and control are usually done by the top manager. In Figure
2.2 Organizational Configurations 47

2.1, a simple configuration is pictured as an organization chart.

Individual names are used to highlight the relations among these
individuals, but the organization usually lacks task specificity, so
functional activities are omitted. Normally, George, Jane, and Jens do
what Peter tells them to do. For this configuration the leader is
dominant, and the number of vertical levels seldom exceeds two.
This configuration is often chosen by small owner-run companies in
their early stages. This is the time when the founder wants to be
involved in all the activities of the organization. In Chapter 9, we
discuss in more detail when a particular configuration should be chosen.
The simple configuration has strengths and weaknesses. Decision
making and control rest with the top manager. Peter can personally
coordinate all activities and assume that these activities meet his
purpose for the organization. The activities or tasks are assigned by
Peter directly to George, Jane, and Jens. Each is expected to complete
those activities that Peter assigns to each. Specific task assignments
may evolve, but Peter can change them at will.

Peter

Top Manager

I I
George Jane Jens

Figure 2.1. A Simple Configuration

The weakness of the simple configuration is its heavy reliance on the top
manager. He or she determines what to do and how to do it. Ifhe or she
decides well, then the firm succeeds; if not, it fails. There is little
redundancy or fallback. The simple configuration is also limited by the
information processing capacity of the top manager. As the firm grows
in size or complexity of the tasks, the top manager can become over-
loaded and simply unable to cope with the information demands. This
leads to a need for task specificity about who is doing what on a regular
basis, so that the top manager needs to process less information. But
48 Chapter 2. Describing the Organization for Design Purposes

that information must be specified in a form that obtains the required

coordination among the specialized tasks or activities.

2.2.2 Functional Configuration

The functional configuration has more levels and more horizontal special-
ization than a simple configuration. There is a well-defined departmental
structure. The departments are created based on the functional specializa-
tion in the organization. A distinction between line and staff departments
is often made. For example, in a manufacturing firm the line departments
could be based on purchasing, production, and sales while the staff units
could be financial, accounting, and human resources departments. The
functional organization is the most common configuration. Organizations
with a simple configuration that grow often change to a functional configu-
ration. See Figure 2.2 for a prototypical functional organization.

HEADQUARTERS
Top Management

I
L--

I I I I
DEPARTMENT DEPARTMENT DEPARTMENT DEPARTMENT
Operations Market ing Finance Human Resources

Figure 2.2. A Prototypical Functional Configuration

While decision making in the simple configuration is done by the top

manager and therefore the coordination and control issues are simple;
coordination and control are much more complicated in a functional
configuration. The horizontal specialization as the basis for the departmen-
talization creates departments that are very different. Each manager of a
particular department may see his department as the primary reason for
the organization's competitive advantage, which can create conflict.
However, horizontal specialization can be the basis for a very efficient use
of resources provided their use is coordinated. The dominant information
flows tend to follow the hierarchy.
2.2 Organizational Configurations 49

The need for coordination requires top management to be involved not

only in strategical issues but to a great extent also in the tactical and
operational issues.

2.2.2.1 A Functional Organization

Mack Trucks, Inc. has a classic functional configuration. Its top manage-
ment and several functions are shown in Figure 2.3.
Mack Trucks is an old company that builds top-of-the-line heavy-duty
trucks in the United States. Mack has a small market share, which has not
changed significantly for many years. It has made few technological
innovations. Over time, truck design does change, but it evolves slowly.
Mack trucks are more expensive than most competitive trucks, and their
quality and reliability are high. Mack Trucks maintains a niche that it has
developed over many years.

Mack Trucks, Inc.

Board of Directors

Chainnan of the Board

President and
Chief Executive OffICer

I I I I
ExeartIYe
ExeartIYe
VIce President executive executIVe executIVe
VIce Presklent VIce President VIce President VIce President
Product &
Administration Maflcellng ManuIactu~ng Rnanoe
EngiMering

Figure 2.3. Mack Trucks, Inc.: The Functional Configuration

Source: Conference Board Chart Collection (1991).

The functional configuration remains the most prevalent in many

industries. It is found in manufacturing, brewing, steel, pharmaceutical,
telecommunications, transportation, public service, and hospitals, to name
a few. Table 2.1 contains the name of several organizations, together with
their functional departments. The functional department names vary
among these organizations.
50 Chapter 2. Describing the Organization for Design Purposes

Table 2.1. Functional Configurations

Company Functions

Chrysler Corporation Sales and Marketing, Manufacturing, Procurement

-(1998) and Supply, Product Strategy, Corporate Staff

Coors Brewing Marketing, Plant Operations, Administration and Fi-

Company (1991) nance, Corporate Affairs, Engineering and Construc-
tion

Bethlehem Steel Operations, Administration, Finance, Law, Planning.

Corporation (1997)

Glaxo Wellcome Technical, General Counsel, Corporate Affairs,

(1997) Marketing and Sales, Finance, Production and Engin-
eering, Personnel, Operations.

Phillips Petroleum Corporate Affairs, Planning and Technology, Produc-

Company (1997) tion, Finance, General Counsel

Cincinnati Bell Tele- Telephone Operations, Accounting, Administration,

phone Company Information Technology, Customer Services, Plan-
(1996) ning.

Federal Express Administration, Personnel, Worldwide Operations,

Corporation (1996) Marketing, Legal, Finance, Information, Quality.

St. Petersburg Accounting, Operations, Personnel, Computer Ser-

Times (1991) vices, Marketing

Lego Systems Sales, Marketing, Finance, Manufacturing, Logistics,

(1991) Quality

Consolidated Edison Public Affairs, Administration, Central Services,

of New York (1996) Operations (Gas, Division, Central), Finance, Law

Beth Israel Hospital Finance, Human Resources, Facilities Planning and

(1991) Engineering, Clinical Support Services

Source: Conference Board Chart Collection.

The functional configuration can be represented generally as in Figure 2.4.

The fundamental nature of the functional configuration is the overall task
specification into subtasks-all of which must be completed to accomplish
2.2 Organizational Configurations 51

the organizational purpose. For example, a manufacturing firm must obtain

supplies and sell its products as well as manufacture them. There is a flow
of materials from supply to manufacturing to sales. Similarly, an engineer-
ing company researches product possibilities, engineers and designs them,
and finally, manufactures and tests them. It, too, must sell the products but
may not have a sales department.

Functions: How many and what are they called

The traditional functional organization has a number of departments which have descriptive
names: procurement, manufacturing, sales, finance and personnel. See Figure 2.1. In terms
of product flow, the raw materials are procured, then manufactured into a product which is sold
to customers. Finance and personnel took care of the money and the people. The organizing
principle is that of function where efficiencies are realized through specialization of task. Each
specialized function did its own part of the total and passed it on to the next function, where
finance controlled the resource issues and personnel dealt with the employees.
In most functional organizations, there are five, plus or minus one or two, functional
departments. Mack Trucks, Inc. has five departments: administration, engineering, marketing,
manufacturing and finance as shown in Figure 2.3. Here, administration includes personnel.
From organization to organization, the departmental names will vary, but manufacturing or
operations will be present as well as marketing or sales and some form of finance or
accounting. These names are typical for product manufacturers. In Table 2.1, corporations in
industries ranging from brewing to newspaper publishing have very similar functional
department names. For engineering firms, typical functional departments will be: research and
development, engineering and operations. Sales, personnel and finance functions will be
included at the firm headquarters. However, for service firms, the names and functions can vary
widely. An advertising agency may have functions such as layout, creation, graphics, client
relations, etc. The main principle is that tasks and individuals are groups by specialized function
where there are efficiencies to be realized, and further there is little substitution across
functions, even though there is a workflow which requires coordination.
How many functions should there be? Five, plus or minus one or two is typical. There is no
fixed number. Although, most organizations do have a common set of functions which must be
performed as suggested above. Of course, accounting can either stand as a functional
department or it can be included in finance or administration; accounting can also be
outsourced. There is considerable variation. Similarly, most functions can be renamed and
subsumed in others. The result is that there is no set number of functions.
The balance is that the efficiency of specialization should be realized, which suggests there
should be at least a few functions. However, the more functions there are, the greater the
difficulty of coordination among the functions. For most functional organizations, five, plus or
minus, seems to work well.

The coordination of these functions minimally requires that the products

and their quantities should be the same for each of the functions. The firm
should not sell fifty trucks, manufacture 100, and buy supplies for 150.
Similarly, it should not make 100 if the market demands 150 at the going
price. This coordination requirement is simply stated, but it is very
complicated in reality. If there is one standard product, the above coordina-
tion description is sufficient. If there are hundreds or thousands of
customized products, coordination is very involved, requiring enormous
amounts of information. Those information flows usually are hierarchical,
52 Chapter 2. Describing the Organization for Design Purposes

as shown in Figure 2.4, and the decision making for coordination is done by
the headquarters units or top management.

HEADOUARTERS
Top Management

+--+ +-+
1. Supply: Manufacturing Sales
Materials.
Labor
Equipment
or

2. Research Engineering and Design Manufacturing and Test

Product flow
Information flow
+-+
---.

Figure 2.4. The Functional Configuration

Galbraith's information processing strategies emphasizes standard

operating procedures or coordination by rules or programs. Standardized
products and standardized production processes reduce the need for
information processing. As standardization decreases and customization
increases, there is a need for greater information processing. Increased
hierarchical information processing is one response. However, top manage-
ment has a limited capacity, and other mechanisms are required. Lateral
communications can be effective means to obtain the required coordination.
There is an increasing emphasis on lateral communications, particularly
modem electronic mail systems.
The rationale for the functional configuration lies in the efficiencies of
specialization; it is more efficient to have one group of individuals
manufacture a product and another group sell it than for everyone to do
both functions. Different skills and interests are required, and not everyone
can be excellent at both. This is the rationale for specialization-or the
partitioning of the overall task into functional subtasks. The coordination
2.2 Organizational Configurations 53

issue becomes the essential short-time management problem-keeping the

product flow in balance. In the longer run, management must match the
functional capabilities with market demand for the product such that the
firm is profitable and a return can be realized for the owners. This is a
minimal requirement for a viable firm.
Coordination approaches vary across firms. In the above discussion, the
headquarters, or top manager, balances the product, or work flows. This is
central planning at the top. However, anyone function could also serve in
this role. For example, the sales function could forecast sales and then tell
manufacturing how many of the product to make. Or alternatively,
manufacturing could set the production quantities and tell sales to sell
them at the best price possible. These leader-follower coordination
approaches are also used in functional configured firms.
Functional configurations have limitations. As the number of products
and their customization grows, the coordination problem can become over-
whelming as information processing needs escalate. This is a sufficient
reason to consider other configurations. Functional configurations also tend
to lack flexibility and innovation, for several reasons. First, the organiza-
tion may be information overloaded. This may be the case if changes in the
environment make constant coordination across the functional specializa-
tion necessary. Second, no one may be directly responsible for innovation-
particularly of new products. Both sales and manufacturing focus on today's
products. Manufacturing may focus on cost reduction, but usually not new
products. Third, the functional incentives do not lead to innovation. Sales
tries to maximize sales, and manufacturing tries to minimize costs. Both
incentives are reasonable and easy to measure but may not lead to longer-
run profitability and viability. The functional specialist's contribution to
longer-term profitability is difficult to measure and hence to reward. Addi-
tionally, product innovation itself requires coordination of the functional
specialties. A new product has to please both manufacturing and sales.
In brief, the functionally configured firm does well in a stable environ-
ment that requires little change with a known and stable technology, that
offers efficiencies in specialization, and that allows coordination to be
obtained with reasonable information processing requirements.

2.2.3 Divisional Configuration

The divisional configuration is characterized by organizational subunits
based on a grouping of products, markets, or customers. The units are
relatively autonomous contrary to the units in the functional configuration.
It is the aim in the design ofthe division to minimize the interdependency
of the units. Coordination of these subunits-called divisions-is very
54 Chapter 2. Describing the Organization for Design Purposes

different than coordination in the functional configuration. In a pure

divisional form the top management is not involved in operational and
tactical issues (Williamson, 1975). If top management engages in issues
that are not of a strategic nature, the configuration is called a corrupt
divisional configuration. The divisions themselves can have any configura-
tion, but very commonly they have a functional configuration as shown in
Figure 2.5.

HEADQUARTERS
Top Management

DIVISION 1 DIVISION 2 DIVISION 3

STAFF

Figure 2.5. The Divisional Configuration

There are a number of different bases for the divisional configurations.

Usually the divisions are based on products or product groups. The markets
by product, geography, or customer can form the basis for a divisional
grouping. The grouping also can be a mixture. Usually a divisional
configuration goes best with a large organization and where the relation-
ship between the divisions is not very involved, such as when the technol-
ogy is divisible. Staff departments may exist both at the top and at the
divisional level.
The divisions may be configured differently, as simple and functional, for
example. They also can be different businesses, such as automobiles and
hamburgers. However, they all may also be alike. A retail chain with
exactly the same type of store on each location is one such example. This
2.2 Organizational Configurations 55

particular version of the divisional configuration is called the carbon-copy

divisional configuration, (Mintzberg, 1979, p. 395). The carbon-copy
structure is often used in retail and service businesses when economies of
scale are not important, either due to geographical dispersion or type of
business. Retail chains and banks have identical shops and branches at
different locations. But carbon copies also may exist at the same location.
The Danish advertising agency Bergsoe reorganized from a functional
structure into a carbon-copy organization. As the number of projects
increased, the horizontal coordination across departments in a regular
functional organization became too difficult to handle. Bergsoe tried to solve
the problem by installing a highly sophisticated computerized project
scheduling system, which it did not work. Projects did not meet deadlines,
and people started to bypass the scheduling system. The situation grew
worse. The solution was to break the company into small self-contained
groups that had all the specialists needed to handle a project (recall one of
Galbraith's approaches). Eight such groups were created, and all were
located in the same office. To eliminate competition among the groups, a
scheme for allocating projects and sharing profit was developed. Most ofthe
eight groups were structured the same way. However, some were run as an
adhocracy while others were simple or functional organizations. But they
all did the same thing: no product specialization was allowed. For a given
customer, each division dealt with all needs ranging from video production
to advertising campaigns in newspapers and magazines. The reorganization
changed both the information-processing demand and capacity.

2.2.3.1 A Divisional Organization

General Electric Company has a classic divisional configuration where each
division is in a different business. The top management and the several
divisions are shown in Figure 2.6. The several divisions report to the
corporate executive office.
The divisions are product or service based. GE is a very large corpora-
tion that develops, manufactures, and sells many different products and
some services, such as broadcasting and finance. Over the last decade, GE
has bought and sold many divisions. This configuration facilitates such
actions.
The divisional configuration is the second most prevalent configuration.
It has become more common as firms have become more product and
customer focused.
56 Chapter 2. Describing the Organization for Design Purposes

Corporate Executive Office

Top Management

I
National Broadcasting General Electric
Company , Inc.
r- GE Aerospace - Capital Services , Inc.

GE International I- GE Aircraft Engines - GE lighting

I - GE mecical Systems I- GE Plastics

GE Transportation GE Industrial & Power

- Systems
I-
Systems

- Corporate R&D - GE Appliances

'-- GE Information r- GE Motors

'-- GE Supply

Figure 2.6. The Divisional Configuration of General Electric

Source: Conference Board Chart Collection (1997)

The divisional configuration also is to be found across a broad range of

industries: aircraft, heavy manufacturing, electronics, pharmaceutical and
chemicals, construction, and consulting. Table 2.2 contains the name of
several organizations, together with their divisions.
The divisional configuration is represented generally in Figure 2.7. The
fundamental nature ofthe divisional configuration is the partitioning ofthe
overall task into relatively autonomous units. Each divisional unit is
self-contained in its operations. As discussed previously, the General
Electric divisions have very few, if any, relations between the divisions, and
each division is in a separate business. There are no flows of goods or
information required among the divisions: each buys and sells into separate
markets. However, all divisions are tied to the headquarters or top
2.2 Organizational Configurations 57

management through information, financial issues, and policy matters.

Information processing requirements grow as top management becomes
involved in the operating decisions of the divisions.

Table 2.2. Divisional Configurations

Company Divisions

Caterpillar (1997) Financial Products, Component Products, Technical

Services, Engine, Human Serv-ices, Caterpillar Over-
seas, Caterpillar Brasil, Solar Turbine, Construction
and Mining Products, Building Construction Products,
North America Commercial Division, Asia-Pacific-
Latin America, Shin Caterpillar Mitsubishi Ltd.

Hewlett-Packard Finance and Administration, Measurement Systems,

(1998) Computer Products, Computer Systems

Ciba-Geigy Group Dyestuffs and Chemicals, Pharmaceutical, Agricul-

(1991) tural, Plastics, Additives, Pigments

Henkel KGaA Chemical Products, Hygiene Cleaners, Adhesives,

(1998) Cosmetics/Toiletries, DetergentslHousehold Cleaners

Bell South Corpora- Advertising and Publishing, Communications Systems,

tion (1996) Mobile Systems, Mobile Data, Long Distance and
Video

Source: Conference Board Chart Collection (1991)

There are no coordination issues among the divisions. The jet engine
division and the lighting division of GE do not need to coordinate produc-
tion levels or sales plans. Consequently, as the number of decisions grows,
the additional coordination requirements involve only relations between the
division and top management.
Information processing requirements between the divisions and top
management can be limited to financial matters and general policy issues,
including strategic planning. And information processing among the
divisions is not required or very minimal. The information processing
demands of the divisional configuration are quite limited, and additional
information processing within a division due to increased products or
customization does not increase the information processing demands at the
divisional-corporate level. Consequently, the divisional configuration can
58 Chapter 2. Describing the Organization for Design Purposes

grow by adding divisions or grow within any division without information

processing difficulties.

HEADQUARTERS
Top Management

r - - -..........-..., f
~
Market
~
Market
~
Market

Product divisions, i.e. group of similar products

Single product divisions
Customer groups, or individual customers

Information flow - - - - -....

Product flow

Figure 2.7. The Divisional Configuration

Divisions: How many and what kind

The variation in the number of divisions is large. There can be either very few or a large
number. The basis for the divisions can be: products, markets, customers or geographical
location. The main principle is to have a close connection with the outside or environment of
the organization. A customer division is organized around the buyer. Even the product
organization is focused on the outside market and who will buy the product. Geographical or
regional divisions are oriented to the local population, buyers and markets.
As shown in Figure 2.6, General Electric has some sixteen divisions. These divisions are
quite varied in their foci from jet engines and financial services to television broadcasting. They
are connected by a common ownership and financial outcome, but their processes, products
or services and customers can be largely unconnected with each other. The coordination
among the divisions is financial and strategic, but usually not operational. The GE headquarters
can "manage" a large number of divisions as the information processing required at the
headquarters is focused on finance and does not involve detail of operations or customers. The
GE divisions are largely product based: lighting, plastics, engines, motors, etc. Some are
service based as: transportation, finance and broadcasting .
Product or service is but one basis for a divisional configuration. Henkel KgaA is product
based, but it is also customer based at the same time as shown in Table 2.2. Chemical
products and cleaners are industrial products aimed at industrial users; cosmetics and
household cleaners are household products aimed at the individual customers. Each requires
2.2 Organizational Configurations 59

a different focus, particularly for marketing and sales. Many service organizations will have a
customer oriented divisional configuration. Banks and other financial institutions will have
corporate divisions and personal divisions, sometimes these divisions will be called "clienf
divisions. The goal is to focus first on the customer and hislher needs and preferences, and
secondly on the service and how it is provided. Bell South as shown in Table 2.2 is an example
of a service company which matches its organization to the service and the type of user.
Geography is a frequent division configuration rationale. Many international corporations,
such as Philips, are organized by region or country. The country provides a local focus on the
customer and the local customer needs as well as legal and other factors which may be
particular to the region or country. Recently, a number of country divisional European
corporations have reorganized by product or customer within the EU region.
The divisional configuration gives a first order extemal focus to the organization. The focus
can be product and its market, the customer and his needs, or the region and its uniqueness.

The rationale or the divisional configuration depends on the efficiency ofthe

coordination processes and rather minimal information processing
demands. Referring to Galbraith's strategic alternatives, the divisional
configuration creates the divisions as relatively self-contained units. Each
can operate independent of the other divisions without creating conflicts or
loss of opportunity.
Additionally, each division can focus on its customers, markets, and
products; specialization is by customer or product. It is an effective
organization, as the division tends to do the right thing through its
responsiveness to customers. However, the divisional configuration has
limitations; most interdivisional efficiencies are overlooked and lost.
Divisions can duplicate costly developments in research, underutilize or
duplicate costly capital investments, and generally, lose opportunities of
process specialization. Further, potential economics of scale can be lost.
However, the financial performance of each division can be assessed using
standard accounting measures of income and profits, and thus a perfor-
mance-based incentive system can be put in place.
The divisional configuration can be quite flexible. It depends on each
division to pursue its market and technological opportunities. The division
will normally focus on variations of its existing customers, markets, and
products. New opportunities beyond the existing divisional activities are top
management's responsibility.
The great advantage of the divisional configuration is the direct focus of
each division on its customers, markets, products, and technology with
measured performance goals. Each division is its own business.

2.2.4 Matrix Configuration

The three basic configurations discussed so far have been hierarchical. The
matrix structure introduces a dual-hierarchy configuration; it incorporates
60 Chapter 2. Describing the Organization for Design Purposes

the essential functional and divisional configuration on an organization

simultaneously.
The matrix configuration can take many variations. One of the more
usual is shown in Figure 2.8. Here the basis for the matrix configuration is
a regular functional form with departments as described in Section 2.4.2.
Together with the functional structure the organization also has a project
orgf.llization. That is, each project in the organization is headed by a project
manager who has responsibility for the project. Normally, he or she has to
request the resources to carry out the project from the various departments.
Each department head is responsible for ensuring that the resources are
used as efficiently as possible.

FUNCTION FUNCTION FUNCTION FUNCTION

1 2 3 4

Project 1
Project 2
Product 1
Product 2
Customer 1
Customer 2

Figure 2.8. A Matrix Configuration

Projects may be temporary, or they may be based on products, programs,

or customers for a longer term. At the limit, a matrix configuration with
only one department is called a project configuration.
Matrix organizations are widely used in corporations and public
organizations. Early use of matrix organization occurred in the production
of airplanes and aerospace projects. The functions were research, develop-
ment, engineering, assembly and testing. The projects were individual
airplanes, rocket engines, guidance systems, and so on. Since then, the
matrix organization (although the name matrix may not be used) has been
quite common. The most common form is to have a functional organization
with a project or product management. In multinational corporations, the
three-dimensional matrix of function, product, and region (country) is
commonly used.
The matrix configuration is used to take care of the coordination in a
functional organization when the coordination requirements are so high
that the regular functional configuration is ineffective and the interdepen-
2.2 Organizational Configurations 61

dencies between products are so many that a divisional configuration is not

an efficient configuration. The goal is to obtain both functional speciali-
zation and efficiency as well as project focus to realize an end objective
effectively.

2.2.4.1 A Matrix Organization

Michael Allen Company is a service organization-a management consulting
company. The functional dimension is the managing consultant specializa-
tion, and the divisional dimension is the practice client management. The
service deliveries are various projects, and the matrix operations are
obtained through project management-that is, a partner who is responsible
for a given client provides services through project management. See Figure
2.9 (note that the functional and project dimensions are switched in Figures
2.8 and 2.9).

Practice Clienl Managemenl

Figure 2.9. The Michael Allen Matrix Organization (1990)

Source: Conference Board Chart Collection (1991).
62 Chapter 2. Describing the Organization for Design Purposes

Michael Allen is one of the few companies to state explicitly that its
configuration is a matrix configuration. Many matrix configurations are not
evident from the organization chart but can be inferred or can be observed.
Nonetheless, matrix configurations are widespread and rather common, if
not for the organization as a whole, for the various divisions or even
departments.
For many organizations, matrix configuration and associated activities
are suggested by the organization charts as both functions and divisions are
equally highlighted. Table 2.3. shows some ofthose organizations, together
with the two dimensions of the matrix. The dual hierarchy is the essential
characteristic of the matrix configuration; there is a simultaneous explicit
managerial focus on both the divisions (project, customer, program,
product) and the function (supply, manufacturing, sales). Each function will
have a manager, and each division will have a manager.

Table 2.3. Matrix Organizations

Company Division or function Supervision

Royal Dutch) Gas, Metals, Chemicals, Various Regions

(1995 Exploration, Manufactur-
ing, Marketing.

Coopers & Lybrand Finance, Human Resources, Various Client

(1996) Tax, Management Informa- Managements
tion Systems, Consulting

Kirk Tyson International Process Reengineering, Project Directors

Ltd. (1996) Benchmarking, Quick Intel- and Managers
ligence

Marketing Decisions, Inc. Project Managers Marketing repre-

(1996) sentatives

Source: Conference Board Chart Collection (1991).

The coordination in a matrix configuration is very complicated and requires

large amounts of management time. To examine the coordination issues,
consider the implications of changing one element in the matrix. Let us
assume that the manufacturing phase of Product 1 is moved from June to
July. What are other required changes? First, for Product 1 all subsequent
activities must be moved back one month: the delivery time and sales
efforts, for example, must be adjusted. But these adjustments are only the
beginning, since Product 1 manufacturing time has changed; all other
2.2 Organizational Configuration. 63

manufacturing must be adjusted. Similarly, since Product 1 sales effort is

adjusted, all other sales efforts must be adjusted. In a matrix configuration,
the adjustment of one element may cause adjustments for all other
elements. Sometimes, this is called the jello effect: you touch it, and it
moves everywhere. There is a ripple effect throughout the matrix. It is
straightforward to extrapolate that the coordination is difficult, requiring
a great amount of information. The advantage is that these adjustments
can be made and uncertainties can be managed.
The rationale for the matrix configuration is its focus on the customer,
product, program, or markets, its ability to adjust to uncertainties in a
timely fashion, and its ability to use source functional resources efficiently.
The goal is to capture the effectiveness of the division as well as the
efficiency of the functional configuration under uncertainty.

ASS: a matrix organization

The matrix configuration is often used for the small organization or project. ASS is a matrix
organization, and it is very large. ABB has some 250,000 employees, revenues in excess of
US$25B and operates in most industrial countries around the world. How does the ASS matrix
work?
ABS's matrix has two dimensions: a local or country dimension and a business area
dimension. ASS wants to focus on being a local company and at the same time global to
realize the economies of scale and knowledge across borders. A local ABS company develops
its own product strategy, deals with labor unions, markets its products, deals with the
government, assures local capital and banking relations, and generally manages a total
business. The business area managers are responsible and make decisions for product
strategies throughout the whole of ABB. At the intersection, there are some 1100 local
companies whose presidents have two bosses. The matrix is to realize the shared knowledge
from around the world and also take advantage of global markets for given products along the
one dimension. Along the other dimension, ABS is deeply local in each country or region. ASS
has purposeful internal contradictions which must be managed: global and local, small and
large, and decentralized and centralized. The two matrix dimensions focus on local organization
and global product lines and is the organizational means to manage these counterforces. The
individual local ABB companies can be small (there are 1200 of them) and yet it is a huge
company.
Decision making is also a contradiction, as decisions are made at the lowest level, but
decisions can not be made at variance with the global product strategy and knowledge about
the product technologies and efficiencies. The two dimensions must complement each other,
not compete. The business area managers can also be country managers as Well, providing
a link across the matrix. Yet, the balance is difficult to maintain.
ABS ties the matrix together in many ways. English is the common language. ASS has a 21-
page bible which spells out the responsibilities of the matrix managers-an amazingly low
degree of formalization. The bible speaks directly to the responsibilities of the business area
manager and what is meant by decentralization and accountability. ASS's Abacus information
system helps provide the glue and lets managers how they are doing and highlights potential
problems. Even so, ASS's main problem remains communications; it must be worked through-
out the organization and frequently face-ta-face. ABB's matrix managers are a special group
who can keep the foci on local and business area and maintain a balance across a broad range
of issues. These are many of the goals, issues and problems as a small matrix organization
has. ABB's matrix illustrates that the information processing fundamentals apply whatever the
size of the matrix organization.
64 Chapter 2. Describing the Organization for Design Purposes

The matrix configuration is an example of two of Galbraith strategies: the

functional dimension is a hierarchical information strategy, and the
divisional dimension is the creation of lateral relations. The great
advantage ofthe matrix configuration is its divisional effectiveness and its
functional efficiency.
The limitations are numerous. Managers must be comfortable with
uncertainty, willing to share authority, and yet be responsible for team
results. These attitudes and skills can be rare among managers. Matrix
configurations are costly and require lots of managers-one for each function
and project. And finally, a matrix configuration can be disastrous-neither
effective nor efficient. When the matrix team breaks down, conflict emerges
and nothing goes well. In brief, a well-run matrix can be a marvel of
effectiveness and efficiency; a poorly managed matrix can be a disaster and
ruinous. Managerial attitude and skill are essential.

2.2.5 Ad Hoc Configuration

Both single- and dual-hierarchical structures have been presented. In some
organizations there is no hierarchy at all. If a group of people get together
to do a project, they are all experts within their own domain, and the goal
is to complete the project, they may create an ad hoc configuration. See
Figure 2.10.
Frequently, there is an associated staff who may be experts as well.
There are no levels at all. Everything happens at the same level and is only
loosely coupled. There is no centralized decision making and there are no
written rules.

l I
Experts Experts Experts
A B C
r- Staff

I I

Figure 2.10. An Ad Hoc Configuration

Coordination is achieved through many and often heated meetings. The

essential glue is obtained through focus on the common goals that were the
basis for the creation of this organization. The incentive system is very
important in such an organization; there must be a reward for each
individual or group. An ad hoc configuration is not a very stable organiza-
2.2 Organizational Configurations 65

tional configuration. In Galbraith's terms, an ad hoc configuration may be

a team or a taskforce.
One of the often cited examples of an ad hoc configuration is film making,
where various experts-such as sound and light technicians, directors,
actors, and so on-form an ad hoc configuration.

2.2.6 Bureaucracies
The above-mentioned configurations, particularly the functional configura-
tion, have been confused with a bureaucracy. However, a functional
configuration mayor may not be a bureaucracy.
A bureaucracy can be defined as an organization with the following
characteristics:

• Division oflabor
• Well-defined authority hierarchy
• High formalization
• Impersonal in nature
• Employment based on merits
• Career tracks for employees
• Distinct separation of members' organizational and personal lives.

The above is also often called a machine bureaucracy.

These are restatements of Weber's (Gerth and Mills, 1946, pp. 196-198)
original bureaucracy characteristics:

• The principle of fixed and official jurisdictional areas, which are

generally ordered by rules, is followed.
• The principle of office hierarchy and of levels of graded authority, a
firmly ordered system of which higher levels supervise lower levels, is
followed.
• The modern office is based upon written documents.
• Office management usually presupposes thorough and expert training.
• The management ofthe office follows general rules, which are more less
stable, are more less exhaustive, and can be learned.

A primary characteristic, perhaps the primary one, is the adherence to

rules.
A divisional configuration may have all of these characteristics.
Similarly, a functional configuration can be a bureaucracy. Some matrix
organizations may be bureaucracies, although this is less likely.
66 Chapter 2. Describing the Organization for Design Purposes

The purpose of high formalization is in most cases (discussed later) to

obtain standardization. Standardization can also be obtained through
socialization or professionalization. If a bureaucracy is staffed by highly
skilled professionals and if some of the standardization is obtained via the
professionalization, giving the professionals some decision authority, it is
called a professional bureaucracy. Medical clinics, law firms, accounting
partnerships are examples of professional bureaucracies.
We find the two forms of bureaucracies everywhere. We will use the
recommendation that an organization should be of bureaucratic form to
further embellish the recommendation regarding a particular functional or
divisional configuration.

2.2.7 International Configurations

Are there unique configurations for international organizations? In their
research, Bartlett and Ghoshal (1988, 1989) have developed four configura-
tions-global design, multinational design, international design, and trans-
national design-and have described their organizational characteristics.
The organizational characteristics are the configurations of assets and
capabilities, role of overseas operations and the development and diffusion
of knowledge (see Table 2.4).
The multinational design is decentralized on a nationally self-sufficient
basis to focus on the separate local market and to keep knowledge within
the national subunit. This is a classic example of a divisional configuration,
where each country unit is a division.
The global design characteristics are centralized on a global scale with
implementation at the local level and knowledge kept at the headquarters.
Although the correspondence is less than exact, the global design operates
something like a functional organization.
The international design is a hybrid design. Some of its activities are
centralized and others are decentralized. The strategy is to lever parent
capabilities. Knowledge is transferred from the headquarters to the units.
The transnational design is a very complex design. The company's assets
and capabilities are dispersed, interdependent, and specialized. National
units are integrated worldwide. Knowledge is developed by and shared
among the units worldwide. Bartlett and Ghoshal (1989, p. 207) argue that
the transnational design goes beyond a matrix configuration, yet the two
have much in common.
2.2 Organizational Configurations 67

Table 2.4.0rganizational Characteristics of International Structures

Organizational Multinational Global International Transnational

Characteristics

Configuration of Decentralized and Centralized and Sources of core Dispersed, interde-

assets and capa- nationally globally scaled competencies cen- pendent, and spe-
bilities selfsufficient tralized, other cialized
decentralized

Role of overseas Sensing and ex- Implementing Adapting and Differentiated

operations. ploiting local op- parent-company leveraging parent- contributions by
portunities strategies company compe- national units to
tencies integrated world-
wide operations.

Development and Knowledge devel- Knowledge devel- Knowledge devel- Knowledge devel-
diffusion of know1- oped and retained oped and retained oped at the center oped jointly and
edge within each unit at the center and transferred to shared worldwide.
overseas units

Source: Bartlett and Ghoshal (1989, p. 65).

Philips, the Dutch electronics giant, has a classic multinational design.

Each country unit is quite independent of the other in its development,
manufacture, and sales of products. The strength is the exploitation oflocal
opportunities, but cooperation and coordination across borders have been
lacking. Its capability to exploit multinational markets has been ques-
tioned.
Matsushita, the Japanese consumer electronics manufacturer, has a
global design. The Japanese headquarters runs the corporation worldwide.
National manufacturing and sales units implement the headquarters'
plans. Local adaptation is not facilitated.
L.M. Ericsson, the Swedish electronics and telecommunications
manufacturer, has a transnational design. Its operations-research, develop-
ment, manufacturing, and sales-are dispersed among several countries. Yet
these activities are integrated to take advantage of opportunities world-
wide. Large-scale personnel transfer is one means of obtaining the
necessary communications needed for coordination among these several
activities. Although there are many matrix characteristics, the transna-
tional design involves additional coordinating and integrating mechanisms.
Phatak (1992) discusses four basic international configurations. In the
preinternational division phase the domestic organization is expanded with
an expert department and may be further expanded with foreign subsidiar-
ies. Such an organizational structure will be chosen when export activities
are rather low and in a development phase.
68 Chapter 2. Describing the Organization for Design Purposes

In the international division configuration, all international activities

are in one international division. If market diversity and product diversity
are low and foreign activities are relatively small, then an international
division configuration may be the optimal choice.
If the market diversity grows, a more global configuration may be
necessary so that activities are not in one division but in more divisions
based on geographical areas.
The global functional configuration is a functional organization where
each function operates in many international markets and areas. For this
configuration to be advantageous, the organization has to produce narrow
standard product lines.
Finally, Phatak (1992) discusses the multidimensional global configura-
tion, which is an international-oriented matrix structure based on
functional areas and market and product areas.
Habib and Victor (1991) also use four basic international configurations
similar to the configurations presented above: worldwide functional con-
figuration, international division structure, geographic region structure,
and international matrix structure as discusses in the capsule on ABB.
Basically, international configurations are only variations on the more
basic configurations discussed in this chapter, particularly when an
information processing view is adopted. International or not, each
configuration has to be designed in detail, and international issues have to
be taken into account. In this book we primarily address our basic
configurations and only occasionally refer to international issues specifical-
ly. However, our example case, SAS, is a truly international and multina-
tional corporation.
The information processing view is fundamental to organizations, both
domestic and international. In this section on configuration, we have
defined a number of alternative organizational configurations. Organiza-
tional configuration, due to its graphical representation in charts, is
perhaps the best-known characteristic of an organization. We now consider
other equally important organizational characteristics.

2.3 Organizational Complexity

We begin with definitions of organizational complexity and differentiation:

• Organizational complexity is the degree of horizontal, vertical, and

spatial differentiation.
• Horizontal differentiation is greater when there are several small tasks
and specialization by experience, education, and training.
2.3 Organizational Complexity 69

• Vertical differentiation is the number of hierarchical levels between top

management and the bottom of the hierarchy.
• Spatial differentiation is greater when there are many locations of
facilities and personnel.

As the degree of organizational complexity increases, then the difficulty of

coordination issues and the requirements for information processing
increase as well. For example, with a high degree of specialization, each
specialty may have to be coordinated. With a low degree of specialization,
the coordination is internalized either within a group or even within a
person. The complication of coordination in the high specialization case
comes from two sources. First, the amount of information that has to be
exchanged is high. Second, each specialty may have difficulties under-
standing the signal sent from another specialty. Each group has its own
jargon, values, and form of communication. In an international organiza-
tion, coordination may be further complicated by language and culture.
The very nature of the airline business demands a large spatial
differentiation. As SAS expanded abroad, spatial differentiation increased.
Horizontal differentiation is also high for SAS; it has a large number of
specialists, including both flight and ground personnel. The vertical
differentiation has traditionally been high; however, recent organizational
changes to flatten the organization have reduced the vertical differentia-
tion. Overall, SAS has high complexity, where technical task specialization
enhances operating efficiencies.

2.3.1 Horizontal Differentiation

Horizontal differentiation refers to specialization within an organization.
The simple organization shown earlier in Figure 2.1 has no well-specified
horizontal differentiation: as Peter, the top manager, assigns work to
George, Jane, and Jens at will. The prototypical functional configuration in
Figure 2.2 has clearly specified departmental tasks and responsibilities.
The horizontal differentiation measure is relative, and it is greater here
than for the simple configuration. Mack Trucks in Figure 2.3 has a
well-specified, moderate, horizontal differentiation. Horizontal differentia-
tion is a general concept and not restricted to any configuration. The
divisional configuration in Figure 2.5 has a higher horizontal differentiation
than the functional configuration does. Note that at the functional task
level, there are more organizational units here, but it also could be low. GE
has a high degree of horizontal differentiation with its many divisions and
large number offunctions. For the matrix configuration in Figure 2.8, there
is a moderate degree of horizontal differentiation as the project, product,
70 Chapter 2. Describing the Organization for Design Purposes

and customer dimensions do not increase the horizontal differentiation. Ad

hoc configurations can have high or low degrees of horizontal differentia-
tion, depending on the collection of task specialization. Finally, horizontal
differentiation and task specialization are fundamental aspects of
bureaucracy-whether machine or professional.
A recommendation about level of horizontal differentiation has to be
associated with the specific configuration that is considered, but for each
there is a choice. Additionally, the particular kind of specialization may also
depend on the level of professionaliz ation. We have mapped the number of
job titles and the proportion of the employees that hold advanced degrees
or have many years of specialized training into the measure of horizontal
specialization. This view ofhorizontal differentiation as a practical measure
of organizational complexity has a general agreement in the literature
(Hall, 1991).
Specialization may increase efficiency. People may be selected to do
what they do best, and by doing it regularly, they may learn how to do it
well. Additionally, if persons or departments do many different things, then
there are costs involved in moving from one task to another. These setup
costs can be minimized by proper specialization. Specialization can be based
on functions, products, customer groups, or geographical areas. It also may
be based on production processes.
However, increased horizontal differentiation and task specialization
come with a cost. Increased horizontal differentiation creates a need for
coordination among the specialized units. Without specialization, each
individual cut and topped his own nails. But consider two nail makers-one
who specializes in cutting nails and another who puts tops on them. The
nail cutter and the nail topper must coordinate their activities so that the
number of nails cut equals the number of nails topped. The two individuals
can coordinate their activities very simply by talking with each other.
However, 10,000 individuals with specialized tasks need to be coordinated,
a problem that is not easily resolved. We now have a very large informa-
tion-processing task. There are numerous information processing strate-
gies. A 10 ,OOO-person organization will have some organizational configura-
tion with a hierarchy that provides the dominant information processing
approach. Normally, information for coordination and control follows the
hierarchy. Galbraith (1973, 1974) offers various strategies that either (1)
increase information processing by becoming faster within the hierarchy or
bypass the hierarchy with liaison and integrative mechanisms (2) decrease
the need for information processing by rearranging the tasks into more
nearly independent units or standardize the operation. Whatever scheme
may be used to coordinate the specialized activities, the general relation is
that greater horizontal differentiation increases the need for information
processing to obtain the required coordination and control.
2.3 Organizational Complexity 71

2.3.2 Vertical Differentiation

Vertical differentiation relates to the depth of the hierarchy. Since the
configuration may not be symmetric, the measure has to take that into
account. We follow Hall, Hass, and Johnson (1967) and use the number of
levels that separate the chief executive from those employees working at
the bottom of the organization combined with the average number of
vertical levels (total number oflevels per number of subunits) as the basis
for our measure of vertical differentiation. The measure shows that the
vertical differentiation may not be symmetrical. It may be deeper in one
area than in another.
The simple configuration shown in Figure 2.1 has only two levels and
thus a low degree of vertical differentiation. The functional configura-
tion-that of both the prototype in Figure 2.2 and Mack Trucks in Figure
2.3-has more vertical levels and thus higher vertical differentiation.
However, we cannot measure precisely the degree of vertical differen tiation
because the configurations are not complete to the lowest level of the
organization. Similarly, the divisional configurations-the prototypical and
GE-have considerable vertical differentiation, but the charts are incom-
plete. Matrix organizations are usually asymmetrical where the functional
side is the deeper than the project, product, or customer side, and thus the
functional dimension yields the degree of vertical differentiation. An ad hoc
configuration is usually not very deep with no levels at all. Again,
bureaucracies, particularly machine bureaucracies, tend to have many
vertical levels.
The horizontal and vertical differentiation measures are not unrelated.
The higher the horizontal differentiation, the higher the vertical differenti-
ation; the relationship depends on the span of control. The span of control
defines the number of subordinates a manager directs. If the span ofcontrol
is wide, many subordinates are directed. If it is narrow, few are directed.
The optimal span of control depends on the complexity of the task that
subordinates performs and the subordinate's skill level. This issue is dealt
with in Chapter 3.

Delayering, organizational complexity and IT

Organizations are delayering; what does that mean? It means to take out a vertical layer in the
hierarchy, by eliminating an organization layer and its jobs. Delayering usually takes place in
the middle of the organization and is likely to affect middle management. So an organization
that had five levels will now have four levels, frequently with new titles and redefined functions.
Delayering does not eliminate the activity or necessarily mean that the old level did no work;
it means that this work will have to be done by individuals above and below in the hierarchy.
Adjustments will have to be made from the bottom to the top of the hierarchy, not just the
nearby levels. Renaming all the levels is a way to signal the extent of the change as well as
new job descriptions and titles. So, delayering is a large organizational change which can affect
everyone in the organization.
72 Chapter 2. Describing the Organization for Design Purposes

Delayering decreases the vertical differentiation of the organization. We defined the vertical
differentiation as the number of levels from the bottom to top of the organization. If one level
is eliminated, then the vertical differentiation is lower and then, the organizational complexity
has decreased. It is a less complex organization as we have defined the term organizational
complexity. The organizational complexity can be changed by delayering, but it can also be
changed by decreasing the horizontal differentiation.
Horizontal differentiation is a measure of how finely the work is divided upon into jobs or how
specialized the jobs in the organization are. Very specialized jobs indicates a high horizontal
differentiation. When the organization enlarges the jobs and tasks, the horizontal differentiation
is less. When two jobs are made into one or the work is less specified, then the horizontal
differentiation is less. This, too, decreases the organizational complexity.
Organizations with a high level of organizational complexity require greater information
capacity to coordinate the across the various tasks and jobs, and also up and down the longer
hierarchy. The organizational design tradeoff is to balance the costs of the greater information
processing with the benefits from the greater organizational complexity. With new information
technology and lower relative information processing cost the tradeoffs involve all three
variables. We would expect organizations to become less complex and delayering is one way
to decrease the organizational complexity as we realize the returns from new information
technology.

Increased vertical differentiation also comes with a coordination and

information processing cost. First, increased vertical differentiation
increases the number of individuals in the organization, and they must be
paid; these individuals process information-orders and information up and
down the hierarchy. Second, increased vertical differentiation increases the
number of individuals who handle a given piece of information. Consider
again our two-person nail example. The organization can become a three-
person organization by creating a boss or top manager, whose job is to
coordinate the cutter and the topper. The boss does not cut or top but
gathers information, makes plans and tells the cutter and topper what to
do. These are valued activities and increase the total production of nails.
With 10,000 individuals, there are likely to be a large number of vertical
levels. Consider again the earlier examples of Mack Trucks and GE.
Recently, many organizations have addressed the issue of number of
vertical levels. Generally, increased efficiency in information technology
and systems have permitted the elimination of costly middle-management
levels. The current trend to eliminate middle management and flatten the
organization decreases the vertical differentiation of the organization.

2.3.3 Spatial Differentiation

The third measure related to the organizational complexity of the organiza-
tion is the geographical dispersion of the activities in the organization. It
is relatively easy to measure this concept. We use the number of geograph-
icallocations, their average distance from the main office, and the location
of the dispersed personnel as our measure of spatial differentiation (Hall,
Hass, and Johnson, 1967). The simple organization is likely to have one
2.4 Formalization 73

location. Mack Trucks has one manufacturing location where the functional
heads are located. However, GE has locations all over the world. In general,
international organizations have a high degree of spatial differentiation.
Little can be said in general for the ad hoc configuration or bureaucracy.
It is quite clear that increased organizational complexity implies that
additional efforts have to be put into coordination and control. This is
particularly true for the horizontal and vertical differentiation. The effect
of increased coordination and control due to spatial differentiation depends
at least on the information system capabilities. In the old days this concept
was particularly important because geographical distance slowed down
communication-both with respect to amount and complexity. This may no
longer be the case. With modern information systems, it is possible to
communicate as easily with a person 5,000 miles away as with the person
next door. In other cases, electronic communication may be inappropriate,
and the location of personnel is very important.

2.4 Formalization
For many organizations it is efficient to obtain a standardized behavior of
the members ofthe organization. This standardization can lead to low cost,
high product quality, and generally efficient operations.
Formalization is one way to obtain such standardized behavior and, as
such, is a means to obtain coordination and control. Formalization
represents the rules in an organization. Whether the rules have to be in
writing has been discussed in the literature, and in most empirical studies
the measurement of formalization has been related to written rules (Hall,
1991). The rules and procedures can be many and fine tuned or be few and
not so fine. The formalization often will vary depending on the particular
part of the organization. The production department is more likely to have
a high degree of formalization than the R&D department.
The empirical studies on formalization have usually measured formal-
ization in terms to the quantity of written rules. Additionally, formalization
involves measurement and degree of compliance.
We measure formalization as the degree to which there exists formally
stated rules, in writing. Additionally, we measure how much latitude
employees are allowed from standards-the level of compliance. It is
important to consider whether supervisors and managers make decisions
under rules, procedures, and policies and whether these are in writing.
Generally having more written rules means having higher formalization.
Formalization may be seen as the means to both increase the informa-
tion processing capacity and decrease the demand for information
processing. When the information processing demand originates from more
74 Chapter 2. Describing the Organization for Design Purposes

sales, more customers, or more clients, standardized rules and general

policies can keep information processing at low levels in the hierarchy. At
the same time, one also may argue that the amount of information that has
to be processed for each product, customer, or client is reduced: "Formaliza-
tion of a decision making language simply means that more information is
transmitted with the same number of symbols" (Galbraith, 1974).
SAS is a highly formalized organization. There are many, many rules on
how to accomplish tasks-change an airplane tire, inspect ajet engine, write
a roundtrip ticket, and so on. There are hundreds of thousands of detailed
instructions. Further, there are many unwritten rules-particularly for
professional employees such as accountants.
Recently, SAS has eliminated many of the fine rules and focused more
on serving the customer than applying existing rules. It has become less
bureaucratic. Even so, there remain very many rules on operations that are
required to obtain coordinated behavior in an efficient manner.
In contrast, a simple organization has few, if any, written rules, showing
very low formalization. Mack Trucks, with a functional configuration, has
elaborate written rules and follows them to a large degree. GE, likewise,
has a large number of stated rules, although formalization varies across
divisions. Matrix configurations are less bound by rules and rely heavily on
give and take to obtain coordination. Ad hoc configurations abhor rules. And
finally, written rules are fundamental to a bureaucracy and provide a
fundamental characteristic of a bureaucracy. For a machine bureaucracy,
the rules are written; for a professional bureaucracy, they are brought to
the organization by the professionals.
Formalization must not be confused with standardized behavior per se.
Standardized behavior can be obtained by a number of different means. The
rules may not be by the organization but by a professional association.
Medical Doctors and CPA's do a number of things in a standard way
because their professional organizations tell them to do so. Professionals
also may do things in a standard way due to their training. Professionaliza-
tion is the term for standardization obtained in this manner.
Social norms and group pressure may also lead to standardized
behavior, and the organization mayor may not be able to control these
norms. The behavior of employees also can be modified by the use of
incentives. Some incentive systems can lead to standardized behavior, and
others may not. If the incentive is associated with process, it usually leads
to standardized behavior. If it is linked to results, it may not.
2.5 Centralization 75

2.5 Centralization
Centralization is the degree to which formal authority to make discre-
tionary choices is concentrated in an individual, unit, or level (usually high
in the organization). Decentralization is low centralization. We measure
centralization by how much direct involvement top managers have in
gatliering and interpreting the information they use in decision making and
the degree to which top management directly controls the execution of a
decision. The above issues are important in determining who has authority
to influence a decision aside from actually making the decision.
Centralization is related to who makes which decisions. This includes
establishing the budget, exercising control over evaluations and rewards,
and being involved in hiring and firing personnel. It also includes issues
such as purchasing supplies and equipment and establishing of programs
and projects. Finally, related to formalization is the issue of how exceptions
are handled. No matter how many rules there may be, they cannot cover all
possible situations.
Traditionally, SAS has been a highly centralized organization with most
decisions made at the top. More recently, SAS has become more decentral-
ized where midlevel managers are given much more discretion over service
decisions and latitude to meet the customers' needs. Yet at the same time,
strategic issues remain at the top.
A simple configuration is usually very centralized, and the top manager
makes all the important decisions. Sometimes, a simple organization can
be decentralized when the top manager lets others decide operational
issues. Functional configurations require coordination across functional
units; centralized decision making is a frequent approach. Mack Trucks is
rather centralized. A divisional configuration does not require operational
coordination across divisions and lends itself to greater decentralization.
However, top management is usually deeply involved in budgetary matters
and frequently in strategic decisions, even for the divisions. Each division
may be centralized or decentralized. Adhocracies can be either, but somehow
the required coordination must be realized. Bureaucracies can be either;
their desired predictable behavior and outcomes can be obtained by
formalization and low centralization. Exception decisions are normally
centralized.
From the above it is obvious that centralization is a means for coordina-
tion. The relationship between centralization and formalization has been
discussed by a number of researchers (Zeffane, 1989). Both are means of
coordination. There can be a rule telling what to do, or the top management
can each time tell what to do. Therefore, in cases where both centralization
and formalization could be high, formalization is most often lower than it
76 Chapter 2. Describing the Organization for Design Purposes

otherwise would be. This issue is particularly dependent on the size of the
organization and is discussed in the next chapter.
Centralization and decentralization are directly related to the informa-
tion processing capacity in an organization. The notion of bounded
rationality assumes that individuals have limited capacity to read, store,
and interpret information. The actual amount of information an individual
can process depends on skill and educational levels. As the information
processing demand increases, more individuals have to be involved in
decision making, and decentralization increases. Empowerment of
individuals usually means greater decentralization and decision making at
lower levels in the organization.

2.6 Coordination and Control

Lawrence and Lorsch (1967) concluded that more differentiation is
required, more integration is needed. Basically, organizations are formed
to achieve a set of goals. For cost efficiency the work in the organization
may be divided into a number of separate tasks. To obtain common goals,
the activities must be coordinated.
So far four major means to obtaining coordination and control have been
presented: formalization, centralization, incentives, and lateral structures.
They may be used in some combination, and there are a number of ways to
implement each of them. Coordination and control have two sides. One is
to make sure that enough relevant information is available at the right
time to be able to make the right decisions. The second is to make sure that
the right decisions are made. Incentives can be related to both process and
results, and can be associated within either groups or individuals. And
incentives may be unrelated either to process and results, to groups, or to
persons. Lawrence and Lorsch (1967) studied coordination intensity in the
different industries and found that the means to obtain that level of
integration was very different, even when the level of integration was the
same. A fit between all the various elements in the organizational design
has to be obtained.
Summarizing, coordination can be obtained by a number of means
ranging from direct supervision to autonomous groups using both hierarchy
and formalization and including incentives. Additionally, various kinds of
lateral procedures may be appropriate. It is also obvious that a proper
information system is important. The information system can be character-
ized according to a number of dimensions. An obvious dimension is the
amount of information that can be transferred and processed. Additionally,
media richness is of interest. Media richness relates to the type of
information that can be processed and the type and speed of feedback. Face-
2.7 The Organizational Structure of SAS 77

to-face meetings have higher richness than a short written note. It is

interesting to observe that the revolution in information technology
increases both of these dimensions: amount of information and media
richness.
Information systems are also categorized as sequential or concurrent. In
some negotiations it may be important that you use a sequential informa-
tion system, but when LEGO plans to release the same spaceship design on
the same day in 40,000 stores, then the system has to be concurrent. New
fax systems and electronic mail are modem information systems that can
be both sequential and concurrent. In the choice of coordination and
information system, the particular combination of tools is important. The
choice has to fit the demand for information processing capacity.

2.7 The Organizational Structure of SAS

At the beginning of this chapter, we briefly reviewed the history of the
Scandinavian Airline System over the last fifty years. SAS evolved greatly
due to market changes, new technologies, and new strategies. We now
review its organizational structure in more detail. As discussed previously,
1975 was a turning point when the environment changed. Economic
conditions for the airline industry since 1950 are summarized in Table 2.5.

Table 2.5. Economic Conditions in Aviation since 1950

1950-1975 1975-Present
Stable and rapidly growing market Stagnant market
Slowly rising oil prices Rapidly rising oil prices
Protectionist aviation policies Liberalization of aviation
Cartel formation and market sharing Competition, including price
Stable profitability Deficits

Despite the fact that general economic conditions changed around 1975,
SAS did not change its strategy until 1981. In 1981, SAS found itself with
declining productivity and overcapacity. Its image in terms of service and
punctuality was declining. From 1975 and onwards, the market became
more competitive. The market was more segmented, and new entrants tried
to exploit niches. There was an emphasis on both costs and services.
Generally, the situation had shifted from an orientation to operations to an
orientation to markets-a truly dramatic shift. Uncertainty was greatly
increased, requiring greater information-processing capacity.
78 Chapter 2. Describing the Organization for Design Purposes

However it was not until 1981 that SAS realized that its old strategies
did not work in the new situation. SAS had operated from 1975 to 1981
with significant situational and contingency misfits, and contrary to
previous periods of stable profit, losses were substantial.
In order to solve the crisis, the board brought in new management in
1981 and a new strategy was immediately developed for SAS. The main
strategy was to make SAS "the businessman's airline", focusing on service
and functionality.
Changes in the strategy were followed by a new organization. The old
organization was a functional structure with a high degree of centralization
and formalization. There was an emphasis on rules and programs for
coordination with an authority and control hierarchy to ensure proper
implementation and deal with exceptions, which were kept to a minimum.
The new organization was more complex, being a mixture of the
divisional form and the functional form. Figure 2.11 shows the structure
that was implemented in 1981-1982.

Figure 2.11. SAS's Organization, 1981-1982

The commercial division was now organized according to geographical

areas, as shown in Figure 2.12. These route sectors were basically profit
centers, as were the other divisions. The new organization was more
decentralized than the old one.
2.7 The Organizational Structure of SAS 79

SAS Group
Management

Chief Operatklg officer

f· ·· · · · ·· · s~s·· ·· ····· ·~
SAS SAS SAS
Denmark Business division : Intema1ionaJ hotels : Leisure
~ .... ....................:
f · ········ s~s · · ········~
:............ ............. :
SAS SAS
: Diners club Nordic :
NoIway Production division Service partner
~... .. ... ...... -... ...... . . ............... ......... .
. ...... ........ ... ..--....
..
SAS SAS : SAS :
Sweden Support division Trading

SAS SAS o SAS Business ares

Intema!ional Subsidiaries
Other business areas

Figure 2.12. SAS Organization, January 1993

From the mid eighties and onwards, deregulations increased the number
of competitors, and price competition became more intense than ever. SAS
was badly situated to take on price competition because of its high fixed
costs. As a means to differentiate its services SAS introduced the total
travelling concept, offering customer's a total traveling package which, in
addition to the plane ticket, included ground transportation, hotel bookings
etc. The total traveling concept necessitated wide spanning investments in
many non airline businesses. In 1986 SAS reorganized its businesses into
five independent business units: the airline, SAS Service Partner, SAS
International Hotels, SAS Leisure and SAS distribution. The rationale was
that each of these businesses faced very different strategic demands and,
therefore, were required to have their own management team to allow more
aggressive business development. The new SAS organization was further
decentralized and attempted to create relatively independent units that did
not require a great deal of coordination. The SAS airline and SAS
International hotels were treated as relatively independent businesses-that
is, self-contained tasks (following Galbraith, 1974). SAS had backed away
from an earlier strategy of integrating the airline and hotel business into
a single identity for the customer.
80 Chapter 2. Describing the Organization for Design Purposes

In April 1994, Jan Carlzon was succeeded by Jan Stenberg, fonner CEO
of the Swedish Ericsson. With the appointment of Stenberg a new strategy
was fonnulated. From the very beginning, Stenberg carried through a
severe rationalization program, including the selling off many of the
company's side businesses, among others Diners Club Nordic franchise,
SAS Leisure, SAS Service Partner and several hotels. Contrary to Carlzon,
Stenberg aimed at satisfying customers' demands for frequent departures
by making alliances, rather than mergers with foreign airline companies.
To carry through his rationalization program, Stenberg found it necessary
to increase the level of centralization of the company.
The deregulations as well as the many strategic alliances increased
SAS's information processing requirements. SAS was recentralized to
initiate the changes needed to reduce costs, but at the same time the
company was dependent on the innovativeness and adaptability enabled
through decentralization. Although SAS was made leaner and more
centralized with the appointment of Stenberg, the structure and operations
of SAS's three parent companies remained unchanged. SAS can therefore
still be described as a mixture of a divisional and functional structure. The
lateral communication of the functional structure allows for the informal
communication necessary for innovation and adaptation to changes in
customers' wishes and demands. The divisional structure and its higher
level of centralization allow the top management to control expenditures
and assure the company's efficiency. As a further means to coordinate the
actions oftop management and front line personnel, the company is making
increased use of professional managers. The role of the professional
managers is to have the necessary professional insight to understand top
management's decisions, and effectively communicate the implications to
front line personnel. At the same time they communicate front line
individuals' ideas as well as market trends to the top management.

2.8 Summary
In this chapter, we considered how to describe an organization for design
purposes. For management and design purposes, we argued that an
information processing description is relevant: how does the organization
use information to coordinate its activities and accomplish its goals?
A number of organizational concepts were introduced and related to
coordination. The organizational configuration is the way the organization
divides its work into smaller tasks for management-simple, functional,
divisional, matrix, ad hoc, or bureaucracy. An organization may have a
number of configurations operating at the same time in different parts of
the organization. Complexity, formalization, and centralization were
References 81

defined and illustrated. Additionally, means for coordination and control

and incentives issues were introduced. Throughout we have referred to
SAS, Mack Trucks, GE, and Michael Allen to illustrate the concepts.

Notes
1. A more detailed description of SAS can be found on the accompanying
CD-ROM.

References
Bartlett, Christopher A., and Sumantra Ghoshal. 1989. Managing Across Borders:
The Transnational Solution. Boston: Harvard Business School Press.
Bartlett, Christopher A., and Sumantra Ghoshal. 1988. "Organizing for Worldwide
Effectiveness: The Transnational Solution." California Management Review,
54-74.
Buraas, Anders. 1972. Fly over Fly. Olso: Gyldendal Norsk Forlag AlS.
Conference Board Chart Collection, 1991. New York: The Conference Board.
Galbraith, Jay R. 1973. Designing Complex Organizations. Reading, MA: Addison-
Wesley.
Galbraith, Jay R. 1974. "Organization Design: An Information Processing View."
Interfaces, 4(3), 28-36.
Gerth, H. H., and C. Wright Mills. 1946. From Max Weber: Essays in Sociology. New
York: Oxford University Press.
Habib, Mohammed M., and Bard Victor. 1991. "Strategy, Structure and Perfor-
mance of U.S. Manufacturing and Service MNCs: A Comparative Analysis."
Strategic Management Journal 12, 589-606.
Hall, Richard M. 1991. Organizations, Structure Processes, & Outcomes. Englewood
Cliffs, NJ: Prentice-Hall.
Hall, Richard M., J. Eugene Hass, and Norman Johnson. 1967. "Organization Size,
Complexity, and Formalization," American Sociological Review, 32(6) 903-912.
Hofstede, Geert. 1991. Kulturer og Organisationer. Copenhagen: Schultz.
Lawrence, Paul R., and Jay W. Lorsch. 1967. Organization and Environment.
Boston: Harvard Business Press.
Mintzberg, Henry. 1979. The Structuring of Organizations. Englewood Cliffs, NJ:
Prentice-Hall.
Obel, B!Ilrge. 1986. "SAS: Changes in Competition, Strategy and Organization." In
R. Burton and B. Obel (Eds.), Innovation and Entrepreneurship in Organiza-
tions. Strategies for Competitiveness, Deregulation, and Privatization. Amster-
dam: Elsevier.
82 Chapter 2. Describing the Organization for Design Purposes

Phatak, Arvid V. 1992. International Dimensions of Management. Boston: DWS-

Kent.
Taylor, William. 1991. "The Logic of global Business: An Interview with ABB's
Percy Barnevik." Harvard Business Review. March April, 91-105.
Williamson, O.E. 1975. Markets and Hierarchies: Analysis and Antitrust Implica-
tions. New York: Free Press.
Zeffane, Rachid. 1989. "Computer Use and Structural Control: A Study of
Australian Enterprises." Journal of Management Studies, 26(6), 621-648.
CHAPTER 3

Management and Leadership Style

3.1 Introduction
The Institute of Applied Computer Science, in Odense, Denmark, is a small
research-oriented organization located at Science Park. Its main activities
are related to high technology applications of computer science in various
organizations and projects. It has been involved in technology-transfer
projects for a number of European companies. The basic idea is that
Applied Computer Science will transfer ideas from research laboratories
that can be used in practice. It has particularly targeted companies with
relatively low research and development budget that do not have staff that
can make the transfer. For these companies, a technology transfer project
is a rare event. The activities are typically done in the framework of
national research programs and European Union (EU) R&D programs such
as ESPRIT, BRITEIEURAM, SPRINT, and COMETT. The Institute of
Applied Computer Science is a small organization currently employing
seventeen highly skilled employees. It was established in 1986 at the
request of local industry, which needed to better organize the transfer of
knowledge from Odense University and a technical college. A group oflocal
businesspeople and representatives from Odense University and the
technical college formed the company, raised a small amount of capital, and
hired a teacher, Benny Mortensen, at the technical college to run the
company. Benny Mortensen had good contacts both at the University and
in local businesses and quickly linked the company to EU grants. His
knowledge and enthusiasm have contributed significantly to the success of
the company. In particular, he has established cooperation with ten
European partners. The company has developed expertise in project
management and has managed both small and large EU projects. Benny
Mortensen has created relationships with more than fifty Danish compa-
nies. The Institute of Applied Computer Science has worked for and with
European companies such as British Aerospace, Agurta Helicopters, MARl
computer systems, and Verilog, among others. Additionally, he has working
contracts with research institutes and universities in Milano, Madrid, Kiel,
Delft, Utrecht, London, Manchester, and Oxford. The organization is
relatively loose, and the CEO spends many days on the road getting new
contracts, but he knows what goes on in each project. He prides himself on
84 Chapter 3. Management and Leadership Style

actively seeking out new business opportunities that challenge the

organization. He wants the organization to grow, but growth is not the
primary goal. The CEO's enthusiasm is his own approach to motivation,
and he prefers to lead by example, believing that others will follow with
equal enthusiasm and hard work. He likes the technical details of each
project but does not impose his views on the group. Ifprojects fit his longer-
range goals for Applied Computer Science, he does not interfere but gets
involved because of personal professional interest and his commitment to
the firm's high-quality reputation. The CEO's management style definitely
helps shape the firm-both its strategy and organizational structure.

3.1.1 Management and Structure

More generally does the structure determine the leadership style and type
of personnel, or does the reverse hold true? From the literature it is not
clear that any cause-effect relationship exists. Correlations and fit are more
easily demonstrated. In our view, the cause-and-effect relationship is not
as important as the fit; cause and effect may work in both directions. For
a given management group, management will try to tailor the structure so
that it fits the needs of the management. Child (1972) argues that
management has considerable discretion in its choice ofthe organizational
structure-more discretion than contingency theory implies. First, the
contingency factors of size, technology, environment, and strategy are
incomplete determinants of organizational structure. Thus, management
has a choice of structure within constraints given by the contingencies.
Second, at a higher level, management can choose the contingencies
themselves over time. Management can change the organization's size, its
technology, and the environment it operates in and adapt its strategy. It
can adopt, and probably should adopt, a new organizational structure. At
the first level, the management has discretion within the contingencies as
constraints. And at the second, the contingencies themselves can be
changed over time to give an even broader range of organizational choices.
Similarly, management can adapt to structure. The many executive courses
and training programs available today are examples of efforts to change the
attitudes and skills of managers.
Robbins (1990) argues that management and its power relations account
for a significant part of the variation in the choice of structure. Recently,
Lewin and Stephens (1994) have argued that "social-psychological attitudes
of chief executive officers and general managers are critical contingencies
in organizational design and strategy." Management will choose the
structure, within the limits given by the other contingency factors. But in
some situations no feasible solution exists: it is not always possible for the
3.1 Introduction 85

management to find a structure that both fits their needs and creates a fit
with the size, technology, environment, and strategy of the organization. In
many cases, practice has shown that a change in the structure causes a
change in the management group. We have seen many examples of
organizations where a change from a functional configuration to a
divisional configuration accompanied the introduction of a new CEO, as
was the case for SAS. In the functional configuration, the CEO is involved
in more detailed, short-time decision making; in the divisional configura-
tion, the CEO is mainly involved in long-term, strategic resource allocation
decisions. "Obviously ... incorporating managerial philosophies into current
structural contingency theories is of major importance" (Miles and Creed,
1995). Later in this chapter we will examine the management and
leadership styles of recent SAS CEOs and the associated changes in
organizational structure and design.
Similar concerns relate to the fit between the structure and other
employees than the CEO. There are managers at all levels, and their views
and style generally affect the operation of the organization. Individual
managers play an important role, as does the way they interact. In general,
the organization is an information processing social unit, and the way the
individual process and exchange information is an important factor for the
functioning of the organization. Interaction among people depends to a
great degree on organizational structure and properties. It can therefore be
expected that people in the organization with decision-making power will
try to influence the choice of organizational structure, as discussed above.
For the organization to function well, there has to be a fit between what
the organization wants to do, how it wants to do it, and the people who have
to do it. It may be desirable to have a very complex and decentralized
organization with a low degree of formalization. However, if employees do
not have the necessary skills or will not assume new responsibilities, such
a structure may not work. Also, individuals with a need to assume respon-
sibilities may not function well in a very centralized structure with a high
degree offormalization. The chosen structure should accommodate the type
of people who work in the organization, or the organization should hire and
train people so they can function well. The information processing capacity
of the individuals in the organizations affects the information processing
capacity of the organization. The use of information is an important issue
in this context (Arrow, 1974). Professionals may develop their own
terminology that enables them to communicate efficiently among them-
selves, but makes it difficult for them to communicate with others. The
degree of horizontal differentiation may require or prohibit such develop-
ments. For international organizations multiple language skills are
important. In general, organizational culture or climate play an important
role in the way individuals and organizations process information. From
86 Chapter 3. Management and Leadership Style

this point of view, personnel planning and education are investments in

information-processing capacities. There should be a fit between the
demand for information processing and individual capacity. This kind offit
is an important aspect of organizational behavior (Schermerhorn, Hunt,
and Osborn, 1991). In Chapter 5 the relation between the climate in the
organization and the organization structure is treated and related to the
way people process information.
In this chapter we focus on the fit between management style, leader-
ship and organizational structure. The neo-information processing view of
organization is the theoretical basis for integrating the management and
leadership style to the choice of the organizational structure.

3.2 Literature Review

Leaders and managers have different skills, values, and personalities. The
literature on leadership is extensive. Hunt (1991) claims that there are
more than 10,000 empirical studies on the subject. Different organizational
structures require different skills and attitudes. Therefore, structure affects
the desired organizational culture, the desired type of manager or leader,
and the requisite skills and education of the employees. A number of
authors have developed leadership behavior typologies. See Hunt (1991) for
a survey.
However, in this chapter we are more concerned with the counter
relation: does the management or leadership style affect the choice and fit
of the organizational structure? In the design phase, we argue that
management or leadership style is an important contingency for the choice
of structure. This question is particularly important for small and medium-
size organizations that are owner run; however, it is an important
contingency for all organizations.
In this chapter, we develop the framework for management style and
leadership for the organizational design recommendations for the Organiza-
tional Consultant.
A classic way to categorize leadership styles is by applying McGregor's
(1969) Theory X and Theory Y. He suggests a specific view ofthe skills and
attitudes of the members of the organization. Theory X presumes that
people dislike work, shirk responsibility, and seek formal direction
whenever possible. They place security above all. They must be coerced,
controlled, or threatened with punishment to achieve desired goals. In
contrast, Theory Y posits that people like to work and will exercise self-
direction and self-control. According to Theory Y, people accept and seek
responsibility. Many Theory Y people have a high degree of creativity and
can make good decisions. A Theory X or Theory Y leader may develop a
3.2. Literature Review 87

leadership style that fits his or her view ofthe members ofthe organization.
From an information processing point of view, Theory Y individuals can
process more information and more complex information than Theory X
individuals. Theory Y information is more forward looking and general;
Theory X information is more control oriented and detailed.
Kotter (1988) makes a similar distinction between a leader and a
manager (Hunt, 1991, pp. 198-199). A manager does planning and
budgeting, monitors results in detail, and develops planning, staffing, and
delegation structures. A leader develops future visions, communicates with
those whose cooperation is needed, and uses motivation and inspiration
rather than controlling and problem solving. A manager is more internally
oriented than a more externally oriented leader. A leader can process more
complex information; it is unclear which leader type-leader or man-
ager-processes the greatest amount of information.
Mintzberg (1980) developed three major groups of leadership roles of
interpersonal, informational, and decisional roles: (1) interpersonal roles
include those of figurehead, leader, and liaison; (2) the informational
category includes the roles of monitor, disseminator, and spokesperson: and
(3) the decisional category consists ofthe roles of entrepreneur, disturbance
handler, resource allocator, and negotiator. Mintzberg argues that each
leader or manager plays each of these roles to some extent over time. The
importance and intensity of each role vary with the level in the organiza-
tion and the type of organization.
Luthans, Hodgetts, and Rosenkrantz (1988) have developed a typology
with four groups: (1) routine communication consisting of exchanging
information and handling paperwork; (2) traditional management including
activities such as planning, decision making, and controlling; (3) network-
ing including the activities interacting with outsiders, socializing, and
politicking; and (4) human resource management consisting of motivating,
disciplining and punishing, managing conflict, staffing, and training and
development. Again, it is hypothesized that all leaders are engaged in all
of these activities to a certain degree. Particular leadership positions
require an optimal balance between the activities and their implementa-
tion.
Both Mintzberg's and Luthans, Hodgett, and Rosenkrantz's categories
are related to different types of information and how they are handled. A
leader performs some roles, and a manager performs others. The type of
organization affects the type of decisions that the manager or leader has to
make and how he or she has to process information. Similarly, YukI's (1981)
typology is related to the information network, information processing, and
decision making. He also developed a typology of four groups each with a
number of subgroups:
88 Chapter 8. Management and Leadership Style

• Making decisions: Consulting and delegating, planning and organizing,

problem solving
• Giving or seeking information:Monitoring, clarifying, informing
• Building relationships: Supporting, networking, managing conflict and
team building
• Influencing people: Motivating, recognizing and rewarding,

The four groups relate both to skills and behavior. A particular manager or
leader will play various roles, depending on the particular organization.
Managers can, to some degree, be taught how to play these various roles.
However, the personalities and values of the particular manager or leader
will also determine how this manager or leader will play each role.

DECENTRALIZATION
ORIENTATION
Flexibility

~~ 1. Understanding ~+
yourseH and others
2. Interpersonal
conmunicalion 1. Living with change OPEN SYSTEMS
HUMAN RELATIONS
MODEL 3. Developing 2. Creative thinking MODEL
subordinates 3. Managilg change

1. Buikllng and mailtailil

1. Teambuiklilg ·•••MENTOR INNOVATOR apower base
2. Participative 2. Negotiating agreement
decision making andcornrniJment
3. Conflict management 3. Negotiating and selling
ideas
FACILITATOR BROKER
....
1. Receiving and MONITOR ·••• PRODUCER 1. Personal productivity
organizilg mormation IrId motivation
2. Evaluatilg routing 2. Motivating others
infonnation 3. Tme and stress
3. Respond'llg to management
routine infonnation
/COORDINATOR DIRECTOR

INTERNAL 1. PlMnilg RATIONAL

1. Taking inHiative
PROCESS MODEL 2. Organizing GOAL MODEL
2. Goal setti1g
3. ControUIng 3. Delegating
eHectively

Control
CENTRALIZATION
ORIENTATION

Figure 3.1. Competing Values

Source: Hunt (1991, p. 163).
3.2. Literature Review 89

Autocratic behavior and democratic behavior have been viewed as two

extreme positions (Likert, 1967). This is also related to the leader-
participation model proposed by Vroom and Yetton (1973), where the
information and who makes the decisions are the most important factors.
Many of these issues are depicted in the competing-values leadership
approach developed by Quinn and Rohrbaugh (1983). It is summarized in
Figure 3.1.
Complementary values are next to each other, and contrasting values
are opposite. For example, one may want the organization to be adaptable
and flexible but also stable and controlled. In the mind of the leader, these
states are assumed to be opposites, and these opposites cannot exist at the
same time. Quinn and his associates argue that managers narrowly pursue
one viewpoint and need to break out and pursue other viewpoints,
particularly opposite ones. Individuals who are able to balance these values
are referred to as master managers or Janusian, after the Roman god. The
competing values approach also captures many of the buzzword leadership
styles like charismatic leadership, supportive management, etc.
A review of the management or leadership typologies reveals that many
of the dimensions are similar to the dimensions used to describe the
organizational structure. The delegation issue relates to decentralization.
Planning and control relate to formalization. In general, the dimensions in
Figure 3.1 relate to all the informational issues that are important for
organizational design, including reward and motivational issues.
The important issue is how effectively the various leadership styles and
their information processing capacity affect organizational structure.
Fieldler (1977) proposed a contingency model ofleadership behavior and its
effect on the task structure. He found that if the task structure is struc-
tured, then the leader position should take a strong power position, and if
the task structure is unstructured, then the leader position should be weak
with respect to power. From our information processing point of view, if the
task structure is unstructured, then the demand for information processing
is high and decentralization is appropriate; a highly structured task is
compatible with centralization.
Miller and Toulouse (1986), Miller and Droge (1986), and Miller, Kets de
Vries, and Toulouse (1982) investigated the relationship between leader-
ship style or personality and strategy and organizational structure. They
consider three personality dimensions: flexibility, the locus of control, and
the need for achievement. First, flexibility refers to the adaptability of a
person's thinking and social behavior. Leaders that are flexible are
informal, adventurous, confident, assertive, and egoistic. Inflexible leaders
are deliberate, cautious, industrious, mannerly, guarded, methodological,
and rigid. Second, locus of control is measured on an internal-external
scale. An internal manager believes that the consequences of his behavior
90 Chapter 8. Management and Leadership Style

stem from his own efforts, while an external manager sees the events of his
life as beyond his or her control. Third, managers or leaders with a high
need for achievement prefer to work on a problem rather than leave it to
others; they have a strong preference for structural work situations with
quick and concrete feedback. High-need-to-achieve managers are striving
to meet standards of excellence.
Miller and Toulouse (1986) encountered two qualifying conditions (Table
3.1). The first one is that the relationship between leadership style and
personality is strongest when the organization is small. The second is that
the leadership style has more effect on the strategy and structure when the
organization faces a dynamic, unpredictable, and changing environment
than when it operates in a stable and simple environment. Their empirical
study of ninety-seven Canadian firms generally supports their hypotheses.
Miller and Droge (1986) investigated the relationship between need for
achievement and structural dimensions of centralization, formalization,
and complexity using the same sample of firms as Miller and Toulouse
(1986). They particularly investigated the correlation related to the age and
size ofthe organization. They found that leadership style is more important
in small, young organizations than in large, old organizations. Miller (1991)
extended the analysis to include the tenure of the CEO. He found that the
longer the tenure, the less likely it was that the CEO and the organization
would adapt to changes and uncertainties in the environment. He also
found that such a mismatch would decrease organizational performance.
The various empirical studies by Miller and his colleagues use different
typologies of leadership style and personality. The effect and relationship
are not entirely clear (Miller and Toulouse, 1986). They generally found
that flexibility is an important measure, but they also found that the
relationship between flexibility, delegation, and formalization becomes
insignificant when controlled for need for achievement and locus of control.
They argue that this is due to the significant negative relationship between
flexibility and the need to achieve. The need to achieve is a more important
predictor of appropriate structural dimensions, but Miller also uses the
locus of control in earlier studies (Miller, Kets de Vries, and Toulouse,
1982). Referring to Table 3.1, the flexible personality is at one end of a
scale, and high need for achievement is on the other, with internallocuses
of control in between. This is also consistent with the significant negative
relationship between flexibility and the need to achieve that Miller and
Toulouse found. The decision-making attitudes of a flexible leader show
that he or she is less involved in the decision making than a leader with a
high need for achievement.
This discussion suggests that a one-dimensional typology related to
decision making and information processing could summarize the various
typologies discussed above: microinvolvement. The flexible leader has a low
8.2. Literature Review 91

preference for microinvolvement, while the high-need-for-achievement

leader has a high need for microinvolvement, with the intemallocus of
control leader being in the middle.

Table 3.1. Expected Relationships Between CEO Personality and Organi-

zationa

Flexible Internal Locus High Need for

Personality of Control Achievement

Strategy Niche-focused Innovative Broad aggressive

marketing

Decision Intuitive Informal Analytical

making Short-time Long-term Long-term
horizon planning
Reactive Proactive Proactive
Risk taking Risk neutral Risk aversion

Structure Informal Informal Formal

Unspecialized Mixed Specialized
Much delega- Much delegation Little delegation
tion of authority of authority of authority
Few controls Mixed Many controls
Few liaison Mixed Many liaison
devices devices

Performance Successful in Successful in Successful in large

small firms any size but firms and stable
and dynamic especially so in environments
environments dynamic
environments

Facilitating Conditions: We expect that the relationships between CEO personality

and organizational variables will be higher in (a) small organizations, and (b)
dynamic environments.

"The attributes listed are for high scores on the variables in question. For example,
the higher the score on CEO flexibility, the more will strategies be focused, decision
making intuitive, and so on.

Source: Miller and Toulouse (1986, p. 1392).

The importance of the relationship between leadership style and organiza-

tion and strategy is quite clear. The problematic issue is to define a
typology that is appropriate for establishing this relationship.
92 Chapter 3. Management and Leadership Style

3.3 Management and Leadership Style as

a Contingency

3.3.1 Definition of Management and Leadership

Style: Microinvolvement
Our review of the various typologies on leadership behavior shows that
leadership style can be described by how the leader makes decisions,
handles information, builds relationships with other people, and motivates
and controls subordinates. These dimensions determine whether the leader
is an autocratic or democratic leader, a Theory X or Theory Y leader, or a
flexible or need-for-achievement leader. The different typologies look very
much the same on most of the above-mentioned dimensions. As discussed
in the previous section, it is involvement in information processing and
decision making that matters.
In Table 3.2 we have summarized how the various typologies score on
the dimensions that are important with respect to organizational structure.
Each management or leadership style is assessed on the following
information-processing and decision-making dimensions:

• Preference for delegation,

• Level of detail in decision making,
• Reactive or proactive decision making,
• Decision-making time horizon,
• Risk preference, and,
• Motivation and control.

Beginning on the left, McGregor's Theory X and Theory Y model is

contrasted, (McGregor 1969). A Theory X manager prefers not to delegate
and is formal, detailed, and reactive in decision making. He or she has a
short-time horizon, is risk averse, is not motivating, and has many controls.
A Theory Y manager prefers to delegate, and is general, proactive, and
prefers to be a longer-term decision maker. He or she has high risk prefer-
ence, prefers to motivate, and has few controls. Similarly, Zalesnik and
Kotter's managers and leaders model demonstrates the contrasting styles
(Zalesnik, 1977;Kotter 1990). Likert's (1967) autocratic and democratic
model is similarly contrasting. Finally, Miller, Kets de Vries, and Tou-
louse's three categories of high need for achievement, internal locus of
control, and flexible personality follow the same pattern (Miller et al1982;
Miller and Toulouse,1986).
8.8. Management and Leadership Style as a Contingency 93

Table 3.2a. Management Preference According to Different Typologies

Theory X TheoryY Managers Leaders

Preference for Low High Low High

delegation

Level of detail Formal General Very detailed Low

in decision detailed directions
IIlll.king

Reactive or Reactive Proactive Reactive Proactive

proactive deci-
sion making

Decision- Short term Long term Short, related Future visions

making time to plans
horizon

Risk prefer- Risk High Risk averse, Risk taking

ence averse planning related to
oriented future visions

Motivation Little motiva- Motivation by Monitoring High motiva-

and control tion, IIlll.ny fewer details tion and in-
controls controls spiration

Consider the similarity among Theory X, managers, autocratic, and high

need for achievement. Then consider Theory Y, leaders, democratic, and
flexible personality. On the six dimensions that are particularly related to
organizational design issues, the similarity is striking.
The typologies ofMcGregor, Zalesnik, and Likert fit together well on the
six dimensions. The three categories investigated by Miller fit less well. The
correspondence matches on the dimensions of delegation, level of detail,
motivation and control and risk taking. They differ, however, on the
proactive and reactive and time-horizon dimensions.
Froman intuitive point ofview the four dimensions on which they agree
are closely related to the amount and complexity of information process-
ing. The time-horizon perspective is more difficult. Depending on the
situation, a long-term or short-time horizon may demand complex
information processing, but the quantity may be lower. In our model, we
put a lower weight on this dimension than on the four where all agree.
There is also disagreement in the reactive and proactive dimension. In most
cases a reactive style reflects an inappropriate leadership behavior. In
94 Chapter 3. Management and Leadership Style

particular, we discount the reactive decision-making perspective for the

flexible personality as suggested in table 3.2b. We therefore treat this
dimension with a low weight and use it to warn against inappropriate
leadership behavior.

Table 3.2b. Management Preference According to Different Typologies

Autocratic Democratic High need Intemallo- Flexible

for achieve- cus of eon- personality
ment trol

Preference Low High Low Some High

for
delegation

Level of High Low Analytic and High task Relatively

detail in high level oriented low
decision
making

Reactive or Proactive Proactive Reactive

proactive
decision
making

Decision- Usually Usually lon- Long-term Long-term Short-time

making time short ger planning planning
horizon

Riskpref- Risk averse High Averse Neutral to High

erence averse

Motivation Controls Motivation Many con- Mixed moti- Few

and control troIs vation and controls
control

It is now important to be able to determine whether a manager or leader

has a low preference for microinvolvement or a high preference for
microinvolvement. Our measure has been constructed such that we can
base our relationships on the above dimensions from the literature. The
scale has worked well in our validation procedure for Organizational
Consultant with managers and students.
High preference for microinvolvement and low preference for microin-
volvement are two ends on a one-dimensional scale. The scale represents
how leadership roles are performed. We can place the classic management
3.3. Management and Leadership Style as a Contingency 95

styles on this scale according to how many of a manager's preferences fit

with the scores for the end points (Table 3.3). We argue that any leadership
style on this scale may be effectual depending on the situation and the
chosen organizational structure. However, it also may be useful to add a
managerial style that is categorized as ineffectual. Not all leadership styles
are effectual. Moving from one end of the microinvolvement scale to the
other assumes that the various dimensions continue to be in balance. If
that is not the case, the leadership style is very likely to become ineffectual.
These management preferences are stated in a set of Propositions, 3.1
through 3.13, which are easily incorporated into the knowledge base for
Organizational Consultant. The relations are given in Table 3.4, and we
give a summary in Table 3.5.

Table 3.3. Management Preference for Microinvolvement

Preference for Microinvolvement

Low Medium High

TheoryY Theory X

Leader Manager

Democratic Autocratic

Flexible Personality Internal locus of control High need for achievement

Propositions 3.1 through 3.13 follow directly from our earlier discussion. In
Table 3.2, the first row, we find that a low preference for delegation is
consistent with a Theory X manager with an autocratic approach and a
high need for achievement, Proposition 3.1. In the same row, we find that
a high preference for delegation is consistent with Theory Y leaders with a
democratic approach and a flexible personality, Proposition 3.2. Individuals
who do not delegate decision-making authority and want to be involved in
details realize that preference by making decisions themselves. Individuals
who do delegate decisions have a lower preference for microinvolvement.
The manager with a short-time horizon usually is very much in control
of detailed decisions. A long-term planning horizon will not allow the
manager to be in such control of the details. Again, referring to Table 3.2,
the fourth row, this fits the view of being a Theory X manager and an
autocratic person. In contrast, an individual with a longer-term view is less
96 Chapter 3. Management and Leadership Style

involved. A Theory Y leader with a democratic approach is longer-term

oriented. Propositions 3.3 and 3.4 state these relations.

Table 3.4. Management's Preference for Microinvolvement, Propositions

3.1-3.13

3.1. If the manager has a low preference for delegation of decision

authority, then his preference for microinvolvement is high.

3.2. If the manager has a high preference for delegation of decision

authority, then his preference for microinvolvement is low.

3.3. If the manager has a preference for long-term horizon in his decision
making, then his preference for microinvolvement is low.

3.4. If the manager has a preference for a short-time horizon in his

decision making, then his preference for microinvolvement is high.

3.5. If the manager has a preference for very detailed information when he
makes decisions, then his preference for microinvolvement is high.

3.6. If the manager has a preference for very aggregate information when
he makes decision, then his preference for microinvolvement is low.

3.7. If the manager prefers to be reactive in the decision-making process,

then his preference for microinvolvement is high.

3.B. If the manager prefers to be proactive in the decision making process,

then he has a low preference for microinvolvement.

3.9. If the manager is risk averse, then his preference for microinvolvement
is high.

3.10. If the manager has a low level of risk aversion, then his preference for
microinvolvement is low.

3.11. If the manager prefers to motivate the employees, then he has a low
preference for microinvolvement.

3.12. If the manager prefers to control the employees, then he has a high
preference for microinvolvement.
3.3. Management and Leadership Style as a Contingency 97

3.13. If the manager has reactive decision making style and does not have
a well-balanced style on the other dimensions, then he may be an
ineffective leader.

The more involved the decision maker prefers to be in gathering and using
detailed information, the more microinvolvement he wishes to have. The
same is true with respect to level of detail as shown in the second row of
Table 3.2. If the decision maker wants to make decisions on line-by-line
items, he has higher preference for microinvolvement than one who makes
decisions in gross terms. Propositions 3.5 and 3.6 are empirically supported
by the studies by Miller and Toulouse (1986). They also fit the competing
leadership model shown in Figure 3.1.
A reactive manager is often faced with emerging detailed crises and
operational problems, while a proactive manger can engage in more general
problems facing the organization. In the third row of Table 3.2, these
relations are consistent with Propositions 3.7 and 3.8.
A person who is very risk averse does not want to let anything be
determined by chance. He wants to be in control. The risk-aversion
dimensions appear in most typologies either indirectly or directly. Creative
thinking, vision on management, or similar concepts all include some risk
taking. The relations are included in the fifth row of Table 3.2 and are
included as Propositions 3.9 and 3.10.
The distinction between motivation and control is very important and
is summarized in the last row of Table 3.2. Miller and Toulouse's (1986)
empirical work supports these relations, Propositions 3.11 and 3.12.
Finally, an ineffectual leadership style can occur where the dimensions
are not in balance. This is particularly true ifthe leader also has a reactive
leadership style. One example of an ineffectual leadership style is a leader
with a high preference for delegation and high level of detail, using many
controls, and taking high risks in a reactive manner. Any nonbalanced
leadership style may potentially be ineffective. And actually there are many
more ineffectual styles than effectual ones.
The propositions in Table 3.4 form the basis for estimating the
management's preference for microinvolvement.
An individual's style may be at either end of the level of micro-
involvement. Referring to Table 3.5, an individual with a high preference
for delegation, a long-term horizon, aggregate information, general decision
making, a proactive approach, risk taking, and motivation clearly has a low
preference for microinvolvement. At the other extreme, an individual with
a low preference for delegation, a short-time horizon, detailed information,
detailed decision making, a reactive approach, risk aversion, and a desire
for control has a high preference for microinvolvement. Most individuals do
98 Chapter 3. Management and Leadership Style

not fall at either extreme but somewhere in the middle. How are these
mixed responses put together?

Table 3.5. Management Preference for Microinvolvement

A manager with a LOW preference for microinvolvement:

- Prefers to delegate,
- Gives general directions in decision making,
- Proactive,
- Focuses on the long term and is forward looking,
- Can assume high risk, and
- Motivates by inspiration rather than controls.

A manager with a HIGH preference for microinvolvement:

- Prefers not to delegate,

- Gives specific directions in decision making,
- Is reactive,
- Does not focus on the long term and focuses on the past,
- Is risk averse, and
- Controls rather than motivates by inspiration

Organizational Consultant determines a preference for microinvolvement

by combining the responses to the separate questions on delegation, time
horizon, information, decision level, decision attitude, risk, and motivation.
The combining algorithm gives the most weight to the preponderance of
evidence. Iftwo responses are consistent with high microinvolvement, and
one with low, then the level of microinvolvement tends toward high. The
averaging technique is not simply arithmetic; it follows the Mycin rules dis-
cussed in Chapter 1. Each of the above rules is associated with a certainty
factor. The propositions that lead toward a high preference for microin-
volvement give a somewhat higher weights than those that lead to a low
preference for microinvolvement. Additional nonbalanced dimensions will
lead to an ineffectual management style. In Organizational Consultant, a
separate expert system determines the manager's preference for microin-
volvement.
At the Institute of Applied Computer Science, Benny Mortensen was
very involved in making decisions. When the new organization was
struggling to survive, he took risks and had a short-time horizon .. He
delegated a little but in general was very involved in the activities. His
decision making was neither reactive nor proactive and he used a mixture
of motivation and control. Generally, he was closer to a need-for-an-
3.3. Management and Leadership Style as a Contingency 99

achievement manager than a flexible leadership style. Benny Mortensen

has a relatively high preference for microinvolvement. For this situation he
had an appropriate leadership style. However, only a few changes in one or
two of the dimensions would jeopardize the appropriateness.

3.3.2 Management and Leadership Style as a

Contingency Factor
Robbins (1990) and Lewin and Stephens (1994) suggest that management's
preference for control has a great effect on the choice of the organizational
structure. A level of microinvolvement preference argument can explain a
good portion of the variance not explained by the contingency factors: size,
technology, environment, and strategy. Robbins (1990) is, however, less
clear about the prescriptive-normative issue. The same is true for the
studies by Miller and his associates. Using the arguments presented in
Chapter 1, the descriptive studies can be used as a basis for a normative
model if we assume that only the efficient and effective organizations
generally survive. CEOs view themselves as organizational architects
(Howard, 1992). They want to shape and organize their organizations. In
fact, it is the responsibility of the CEO to see that the organization is
appropriately organized to serve its purpose. It is, of course, not surprising
that CEOs try to create an organization that fits their skills and views.
Using the preference for microinvolvement as a proxy for management
style relating to the information processing capacities, we have developed
a series of propositions. The propositions, that we present, relate to
formalization, centralization, and complexity as well as to configuration.
The strength of the propositions may depend on how extreme the position
is (Lewin and Stephens, 1994). The propositions have worked well in our
validation process.
Propositions 3.14 and 3.15 in Table 3.6 relate the manager's preference
for microinvolvement with the organization's complexity and thus its
differentiation; a higher preference for microinvolvement suggests lower
complexity. Proposition 3.14 is supported by Robbins (1990).

Table 3.6. Management's Preference for Microinvolvement Effects on

Organizational Structure, Propositions 3.14-3.20.

3.14. If the manager has a high preference for microinvolvement, then

organizational complexity should be low.
100 Chapter 3. Management and Leadership Style

3.15. If the manager has a low preference for microinvolvement, then

organizational complexity should be high.

3.16. If an advanced information system is installed, Propositions 3.14 and

3.15 have less effect than they would have otherwise.

3.17. If the manager has a high preference for microinvolvement, then

formalization should be high.

3.18. If the manager has a low preference for microinvolvement, then

formalization should be low.

3.19. If the manager has a high preference for microinvolvement, then

centralization should be high.

3.20. If the manager has a low preference for microinvolvement, then

centralization should be low.

With a high level of differentiation, it is more difficult for the manager to

coordinate and control the activities of the organization. The information
processing demands are greater. The result comes from an increase in the
horizontal differentiation as well as from an increase in the vertical
differentiation. A high horizontal differentiation requires the coordination
of specialists, and a high vertical differentiation implies that many middle-
level managers coordinate. Generally, a manager with a high preference for
microinvolvement prefers an organization where the demand for informa-
tion processing does not exceed his capacity. His capacity may, however, be
enhanced by an information processing system or a decision support
system.
This is in line with Robbins (1990), who suggests that the importance of
Proposition 3.14 and 3.15 could be weakened by the introduction of an
advanced information system, which enhances the coordination and control
even in very complex organizations. This leads to Proposition 3.16. It is
supported empirically by Miller and Droge (1986), who in their orthogonal,
exploratory factor analysis found that "the complexity factor simply did not
materialize" in their study of Canadian firms. The propositions are not
supported by the empirical studies by Miller and Toulouse (1986), who
found a significant negative correlation between a flexible personality and
specialization-one component of differentiation. Additionally, they found a
significant positive correlation between specialization and need for
achievement. The control argument leads to a low degree of complexity. The
counterargument relates to the horizontal differentiation where a
3.3. Management and Leadership Style as a Contingency 101

specialization could lead to a higher performance, which would fit the

attributes of a high-need-to-achieve leader.
Propositions 3.17 and 3.18 in Table 3.5 relate a preference for
microinvolvement to the organization's formalization; a higher preference
for microinvolvement suggests greater formalization. A high degree of
formalization provides management with a higher degree of detailed
control. This is particularly true if management is actively involved in
making the rules. Additionally, a high formalization also increases the
information processing capacity within the rule set. These propositions are
supported by Miller and Toulouse (1986), who found a significant negative
correlation between a flexible personality and formalization, and a positive
correlation between need for achievement and formalization. Similarly,
Miller and Droge (1986) found a significant positive correlation between
need for achievement (or high microinvolvement) and formalization.

Table 3.7. Top Management's Preference for Microinvolvement Effects on

Configuration Coordination and Control and Personnel,
Propositions 3.21-3.31

3.21. The relationship between the management's preference for microin-

volvement and organizational structure is stronger in small
organizations than in large ones (see Chapter 5 for a discussion of
size).

3.22. If the organization is small, centralization is not high, formalization

is not high, and management's preference for micro involvement is
high, then control is required and some liaison activities are needed.

3.23. If configuration is functional and management's preference for

control is high, then management has to control both information and
decisions.

3.24. If configuration is functional, formalization is not high, and manage-

ment has a low preference for microinvolvement then coordination
and control have to be obtained via group meetings and liaison
activities.

3.25. If the manager has a high preference for micro involvement, then a
simple configuration is more likely to be a good configuration.

3.26. If the manager has a high preference for microinvolvement, then a

professional bureaucracy is a less likely configuration.
102 Chapter 8. Management and Leadership Style

3.27. If the manager has a high preference for microinvolvement, then a

machine bureaucracy is a likely configuration.

3.28. If the manager has a low preference for microinvolvement, then the ad
hoc configuration is feasible.

3.29. If the manager has a high preference for microinvolvement, then a

matrix configuration is less likely.

Propositions 3.19 and 3.20 relate a manager's preference for microin-

volvement to the organization's centralization; a higher preference for
microinvolvement suggests higher centralization or decision making at the
top. This is perhaps the most obvious pair of propositions. The higher the
preference for microinvolvement, the more centralized the manager wants
the organization to be. This turns out to be very important, and a misfit
here will usually hamper organizational performance. If the organization
is designed to allow a high degree of decentralized decision making, and the
leader has a high preference for microinvolvement, then he or she will tend
to get involved despite the fact that he is not supposed to. The results are
confusion and frustration for all. Miller and Toulouse (1986) and Miller and
Droge (1986) find very strong support for these propositions.
Propositions 3.21 through 3.31 (Table 3.7) add to and give greater
precision to microinvolvement in the organizational structure for particular
situations. Proposition 3.21 suggests that management's microinvolvement
is stronger in smaller organizations. Miller and Toulouse (1986) in their
study find two qualifiers: size and the dynamics of the environment. They
suggest that the relationship between management style and organiza-
tional structure is strongest when the organization is small and the
environment is dynamic. Ifthe organization is small, then management has
a greater power and, therefore, may influence the structure more than if
the organization is large. This is also often confounded with the fact that in
small organizations management is both owner and founder of the
organization. In a small organization the demand for information process-
ing is normally smaller than in large organizations. Therefore, centraliza-
tion can be higher without creating a misfit between the demand and
capacity for information processing.
In a small organization, when there is a conflict about a recommenda-
tion between management's preference and the environment, then manage-
ment's preference wins. We have found that to be generally true in our
validation of the expert system. The certainty factors associated with the
rules derived from the propositions in this chapter are generally larger than
the certainty factors for rules derived from any other propositions, by a
3.3. Management and Leadership Style as a Contingency 103

magnitude of two to three. This means that there should at least be two
contingency factors in conflict with management's preference when
management's preference does not win.
Miller and Toulouse (1986) found that the lower the management
flexibility, the higher the actual level of control. However, the higher the
need for achievement, the lower the actual control; but the coefficient is not
significant. In our validation we obtained a more complex relationship
between actual control and management preferences or style.
Miller and Toulouse (1986) found that leadership style is more
important in small young organizations than in large old ones. This is
reflected in Proposition 3.21, as discussed above. Coordination and control
are important issues. Coordination can be obtained by means of centraliza-
tion, formalization, or incentive systems. Proposition 3.22 states what to do
when neither centralization nor formalization is used in a small organiza-
tion with a particular leadership style. Propositions 3.23 and 3.24 add two
more instances where coordination has to be obtained by special means.
Proposition 3.23 says that if the organization has to be coordinated
across functional specialties by management, then management has to
control both information and decisions if it has a preference for high
microinvolvement. It is not enough to control either ofthe two.
Propositions 3.25 to 3.29 consider how management's preference for
microinvolvement affects the choice of the right configuration. Generally,
a manager with a high preference for microinvolvement will request a
configuration where the manager is in control. In contrast, a configuration
where the manager is not in direct control is not likely to function well if
run by a manager with a high preference for microinvolvement.
The simple structure gives the manager the most control, and that
works well if the conditions for a simple structure are present. Such
conditions will be discussed in subsequent chapters. Ifthe organization is
so large that a simple structure will not function, then a machine bureau-
cracy may serve the preferences of the manager. On the other hand,
structures that require a high degree of delegation can be run successfully
only if the manager has a preference for delegation-a low preference for
microinvolvement.
The general life cycle approach to leadership and organization suggests
a congruency between the leadership style and the configuration. In the
lifecycle view, the organization goes through the entrepreneurial stage, the
collectivity stage, the formalization stage, and the elaboration stage via
growth and decline. In the entrepreneurial stage the organization is
informal and nonbureaucratic. The manager has to be creative and
technically oriented. When the organization moves to the collective stage
due to growth, a few formal systems begin to appear. The manager has to
104 Chapter S. Management and Leadership Style

delegate some decision-making authority. However, the manager is still

involved in many of the details.
The installations and use of rules procedures and control are important
when the organization reaches the formalization stage. Communication
becomes more formal and less frequent. Specialization becomes important.
Top management is less involved in the day-to-day operation but more
concerned about long-term strategic decisions. More decisions are dele-
gated, and coordination and control are keys to success. In the elaboration
stage, the organization matures. The organization has to fight the
bureaucratization to be more innovative, or the organization will slide into
a decline. The organizational structure has to be very elaborate (Daft, 1992,
p. 193). The evolution shows how the organization adapts to the changing
demands for information processing capacity.
The above shows that organizational design is not static but has to be
adapted to the organizational situation. Additionally, the requirements of
the management are not the same in the different stages. This may cause
different kinds of misfits. In many cases management either has to change
preferences or leave the organization.

3.4 Managing the Management

Individual skills as well as attitudes can be adapted. A fit between the
management, employees, and organizational structure can be obtained by
hiring the right people, training them, and having organizational experi-
ence. Organizational changes may require changes in personnel. Changes
also may be obtained through training and education. Most large organiza-
tions have extensive educational activities to ensure that employees have
the right skills to perform well. Human resource management activities
must fit the developments in the organizational structure and its task
requirements. Numerous techniques have been developed for personal
development (Milkovich and Boudreau, 1988).
The organization's incentive system influences individual behavior.
Behavior focuses on activities for which incentives are related. For example,
if the incentive package for the manager is based on growth, managerial
style may drift toward a preference for growth. Management's preferences
may be changed by the incentive system. Generally, management affects
organizational culture, which affects organizational structure (Scott, 1992,
p.137).
In SAS, as discussed in Chapter 2, major restructuring happened in
1975 when a crisis built up and one ofthe board's solutions was to hire a
new CEO. A crisis is often the antecedent to a major restructuring (Miller
and Friesen, 1980: Tushman and Romanelli, 1985). The question then
8.5 SAS: The Management Style ofCarlzon and Stenberg 105

arises about the attributes that the CEO should have. The propositions in
this chapter give some answers to this question by using the following
procedure. Apply the contingency factors excluding management style to
find the best organizational structure. Then use the propositions to find the
management style that would best support the recommended structure.
The logic of "If X, then Y" allows you to conclude that "lfYis false, then X
cannot be true." This allows us to exclude management styles that do not
fit the recommended structure.

3.5 SAS: The Management Style of

Carlzon and Stenberg
As in many other companies, major economic difficulties were the reason
for changes in management style within the airline company SAS.
SAS's problems emerged with the first oil crisis in the early 1970s. While
SAS's management introduced various measures to cut costs without firing
employees, the good results, that everyone had hoped for, failed to appear,
and gradually the entire SAS organization began to understand that more
dramatic changes were needed. In 1980 a major organizational change was
initiated. This change meant that SAS should be divided into three
divisions-corporate management, the airline company SAS, and SAS's
subsidiaries (Carlzon, 1985). Jan Carlzon was hired as the new president
ofSAS.
Jan Carlzon had been president of the Swedish Linjeflyg, where he had
changed deficit into profit in just two years by changing the concept of the
organization. He had gained a reputation of being a man of vision with a
unique ability to realize that vision. A new vision was needed at SAS.
Carlzon was hired in the second year of deficit after seventeen years of
continuous profit; something had to be done. Carlzon turned out to be the
person to do it, and by spring 1981 he was made president of the entire SAS
group (Carlzon, 1985, p. 46).
The new strategy had to be followed up by a drastic change in manage-
ment style. Top managers would no longer make decisions alone, but each
person would be responsible and take part in the day-to-day running of the
company. It was particularly emphasized that employees were to be
informed and not instructed. As a first sign of this changed attitude toward
personnel, Carlzon began informing employees about new policies as soon
as the board of directors approved his strategy outline. All 20,000 employ-
ees received a red booklet called "Now We Start Fighting," which outlined
the main concepts of the new strategy. Through this action, Carlzon showed
that he considered employees to be a resource that should take active part
106 Chapter a. Management and Leadership Style

in creating the changes necessary for SAS's future success. This marked the
beginning of team spirit among SAS employees, which was an important
asset in the implementation of the new strategy. Carlzon himselfexplained
his management style by using the following metaphor: If you give two
employees a block of stone each and ask them to carve a square out of it. If
you tell one of them that his stone is going to become a part of a large castle
whereas you only ask the second person to carve a square, the first person
is by sure going to like his job better than the last. (Lennby, 1990). And this
was Carlzon's idea: To make a clear strategy that everybody would
understand, and to tell everybody how they could contribute to make the
strategy become successful. Because the new strategy focused strongly on
the front-line personnel's role in the customer's perception of service, it was
important for those on the front line to give good service in their daily
contact with the customer (Carlzon called it the "moment of truth"). This
meant that front-line personnel should be able to solve problems as they
occurred without consulting their superiors. This decentralized decision
making made it necessary to delegate responsibility, but if employees were
to assume responsibility, they needed education. As a result of these
educational efforts and the increased focus on front-line personnel,
employees suddenly felt they were appreciated and were willing to lend a
helping hand in the attempt to bring SAS back to its feet. Employ-
ees-colleagues or not-helped each other in order to deliver the best product
to the customer without giving thought to who got credit for the work. At
the presentation, Carlzon talked about the importance of practicing
"Management by Love rather than Management by Fear," meaning that by
showing employees respect and trust they would strain every nerve in order
to show that they were worth the credit.
One of Carlzon's strengths was his ability to motivate employees by
appealing to the idea of the organization's common goal. Carlzon managed
by motivation rather than through control, and this was why employees'
motivation vanished when Carlzon started making cutbacks. In terms of
our description earlier in this chapter, Carlzon had:

• A high preference for delegation

• A general level of detail in his decision making
• A proactive decision style
• A future orientation in his vision and a relatively long-term view and
• A high risk preference

He also promised:

• Motivation through inspiration and not detailed control.

8.5 SAS: The Management Style ofCarlzon and Stenberg 107

He was a leader in leader in ZalesnikIKotter sense, a Theory Y individual

in McGregor's scheme, democratic in Likert's scheme, and a flexible
personality in Miller's categorization.
Using table 3.4 , Propositions 3.2, 3.3, 3.6, 3.8, 3.10 describe Carlzon's
management style well, and they all lead to a low preference for
microinvolvement. When compared to Table 3.6, the new structure of SAS
fit very well with Carlzon's management style as the recommended
organization should have high complexity (proposition 3.15), low formaliza-
tion (proposition 3.18) and low centralization (proposition 3.20).
Jan Stenberg became SAS president and CEO in April 1994. Stenberg's
management style is very different to Carlzon's. Stenberg's main focus was
to bring SAS back on its feet financially. This implied rationalizations of up
to SEK 2.6 billion, include the firings of 3000 employees. These decisions
were obviously taken in the top management group without much
consultation of employees. Stenberg's personality is also very different to
Carlzon's. Contrary to Carlzon who was always in high spirits when he was
in the limelight, Stenberg does not like the media attention, and he only
talks directly to the employees when absolutely necessary.
While Stenberg has managed to get SAS back on its feet and produce
profits unlike ever before (SEK 2365 mill.), many of his decisions have been
taken within the top management group. They have not paid a great deal
of attention to inform the employees how decisions would be implemented.
The general lack of formal information of employees has led to employee
frustration, resulting in strikes, especially among cabin crew and pilots,
threatening to deteriorate SAS's punctuality and reliability and hence the
company's competitiveness.
Although the appointment of Stenberg has increased centralization at
SAS, the centralization has mainly been on strategic issues and not
operational issues-on which employees still have a relatively expanded
decision making authority. Following table 3.2a, Stenberg is therefore more
of a Theory Y than a Theory X individual in McGregor's scheme: Although
not as high as Carlzon's, Stenberg has a relatively high risk preference. He
is future oriented. Stenberg is mainly concerned with general directions in
his decision making, and although he motivates with more control than
Carlzon, he still motivates by "fewer controls" rather than "many". This
makes him more of a leader than a manager in ZalesnikIKotter's sense,
despite the fact that he does not necessarily motivate through inspiration.
Furthermore, Stenberg is relatively democratic in Likert's scheme.
However, the typology that best describes Stenberg's management style is
Miller's internal locus of control. This management style is categorized as
one with some preference for delegation, high task orientation, proactive
decision making, long term time horizon, neutral risk preference and mixed
108 Chapter 8. Management and Leadership Style

motivation and control (table 3.2b). Again, the organizational implications

for this management style can be derived from Table 3.4., 3.6 and 3.7.
From Propositions 3.1 to 3.10 it appears that Stenberg has a low
preference for microinvolvement. Because Stenberg does not emphasize the
motivation of employees, but rather expects them to carry through
decisions that are made in the top management group, propositions 3.11
and 3.12 however lead towards a low to medium preference for
microinvolvement. In Tables 3.6 and 3.7, Propositions 3.15, 3.18, 3.20, 3.28
and 3.29 show that SAS needs:

• High complexity (proposition 3.15)

• Low formalization (proposition 3.18)
• Low centralization (proposition 3.20)
• An ad hoc configuration is feasible (proposition 3.28)
• A matrix configuration is likely (proposition 3.29)

SAS has high complexity on all three dimensions: vertical, horizontal and
spatial differentiation. Due to SAS's technology, formalization is relatively
high. Furthermore SAS's service has been relatively formalized by the
creation of the control and information system containing solutions to
various problems. Formalization should therefore be described as medium
to high. Employees are given responsibility and decision making authority,
although they are now given more directives as for what frames and goals
they are to work within. Centralization can therefore be described as low
to medium. SAS' configuration can best be described as a matrix. An ad hoc
configuration would be inefficient in an organization where the technology
is largely routine. SAS is not totally in accordance with the above proposi-
tions. Misfits are mainly due to the fact that airlines have to operate with
high formalization despite their equivocal, uncertain and complex
environments. As long as operational decisions are kept fairly decentral-
ized, Stenberg's new management style will therefore be consistent with
SAS's overall structure.

3.6 Summary
In this chapter, we considered the fit between management and leadership
style and organizational design. Consistent with the neo-information
processing perspective, executives have preferences for the level of detail
they require in decision making, for a reactive or proactive decision making
style, for risk preference, for a time horizon, for delegation, and for style of
motivation. These leadership dimensions define a style that should fit the
References 109

organization's complexity, formalization, and centralization as well as its

organizational configuration.
A preference for microinvolvement is a summary measure of manage-
ment style. A preference for low microinvolvement is compatible with high
complexity, low formalization, and low centralization. A preference for high
microinvolvement is compatible with low complexity, high formalization,
and high centralization. SAS's Carlzon's and Stenberg's low microinvolve-
ment illustrates the propositions provided.

References
Arrow, Kenneth J. 1974. The Limits of Organization. New York: Norton.
Carlzon, Jan. 1985. Riv pyramiderne ned!. K0benhavn: Gyldendal, Denmark.
Child, John. 1972. "Organization, Structure, Environment and Performance: The
Role of Strategic Choice." Academy of Management Review, 561-575.
Daft, Richard. 1992. Organization Theory and Design (4th ed.). St. Paul, MN: West.
(2nd ed., 1986; 3rd ed., 1989).
Fieldler, F.E. 1977. "What Triggers the Person-Situation Interaction in Leader-
ship?" In D. Magnusson and N.S. Endler (Eds), Personality at the Crossroads:
Current Issues in International Psychology. Hillsdale NJ: Erlbaum.
Howard, Robert. 1992. "The CEO as Organizational Architect: An Interview with
Xerox's Paul Allaire." Harvard Business Review, 107-121.
Hunt, James G. 1991. Leadership: A New Synthesis. Beverly Hills. CA: Sage.
Kotter, J.P. 1988. The Leadership Factor. New York: Free Press.
Kotter, J.P. 1990. A Force for Change: How Leadership Differs from Management.
New York: Free Press.
Lennby, Peter. 1990. "Flyv mod h0jere h0jder." Video, SV Tv, Sweden.
Lewin, Arie, and Carrol V. Stephens. 1994. "CEO Attitudes as Determinants of
Organizational Design: An Integrated Model." Organizational Studies, 15(2),
183-212.
Likert, Renis. 1967. The Human Organization. New York: McGraw-Hill.
Luthans F., R.M. Hodgetts, and S.A. Rosenkrantz. 1988. Real Managers. Cam-
bridge, MA: Ballinger.
McGregor, Douglas. 1969. The Human Side of Enterprise. New York: McGraw-Hill.
Miles, Raymond E., and W.E. Douglas Creed. 1995. "Organizational Forms and
Managerial Philosophies." In L.L. Cummings, and Barry M. Staw. Research in
Organizational Behavior. 17, 333-372.
Milkovich, George T., and John W. Boudreau. 1988. Personnel/Human Resource
Management. Plano, TX: Business Publications.
Miller, Danny. 1991. "Stale in the Saddle: CEO Tenure and the Match Between
Organization and Environment." Management Science. 37(1), 34-52.
Miller, Danny, and Cornelia Droge.. 1986. "Psychological and Traditional Determi-
nants of Structure." Administrative Science Quarterly, 31, 539-560.
Miller, Danny, and P. Friesen. 1980. "Archetypes of Organizational Transition."
Administrative Science Quarterly, 25, 268-299.
110 Chapter 3. Management and Leadership Style

Miller, Danny, M.F.R. Kets de Vries, and Jean-Marie Toulouse. 1982. "Top
Executive Locus of Control and Its Relation to Strategy-Making, Structure and
Environment." Academy of Management Journal, 25(2), 237-253.
Miller, Danny, and Jean-Marie Toulouse. 1986. "Chief Executive Personality and
Corporate Strategy and Structure in Small Firms." Management Science, 32(11),
1389-1409.
Mintzberg, Henry. 1980. The Nature of Managerial Work. Englewood Cliffs, NJ:
Prentice-Hall (1st ed., 1973).
Nordhagen, P. 1993. "Mellemlanding." Jyllandsposten, 21111.
Norman, Richard. 1983. Service Management. Malmo: Liber Forlag.
Quinn, R.E., and J. Rohrbaugh. 1983. "A Spatial Model of Effectiveness Criteria:
Towards a Competing Values Approach to Organizational Analysis." Manage-
ment Science, 29(3), 363-377.
Robbins, Stephen P. 1990. Organization Theory, Structure, Design, and Application.
Englewood Cliffs, NJ: Prentice-Hall.
SAS. 1981-1982. SAS Annual Reports. Stockholm: SAS.
SAS. 1994. SAS Annual Reports. Stockholm: SAS.
SAS. 1995. SAS Annual Report. Stockholm: SAS.
SAS. 1996. Milk~regnskab. Stockholm: SAS.
SAS. 1996. Annual Report. Stockholm: SAS.
SAS. 1996. De f~rste 50 ar. Stockholm: SAS.
SAS. 1996. Scandinavian Airlines 1946-1996. Stockholm: SAS.
SAS. 1997. Delilrsrapport 1: 1997. Stockholm: SAS.
Schermerhorn, J.R. Jr., J.G. Hunt, and R.N. Osborn. 1991. Managing Organiza-
tional Behavior (3rd ed.). New York: Wiley.
Scott, W. Richard. 1992. Organizations: Rational, Natural and Open Systems (3rd
ed.). Englewood Cliffs, NJ: Prentice-Hall.
Tushman, M.L., and Elaine Romanelli. 1985. "Organizational Evolution: A
Metamorphosis Model of Convergence and Reorientation." In L.L. Cunnings and
B.M. Stacer (Eds.) Research in Organizational Behavior (pp. 171-222). San
Francisco: SAl Press.
Vroom, Victor H., and Philip H. Yetton. 1973. Leadership and Decision-Making.
Pittsburgh: University ofPjttsburgh Press.
YukI, G.A. 1981. Leadership in Organizations (2nd ed.). Englewood Cliffs, NJ:
Prentice-Hall.
Zalesnik, A. 1977. "Managers and Leaders: Are They Different?" Harvard Business
Review, 15,67-68.
CHAPrER4

Organizational Climate

4.1 Introduction
"Navy retirees recruited by tech firms fit discipline, flexibility to new
tasks." This was one of the headlines in the San Jose Mercury News,
Sunday, February 15, 1998. The story tells how Silicon Valley corporations
have started hiring retired and former military personnel. The story focuses
on the similarities and differences in working in the military with its chain-
of-command hierarchy and being in a seemingly less structured environ-
ment. "Military personnel are veterans ofthe teamwork that is regarded as
an essential part of valley life. But the military requires clarity at every
level about the task at hand to be critical for effective performance. What
these newcomers to high tech often see are missed opportunities for
nurturing a common sense of mission." "There's more freedom, more
latitude to make things happen in the high tech companies," says Jack Gale
a former Navy commander, but like in the Navy it is all about performance.
The story also stresses the problems and transitions the former military
personnel had to go through to fit to their new positions but also what new
ways oflooking at things they bring in.
The important message is that people, their way of looking at things,
their perceptions and values are indeed important. Applied Komatsu
manager Mayekawa says companies hiring military personnel try to gauge
their ability to operate in a civilian environment by asking how they would
handle a hypothetical situation with a customer. ''Ifthey are too structured,
the issue of culture comes into play," he says.
Tosi (1992) argues that the scope of contingency theory should be
broadened to include culture, organizational culture, and individual and
group variables. So do the organizational climate and culture affect
performance and are there issues offit related to the appropriate choice of
the organizational structure-its configuration, properties, and perfor-
mance? In the literature the discussion has caused a good deal of contro-
versy. First, it is not clear, what organizational culture and climate mean.
Second, the difference between organizational climate and organizational
culture is often blurred. Third, the relationship to performance is
somewhat ambiguous.
112 Chapter 4. Organizational Climate

In this chapter we will argue that a climate measure is a complement

factor to obtain an efficient and effective organization. The organization's
climate, which captures how individuals feel about the organization, is a
determinant ofthe structure. The way individuals feel affects their ability
to process information and thus is an important measure in the determina-
tion of the information processing capacity of the organization. Therefore,
the climate affects the choice of structural properties. As is the case with
other contingency factors, fit is the important issue-cause and effect are
often difficult to determine. Additionally, there are important fit/misfit
issues between the organization's climate and the other contingency factors
in the multidimensional contingency model. These will be dealt with in
Chapter 9.
We review the concepts of climate and culture. We then examine
climate and develop an approach to measuring climate for an organization.
The competing values approach is a means to summarize and categorize
climate. We then develop a number of relations between the four climate
categories and the organizational design, i.e., the implications of the
organization's climate for its design. Finally, we discuss how the organiza-
tional climate may be changed.

4.2 Climate and Culture

Organizational culture and climate are often used in the literature to
describe similar issues. Denison (1996) asks in his survey paper: "what is
the difference between organizational culture and organizational climate?"
He concludes that "on the surface, the distinction between organizational
climate and organizational culture may appear quite clear: climate refers
to a situation and its link to thoughts, feelings, and behavior of organiza-
tional members", while "culture, in contrast, refers to an evolved context
(within which a situation may be embedded)." He is then led to the
conclusion that climate and culture research "should be viewed as
differences in interpretation rather than differences in the phenomenon."
Although organizational climate and culture are often used interchange-
ably, they have different roots. Let us begin with some definitions
(Webster's Ninth New Collegiate Dictionary):

Culture: An integrated pattern of human knowledge, belief and behavior

that depend upon man's capacity for learning and transmitting
knowledge to succeeding generations, the customary beliefs,
social forms and material traits of a racial, religious or social
group.
4.2 Climate and Culture 113

Climate: The prevailing influence or environmental conditions characteriz-

ing a group or period, atmosphere, the prevailing set of conditions.

Given that climate and culture are natural concepts, these definitions are
a good point of departure. The culture is a pattern of knowledge, belief and
behavior that emerge including social forms. In the context of the organiza-
tion social forms and knowledge in general include the organizational
structure, but there has been little agreement about the use of the terms
in the organizational context.
Organizational climate has been defined as the "relatively enduring
quality of the internal environment of an organization that a) is experi-
enced by its members, b) influences their behavior, and c) can be described
in terms of the values of a particular set of characteristics (or attitudes) of
the organization." (Taguiri and Litwin, 1968, p. 27). The climate is the
"ether" within which an organization exists. "In an overall organizational
model climate can be seen as an intervening variable in the process
between input and output and one that has a modifying effect on this
process. Climate effects organizational and psychological processes, and
thus acquires an influence over the results of organizational operations"
(Ekvall, 1987).
The organizational culture is the organization itselLthe form, beliefs,
norms, social patterns, the way things are done, the symbols, rituals, etc.
Schein (1992) defined culture as: "A pattern of shared basic assumptions
that the group learned as it solved its problems of external adaptation and
internal integration, that has worked well enough to be considered valid
and, therefore, to be taught to new members as the correct way to perceive,
think, and feel in relation to those problems." Culture is thus derived and
observed in the emergent behavior of the group or the social form or
organization. It is learned and transmitted from one generation to the next.
The organizational culture is bound up with the form and properties of the
organization itself Patterns of behavior emerge from the organizational
form and the organizational properties themselves. The primary reason
management creates and changes the form and properties are the influence
the "pattern of behavior" has in the organization. It seems that the culture
is integrated into the organizational design and the organizational design
levers are also organizational culture levers.
The culture emerges from the form and properties as the organizational
form is implemented. Perhaps, the most explicit new organization element
discussed as parts of the culture are the rituals, rites and symbols of
management and these should not be overlooked. They are part of the
organizational structure properties as they can be seen as means of
coordinating the organizational activities.
114 Chapter 4. Organizational Climate

From a design view, one would like to argue that some cultures are more
efficient or more effective than others. Many popular books prescribe how
to create an organizational culture that will enhance performance. Siehl
and Martin (1990) however, argue that the evidence is very weak in the
normal positive science sense: "... culture-financial performance proposition
consists of variants of a contingency argument. The most common variant
is a claim that firms with cultures congruent with their business strategies
are better performers than firms that lack this congruency. When a firm's
culture clashes with its strategy, confusion and conflicts of interest
increase, strategies are resisted, and the firm's financial performance
ultimately may be impaired. In contrast, when the culture and strategy are
synchronized, such difficulties, are said to be reduced, with a concomitant
beneficial effect on financial performance. Evidence supporting this has,
however, been confined, for the most part to short, almost anecdotal,
descriptions of case studies of single organizations." (Siehl and Martin,
1990 p. 263). They also state that "when researchers look at organizations
"as if' they were cultures, they can enrich our understanding of organiza-
tionallife in new and unexpected ways. Already, interest in cultures has
opened and reopened neglected areas of inquiry - for example, by exploring
the organizational relevance of symbolism, structuration, semiotics,
deconstruction, and ideology." "These studies bring us far beyond the
traditional variables of organizational theory, such as structure, firm size,
technology, job satisfaction, motivation and leadership," (Siehl and Martin,
1990, p. 272).
In the discussion of organizational culture and climate so far little was
said about the actual relationship between dimensions of culture, climate,
structural design and performance. Some studies relate directly the various
dimensions to, performance while others try to develop culture and climate
typologies (Poole and McPhee, 1983) from which a relationship to perfor-
mance can be established.
One of the successful typologies that is heavily researched is the
competing values approach, which we introduced in Chapter 3, Figure 3.1
(Quinn and Rohrbough, 1983). The competing values approach uses four
categories to describe four cultures of the organization:

• The group culture (a friendly place to work where people shares much
themselves).
• The developmental culture (dynamic, entrepreneurial and creative place
to work).
• The internal process culture (formalized and structured place to work).
• The rational goal culture (results-oriented where leaders are hard
drivers, producers and competitors).
4.2 Climate and Culture 115

Each culture describes an ideal model and may not be found in its pure
form in any organization. Rather, each real world organization will display
some aspects of each culture model with greater weight for one model than
another, i.e., an organization is more like a group culture than a rational
goal culture, but there are aspects of both. Additionally, different parts of
the organization may be more like one type while other parts of the
organization may be more like another type.
The competing values framework captures two basic dimensions of the
organization (control-flexibility and internal-external). On the control-
flexibility axis, the group and developmental cultures are more flexible and
the internal process and rational cultures are more control oriented. On the
orthogonal axis, the group and the internal process are more internally
oriented where the developmental and rational cultures are more external
oriented.
The competing values framework is a very versatile typology for
capturing the complexity of a variety of management issues. The applica-
tions include strategy (Bluedorn and Lundgren, 1993), human resource
policies (Yeung, et aI, 1991, Giek and Lees, 1993), organizational change
(Hooijberg and Petrock, 1993), management information systems, (Cooper
and Quinn, 1993) as well as culture (Denison, 1990, Cameron and Freeman,
1991), and climate (Zammuto and Krakower, 1991). Thus, the competing
values approach enables us to sort out the relationship between culture and
climate.
The initial development by Quinn and Rohrbaugh (1983) examined
organizational effectiveness criteria. The competing values approach has
been robust across these many applications and reliable as a measurement
instrument.

Flexibility

Group culture Developmental culture

Internal External
Internal process culture Rational goal culture

Control

Figure 4.1. The Competing Values Model

116 Chapter 4. Organizational Climate

Poole (1985, p. 84) states that climate seems to be a feature of, rather then
a substitute for, culture. The same view was adopted by Zammuto and
Krakower (1991) that related climate to the four categories in the
competing values approach. They use organizational characteristics
(centralization, formalization, long-term planning), climate measures
(trust, conflict, morale, equity of rewards, resistance to change, leader
credibility, and scapegoating) and strategy dimensions (reactive/pro-active
orientation) to categorize the culture into group culture, developmental
culture, internal process culture, and rational culture. They concluded that
cultural type is related to differences in organizational climate. For each of
these four types of culture their study allows us to develop four types of
climates that correspond to the four types of culture and thus allow us to
obtain relationships between climate dimensions and organizational
characteristics.
Quinn and Spreitzer (1991) analyzed in two studies two different
instruments to describe the competing values categories. Both instruments
seem to capture the categories well. The implication, with some caution, is
that it is possible to combine results from different studies using different
instruments to describe the competing values typology. The categorizations
seem to be rather robust. Using this reasoning the competing values
approach provides us with relationships between strategy, environment,
technology, leadership, organizational form, and psychological climate. The
competing values approach specifically allows us to sort out the various
dimensions of organizational climate and culture. From an information
processing point of view we are interested in including those dimensions
that effect the information processing capacity of the members of the
organization.
The many empirical studies using the competing values approach give
us a rich set of knowledge to fill in the multiple contingency model where
the psychological climate is introduced on the left-hand side in Figure 1.3
in Chapter 1.
In brief, the general support that the culture is an important factor for
efficient organizations is mostly case and anecdotal evidence, not the
normal positive science studies. However interpreted more broadly culture
may be considered as looking at organizations as particular sets of
situational and structural patterns which are consistent-provide a total fit.
This is called the "configurational perspective" by Delery and Dotty(1996).
In the context of culture, the competing values approach provides such ideal
patterns including climate as one ofthe dimensions.
The climate and culture are thus important to consider in the design of
the organization. The relationship between climate and performance is
linked via behavior. However, the way culture is commonly defined and
used it includes both dimensions related to how the individuals in the
4.3 Literature Review on Climate 117

organization process information as well as more structural dimensions. In

this way culture is both the means and the ends in the design process.
Particular culture typologies thus implicitly assume that a proper fit has
been obtained. We are searching for a fit with the way individuals in the
organization process information and the way the organization should be
structured.
Introducing climate and culture opens and enriches the managerial
framework for a deeper and more comprehensive understanding of design
choices. Culture is part of any organization, whether designed or not.
However, organizational culture mixes organizational properties with the
behavior of the individuals. The organizational climate is the part of the
culture that is more directly related to the general behavior of the
individuals in the organization and thus a candidate to be included in the
multiple contingency model.
The correct fit between behavior ofthe members ofthe organization and
the organizational structure is the key to improved performance. From a
design point of view it is important to develop a set of climate dimensions
that captures the individuals' perception of the work environment and its
effect on the individuals' performance and thus on the performance of the
organization.
Again, the individuals' capacity to process information is an important
element in accessing the information processing capacity of the organiza-
tion.

4.3 Literature Review on Climate

Climate was defined in the dictionary as the prevailing influence on
environmental conditions characterizing a group or period, atmosphere, the
prevailing set of conditions.
Even within the more narrow set ofliterature on climate there has been
some confusion about the relevance and definition of the concept and the
relationship with organization structure and organization culture (James
and Jones, 1974, Schneider, 1990, and Denison, 1996). This review does not
pretend to survey all the research within the climate area, but will try to
outline different views and research directions that will allow us to
integrate a measure of climate into the multi dimensional contingency
model.
One reason for the confusion in the literature can be found in use of
climate to represent seemingly different concepts. First, climate can be seen
as organizational climate or psychological climate. Ekvall (1987) states that
the organizational climate arises in the confrontation between individuals
and the organizational situation. According to James and Jones (1974)
118 Chapter 4. Organizational Climate

organizational climate can also be viewed in two different ways: "a

multiple-measurement-organizational attribute approach" or "a perceptual
measurements-organizational attribute approach."
Both these approaches are confounded with organizational structure and
processes and the general organization situation. The organizational
climate is measured using variables like individual autonomy, the degree
of si~ructure imposed as the positions, reward orientation, consideration,
warmth, and support. This is also the case in the treatment of organiza-
tional climate dimensions presented in Litwin and Stringer (1968) where
organizational climate is measured along the following dimensions:
structure, responsibility, warmth, support, reward, conflict, standards,
identity, and risk.
It is obvious from the above that measures and dimensions of organiza-
tional climate and organizational culture can be confused. This has been
discussed often in the literature (Schneider, 1990 and Denison, 1996).
Denison concludes that although the two concepts on the surface look very
different at a deeper level, the clear distinctions begin to disappear.
In contrast the definition of the psychological climate seems more strict
in the sense it refers to perceptions held by the individuals about the work
situation. James and Jones (1974) summarize the psychological climate to
be a set of summary or global perceptions held by individuals about their
organizational environment. The psychological climate is a summary
evaluation of actual events based upon the interaction between actual
events and the perception of those events. The psychological climate has
been measured using dimensions such as disengagement, hindrance, esprit,
intimacy, aloofness, production emphasis, trust and consideration.
Koys and DeCotiis (1991) define the psychological climate as "an ex-
perimental-based, multidimensional, and enduring perceptional phenome-
non which is widely shared by the members of a given organizational unit."
They continue to state that the psychological climate is the description-and
not the evaluation-of experience. As such, the psychological climate is
different from, e.g. job satisfaction. In their survey Koys and DeCotiis
report more than 80 different dimensions found in the literature which has
been labelled a climate dimension.
They set out to find a theoretical-meaningful and analytical-practical
universe of all possible climate dimensions. They established three rules for
a dimension to be included in the universe:

• Has to be a measure of perception

• Has to be a measure describing ( not evaluating)
• Must not be an aspect of organizational or task structure
4.3 Literature Review on Climate 119

These rules attempt to sort out the confusion and also distinguish the
measures from the organizational climate and culture measures. The rules
make sure that the psychological climate measure is not confounded with
the organization's structural properties.
Applying these rules to the more than 80 dimensions used in climate
measurements and combining dimensions that were actual the same
despite the fact they had different names reduced the number of dimen-
sions to 45. Koys and Decotiis were then able to categorize these 45
dimensions into the 8 summary dimensions shown in Table 4.1.

Table 4.1. Definition of Each of the Eight Dimensions of the Universe of

Psychological Climate Perceptions (Koys and DeCotiis, 1991).

Dimension Definition
name
Autonomy The perception of self-determination with respect to work
procedures, goals, and priorities.
Cohesion The perception of togetherness of sharing within the organiza-
tion setting, including the willingness to members to provide
material aid.
Trust The perception offreedom to communicate openly with members
at higher organizational levels about sensitive or personal issues
with the expectation that the integrity of such communications
will not be violated.
Pressure The perception of time demands with respect to task completion
and performance standards.
Support The perception of the tolerance of member behavior by superiors,
including the willingness to let members learn from their
mistakes without fear of reprisal.
Recognition The perception that member contributions to the organization
are acknowledged.
Fairness The perception that organizational practices are equitable and
nonarbitrary or capricious.
Innovation The perception that change and creativity are encouraged,
including risk-taking into new areas or areas where the member
has little or no prior experience.

Koys and DeCotiis (1991) tested the validity and reliability of their
summary scales and found that they were both valid and reliable.
They discuss, however, whether some of the dimensions should or could
be combined into a single dimension to further reduce the dimension of
psychological climate. They suggest that trust and support may be
combined bringing the dimension of psychological climate down to seven.
120 Chapter 4. Organizational Climate

The dimensions of climate developed by Koys and DeCotiis fit the notion
by Rousseau (1988) where she says that "climate is a content-free concept,
denoting in a sense generic perceptions of the context in which an
individual behaves and responds."
From these dimensions it is seen that the concept of psychological
climate does not interfere with the right-hand side of the multi-dimensional
contingency model (Figure 1.3) and thus more readily can be candidates for
measuring climate in a contingency model for organizational design. As will
be argued later such dimensions can also easily be related to the informa-
tion processing of the members of the organization.
The research on psychological climate has tried to relate the measure to
other concepts such as job satisfaction, job attitude etc. The important issue
is whether the psychological climate influences the behavior of the
individuals and the organization and if there exist interaction effects with
contingency factors and/or structural factors of the organization. Ekvall
(1987) concludes his paper by stating that "In an overall organizational
model climate can be seen as an intervening variable in the process
between input and output, and one that has a modifying effect on this
process. Climate affects organizational and psychological processes, and
thus acquires an influence over the results of organizational operations."
The conclusion ofthis survey on climate in an organizational context is
that seven dimensions exist that measure the individuals' perception of the
organization's psychological climate. As such they should be a basis for a
definition of climate in a multi-contingency model for organizational design.

4.5 Measuring and Categorizing Climate

The psychological climate refers to the beliefs and attitudes held by
individuals about their organization. The climate is an enduring quality of
an organization that (1) is experienced by employees, and (2) influences
their behavior. It should be thought of and measured "at the organizational
level of analysis." (Glick, 1985, p. 607). Climate is an organizational
characteristic-not a characteristic of each individual in an organization.
However, we do look to individuals as the source of information on the
climate. It is their perception about the organization that we measure.
Zammuto and Krackover(1991) measured climate using the following
dimensions:

• Trust
An organization has a high level of trust when the individuals are open,
sharing and truthful, where individuals place their confidence. An
organization has a low level of trust when the individuals are closed,
4.5 Measuring and Categorizing Climate 121

guarded, unsharing, untruthful, and creates an atmosphere of anxiety and

insecurity.

• Conflict
An organization has a high level of conflict when there is a high opposition
of forces, goals and beliefs, which are experienced in friction and disagree-
ment among the individuals. An organization has a low level of conflict
when there is harmony in goals, beliefs, which yields a spirit of cooperation
among the individuals.

• Morale
An organization has a high level of employee morale when the individuals
are confident and enthusiastic about the organization-an Esprit de Corps.
An organization has a low level of employee morale when the individuals
lack confidence and enthusiasm about the organization and individuals lack
a sense of purpose and confidence about the future.

• Rewards
An organization is equitable in its rewards when individuals accept
rewards as fair and just without bias or favorism. An organization is
inequitable in its rewards when individuals see favorism, bias, and non
work related criteria as the basis for rewards.

• Resistance to change
An organization has a high resistance to change when individuals believe
the inertia is high and presume and desire that "we will do things tomorrow
as we did them today." An organization has a low resistance to change
when individuals embrace change as the normal circumstance and relish
that "tomorrow will be different."

• Leader credibility
The leader credibility is high when individuals have beliefin its leadership;
there is a sense of respect, inspiration and acceptance of decisions and
actions. The leader credibility is low when the individuals lack respect and
do not accept the legitimacy of authority.

• Scapegoating
An organization has a high level of scapegoating when individuals believe
that the responsibility for actions will be shifted to others - top manage-
ment, staff, employees, or outsiders. An organization has a low level of
scapegoating when individuals believe that the resposible individuals
assume the responsibility for the failure of actions.
122 Chapter 4. Organizational Climate

It is interesting that Zammuto and Krackover (1991) and Koys and

DeCotiis (1991) both define climate using 7 relatively similar dimensions.
Table 4.2 shows the two sets of climate dimensions.

Table 4.2. Dimension of Climate

Koys and DeCotiis (1991) Zammuto and Krackover (1991)

Autonomy Credibility
Cohesion Conflict
Trust/support Trust
Pressure Scapegoating
Recognition Morale
Fairness Equitable Rewards
Innovation Resistance to change

The dimensions do not fit completely together one-by-one, but they are very
similar and the totality of the seven dimensions are indeed very similar.

Table 4.3. Comparing the results by Zammuto and Krackover (1991) and
Burton et al (1998).8

Group Develop- Rational Internal

mental goal process

Trust H H M H L L L L

Conflict L L L L H H H HM

Morale MH H MIH H M L L HM

Rewards H H M H L L L L
equitability

Resistance to M H L L M L H H
change

Leader credibility H H H HM ML L L L

Scapegoating L L MIL L H H H MH

8The results by Burton et al (1998) are in italics. The scores are high(H), me-
dium/high(MH), medium(M), medium/low(ML), and Low(L).

Zammuto and Krackover (1991) mapped their seven dimensions into the
competing values framework and in that way created four different climate
types which could be labelled:
4.5 Measuring and Categorizing Climate 123

• The Group Climate

• The Developmental Climate
• The Rational Goal Climate
• The Internal Process Climate

Burton et al (1998) in a study of 246 Danish firms found that the seven
dimensions by Zammuto and Krackover (1991) could be used to categorize
the climate. Using a cluster analysis they found that the seven dimensions
could describe 4 climate types very similar to the types found in the study
by Zammuto and Krackover (1991).
Table 4.3 shows that the two studies resulted in very similar results.
There are only marginal differences, and the general spirit of the four
climate groups is identical.
Hooijberg and Petrock(1993) characterize the four corresponding climate
types from the point of view of the competing values: "The group climate
could be described as a friendly place to work where people share a lot of
themselves. It is like an extended family. The leaders, or head of the
organization, are considered to be mentors and, perhaps even parent figures.
The organization is held together by loyalty or tradition. Commitment is
high. The organization emphasizes the long-term benefits ofhuman resource
development with high cohesion and morale being important. Success is
defined in terms of sensitivity to customers and concern for people. The
organization places a premium on teamwork, participation, and consensus.
The developmental climate could be described as a dynamic, entrepre-
neurial and creative place to work. People stick their necks out and take
risks. The leaders are considered to be innovators and risk takers. The glue
that holds organizations together is commitment to experimentation and
innovation. The emphasis is on being on the leading edge. Readiness for
change and meeting new challenges are important. The organization's
long-term emphasis is on growth and acquiring new resources. Success
means having unique and new products or services and being a product or
service leader is important. The organization encourages individual
initiative and freedom.
The rational goal climate could be described as a results-oriented
organization. The leaders are hard drivers, producers, and competitors.
They are tough and demanding. The glue that holds the organization
together is the emphasis on winning. The long-term concern is on competi-
tive actions and achievement of measurable goals and targets. Success is
defined in terms of market share and penetration. Competitive pricing and
market leadership are important. The organizational style is hard driving
competitiveness.
The internal process climate is a formalized and structured place to work.
Procedures govern what people do. The leaders pride themselves on being
124 Chapter 4. Organizational Climate

coordinators and organizers. Maintaining a smooth running organization

is important. The long term concerns are stability, predictability, and
efficiency. Formal rules and policies hold the organization together."
These descriptions fit very well with the scores in Table 4.3. Comparing
the four climate types one notices that the group and developmental have
very similar scores, but they differ on their resistance to change. It is also
interesting to notice that despite the fact that the scores are similar in the
group and developmental climates then they may mean different things as
described above. Similarly, the internal process and rational goal climates
are very similar. The main difference is again their resistance to change.

4.6 Climate as a Contingency

The discussion in the previous sections of this chapter provides us with a
limited set of dimensions that fit nicely into the multi-contingency model.
The three rules developed by Koys and DeCotiis (1991) sort out the
confusion about using dimensions from both the right-hand-side and left-
hand-side of the model in figure 1.3. This allows us to develop a two-stage
model relating the climate dimensions to structural properties via the
climate typology based on the competing values approach. The model is
depicted in Figure 4.2.
This model first maps the seven climate dimensions into four climate
categories from which propositions between climate and structural form are
developed. These seven dimensions, or "questions should focus on the
specific organizational units with recognized boundaries, not an ambiguous
'work environment'" (Glick, 1985, p. 608). The four climate types each
represents four ways to process information and thus which organizational
structure that will be most effective.

Climate Description Climate Types Structural Properties

Trust Group Configuration

Conflict Rational goal Centralization
Morale Internal process Formalization
Rewards equitability Developmental Complexity
Change resistance Coordination and
Leader credibility Control
Scapegoating Media Richness
Incentives

Figure 4.2. The Two-stage Model: Description, Climate Types, Structure

Properties
4.6 Climate as a Contingency 125

4.6.1 Describing a Group Climate

We describe the organization's climate in terms of the seven characteristics
about the organization: trust, level of conflict, employee morale, rewards,
the resistance to change, the leader credibility and the level of scapegoat-
ing. (Zammuto and Krakower, 1991, Burton et al., 1998). There are seven
propositions about a group climate. The propositions have been developed
from Table 4.3. The greater the number of these propositions that are true,
the greater the likelihood that the organization's climate is a group climate.
A group climate has a high degree of trust; it is open and sharing of
information among its members. Proposition 4.1 summarizes this notion of
trust for a group climate. Further, a group climate usually has a low degree
of conflict as stated in Proposition 4.2. If conflict exists, it is constructive
and tends to strengthen the group, rather than destroy the group, i.e., there
can be disagreement for the group purpose itself. This is usually coupled
with a high, or moderately high degree of employee morale as suggested in
Proposition 4.3. Individuals feel that they belong and are part of the group.
The equitable distribution of rewards is also consistent with the group
climate as given in Proposition 4.4. These rewards need not be equally
distributed, but there must be a sense of fairness where the basis for the
distribution is understood and accepted by the individuals in the organiza-
tion. Group climates are normally resistant to change as summarized in
Proposition 4.5. In a group climate, leaders have a high degree of credibility
with the group and can act on behalf of the organization. Proposition 4.6 is
consistent with a high degree of trust. Finally, Proposition 4.7 states that
there is little scapegoating in a group climate. Failure can be recognized
and accepted without the transfer of blame to those who are not responsi-
ble.

Description Climate Structure

High trust Group Adhocracy (or matrix) configuration
Low conflict Low to medium complexity
High or medium morale Low centralization
Equitable rewards Low formalization
High change resistance Wide span of control
High leader credibility Coordination and control through
Low level of scapegoating meetings and liaison roles
High media richness with a large
amount of information
Results based on incentives related
to groups

Figure 4.3. Group Climate: Description and Effect

126 Chapter 4. Organizational Climate

Table 4.4. Propositions 4.1-4.7

4.1. If the level of trust in the organization is high, then the climate is
likely to be a group climate.

4.2. If the conflict in the organization is low, then the climate is likely to be
a group climate.

4.3. If the employees' morale is high or medium, then the climate is likely
to be a group climate.

4.4. If the organizational rewards are given with equity, then the climate
is likely to be a group climate.

4.5. If the resistance to change is high, then the climate is likely to be a

group climate.

4.6. If the leader credibility is high, then the climate is likely to be a group
climate.

4.7. If the level of scapegoating is low, then the climate is likely to be a

group climate.

Overall, the group climate has a consistent pattern of beliefs and attitudes
about desirable behavior. A group climate has an open and a free flow of
information among the individuals in the organization. Trust, low conflict
and equity of rewards encourage individuals. Further, high morale and
little scapegoating help. Information becomes a kind of public good which
is shared and widely available. Information is more likely to be "broadcast"
than "channelled." "Need to know" is replaced by "everybody knows." There
are few secrets. The group climate can handle complex sets of information.
The implications of a group climate for the fit of the structure are
developed next.

4.6.2 The Group Climate Effects on Structure

A group climate has implications for the structure of the organization. The
group climate sets an atmosphere which is compatible or fits a specific set
of structural properties. Proposition 4.8 suggests that an adhocracy fits
with a group climate. The fluidity of the adhocracy is consistent with a high
4.6 Climate as a Contingency 127

level of trust, employee morale and commitment: Sometimes, an adhocracy

has conflict, but such controversy can be directed to the organizational
purpose that fits a group climate. The group is a very stable climate with
a high resistance to change. A group climate could also fit with other config-
urations, e.g. a matrix, but a bureaucracy seems unlikely to work well.
Proposition 4.9 suggests that a group climate requires a low to medium
organizational complexity with a low vertical differentiation. I.e. there are
not many layers in the organization. There is not an emphasis on the
division of work or detailed task definition. The group itself takes responsi-
bility in more aggregate form. Task boundaries are not necessarily sharp
and well defined for the individual. This is also consistent with a low
formalization as given in Proposition 4.10 where there are few written rules
and again the organization is more fluid in its operations (Zammuto and
Krakower, 1991, Hunt, 1991 and Quinn and Rohrbaugh, 1983). Normally,
a low formalization is coupled with high centralization, but for a group
climate, Proposition 4.11 summarizes evidence that the group climate
requires low centralization. (Zammuto and Krakower, 1991 and Hunt,
1991). In the group climate, decisions can be made appropriately through-
out the organization. The span of control can be wide as suggested in
Proposition 4.12. There is not the need for detailed hands on control.
However, activities must be coordinated and control realized by some
means. Proposition 4.13 suggests a need for meetings and many liaison
relations. These mechanisms must then serve the function that otherwise
would be obtained through either centralization or formalization (Hunt,
1991). The group climate can process complex and a large amount of
information relative to its size. Coordination via meetings and liaison is
appropriate. This is captured in Proposition 4.14. Finally, as stated in
Proposition 4.15, the incentives must be results oriented; if for no other
reason there are few procedures. Further, these incentives should reward
the group as matters of equity and reinforcement of the group itself
In brief, the group climate provides the necessary information; so
structural mechanisms can be reduced, and the organization should employ
configurations and structural mechanisms which complement and enhance
the abundance of information already available in the group climate.

Table 4.5. Proposition 4.8-4.15

4.8. If the climate is a group climate, then the configuration should be an

adhocracy or a matrix.

4.9. I{the climate is a group climate, then the complexity should be low to
medium with low vertical differentiation.
128 Chapter 4. Organizational Climate

4.10. If the climate is group, then the formalization should be low.

4.11. If the climate is group, then the centralization should be low.

4.12. If the climate is group, then the span of control should be wide.

4.13. If the climate is group, then the coordination and control should be
via integrators and group meetings.

4.14. If the climate is group, then the media richness should be high with
a large amount of information.

4.15. If the climate is group, then the incentives should be result based
with a group orientation.

4.6.3 Describing the Developmental Climate

Some of the propositions above for the group climate are repeated for the
developmental climate. For both, the trust is high, conflict is low, the
morale is high, with relatively equitable rewards. The significant difference
is the resistance to change which was high in a group climate, but is low in
a developmental climate. Propositions 4.16-4.20 state those conditions for
the developmental climate. However, the developmental climate scores are
not so extreme as for the group climate. The evidence (Zammuto and
Krakower, 1991 and Burton et al., 1998) is that the trust for a developmen-
tal climate may not be so high as for a group climate although the
difference seems minimal. For the developmental climate, there is a greater
focus on the growth of the organization itself. This is the basis for the low
resistance to change. Similarly, the rewards can be more individual with
less attention to the internal equity as perceived by the group. Individual
contribution for the organization is more important. This is a more external
orientation where success is in part realized more outside the organization.
There are also small differences with respect to leader credibility and
the level of scapegoating-Propositions 4.21 and 4.22.
The developmental climate has different information characteristics to
the group climate. The group climate will focus relatively more on internal
information while the development climate focus more on external
environmental information. Environmental information is likely to have
more value for development and growth. Additional compromise is
important, (Quinn and Kimberley, 1984 ).
4.6 Climate as a Contingency 129

Description Climate Structure

High to medium trust Developmental Matrix configuration

Low conflict Medium complexity
High to medium morale Low formalization
Moderate to high Low-medium centralization
equitable rewards: Medium span of control
Low change resistance Coordination and control
High to medium leader through planning, meetings,
credibility and integrators
Low to medium level of High media richness with a
scapegoating large amount of information
Incentives results based on
individuals

Figure 4.4. Developmental Climate: Description and Effect

Table 4.6. Proposition 4.16-4.22

4.16. If the level of trust in the organization is high to medium, then the
climate is likely to be a developmental climate.

4.17. [fthe conflict in the organization is low, then the climate is likely to
be a developmental climate.

4.18. [fthe employees' morale is high or medium, then the climate is likely
to be a developmental climate.

4.19. [fthe organizational rewards are given with high to moderate equity,
then the climate is likely to be a developmental climate.

4.20. If the resistance to change is low, then the climate is likely to be a

developmental climate.

4.21. If the leader credibility is high to medium, then the climate is likely
to be a developmental climate.

4.22. If the level of scapegoating is medium, then the climate is likely to be

a developmental climate.
130 Chapter 4. Organizational Climate

4.6.4 Developmental Climate Effect on Structure

The matrix configuration is suggested in Proposition 4.23 as an appropriate
structure for the developmental climate. For the group climate, an
adhocracy is suggested as well. Both climates require fluidity and "give and
take" to be successful. It is this similarity that we want to emphasize.
Organizational complexity, formalization and centralization have the same
recommendations for the two climates: complexity- medium or perhaps low,
formalization-low, and centralization-low-medium, as stated in Propositions
4.24, 4.25, and 4.26 (Zammuto and Krakower, 1991, Hunt, 1991 and Quinn
and Rohrbauch, 1983). The supporting arguments follow the same logic and
empirical evidence. The medium complexity again includes a low vertical
differentiation. Here the external focus suggests that a tall organization
might be less adaptable than required for the external and growth
orientation of the developmental climate. The low formalization and low to
medium centralization may be recommended for different reasons. Here
again, the change requirements of the environment and for growth requires
flexibility that are not compatible with greater formalization and central-
ization. Proposition 4.27 suggests a medium span of control. The coordina-
tion and control mechanisms are again meetings with negotiations (Hunt,
1991), but more detailed planning is likely to be required here than for the
group climate where there is a more implicit understanding about what to
do. Similarly, high media richness and a large amount of information is
required (Proposition 4.29); here there should be a richer kind of informa-
tion for a more outside focus. Finally, the incentives should be results
based, but more individually oriented as shown in Proposition 4.30. There
is less focus on the group and there are likely to be external success factors
that can be interpreted for the individual.
Here too, the information processing characteristics ofthe development
climate affect the configuration and structure. As with the group climate,
the development climate yields an open and abundance of information; the
organization structure should be complementary. A more open structure is
desired which recognizes the already abundance ofinformation. The matrix
configuration, low vertical differentiation, a medium span of control, low
formalization, low-medium centralization, integrators and meetings; all are
consistent with an abundance ofinformation. Here, this information should
include extensive external content.

Table 4.7. Propositions 4.23-4.30

4.23. If the climate is developmental, then the configuration should be a

matrix.
4.6 Climate as a Contingency 131

4.24. If the climate is developmental, then the complexity should be medium

with low vertical differentiation.

4.25. If the climate is developmental, then the formalization should be low.

4.26. If the climate is developmental, then the centralization should be low

or medium.

4.27. If the climate is developmental, then the span of control should be

medium.

4.28. If the climate is developmental, then the coordination and control

should be via planning, integrators and meetings.

4.29. If the climate is developmental, then the media richness should be

high with a large amount of information.

4.30 If the climate is developmental, then the incentives should be results

based with an individual orientation.

4.6.5 Describing the Internal Process Climate

The internal process climate is consistent, although it is quite different
from the two above: the group and the developmental climates. The internal
process climate is opposite in many ways. To begin, the trust is low as
stated in Proposition 4.31. There are not a sharing and open atmosphere
among the individuals as each is more inwards and guarded. Conflict is
high in the organization and disagreement over both means and ends are
prevalent; see Proposition 4.32. The next Proposition 4.33 indicates that the
employees' moral is medium to low although there is somewhat conflicting
evidence. Intuitively, it fits that morale would be low under an atmosphere
of distrust and conflict. Differences in the empirical results may also be
attributed to the fact that the two studies have been carried out in two
different settings and in two different countries. This raises the important
question how the national culture may affect the organizational climate,
(Hofsteede, 1980).
Rewards are perceived to be given inequitably as stated in Proposition
4.34. Here too, it seems reasonable that rewards would be viewed with low
equity in this atmosphere. The next Proposition 4.35 states that there is a
high resistance to change. Perhaps this outcome is less intuitive as it might
be argued that a change, any change, would be welcome. But the evidence
132 Chapter 4. Organizational Climate

suggests that individuals prefer to keep what they have and not engage in
activities that could lead to a different situation-although it could be better.
Perhaps it is the leader credibility, which is low from Proposition 4.36 and
the low level of trust that helps explain this reluctance. Yet, there is little
faith in the leader and consequently not much hope that the situation
would improve. The high level of scapegoating from Proposition 4.37 seems
consistent with this story about the organization.

Description Climate Structure

Low trust Internal Bureaucracy, functional configura-

High conflict process tion
Medium to low morale High complexity
Not equitable rewards High formalization (Quinn and Kim-
High change resistance berly)
Low leader credibility High-medium centralization
High level of scapegoating Wide span of control
Coordination and control through
rules and procedures, planning
Low media richness and moderate
amount of information
Individual and procedure based
incentives

Figure 4.5: Internal process Climate: Description and Effect

The internal process climate does not possess the capacity to process a lot
of information. The organization structure must supply the requisite
information processing capacity. There is not a norm of sharing and
openness. Information tends to be private and within the role. Information
is passed on within prescriptions and according to procedures. Information
is closely associated with the job or task, or "a need to knOw." The
spontaneous information links are largely missing, or not utilized.

Table 4.8. Propositions 4.31-4.37

4.31. [{the level o{trust in the organization is low, then the climate is likely
to be an internal process climate.

4.32. [{the conflict in the organization is high, then the climate is likely to
be an internal process climate.
4.6 Climate as a Contingency 133

4.33. I{the employees' morale is medium to low, then the climate is likely
to be an internal process climate.

4.34. I{the organizational rewards are given inequitably, then the climate
is likely to be an internal process climate.

4.35. I{the resistance to change is high, then the climate is likely to be an

internal process climate.

4.36. I{ the leader credibility is low, then the climate is likely to be an

internal process climate.

4.37. I{the level o{scapegoating is high, then the climate is likely to be an

internal process climate.

4.6.6 Internal Process Climate Effects on

Structure
The internal process climate has implication for the structure. Generally,
the structure should be tighter and more "structured" with an emphasis on
the mechanism of the organization itself. In many ways, the organization
must be the complement to the lack of "togetherness" among the individu-
als. Proposition 4.38 states the configuration should be a bureaucracy and
is consistent with a functional organization where the coordination is
realized through the organizational mechanisms themselves. This view is
consistent with a high organizational complexity as stated in Proposition
4.39 where there are many specialized roles and a tall organization.
Formalization is high (Zammuto and Krakower, 1991 and Hunt, 1991) with
lots of written rules which govern actions and behavior. With a low level of
trust and a high level of conflict many written rules are needed to ensure
that the information is processed in the right way. Centralization is also
high where decisions are made at the top (Zammuto and Krakower, 1991
and Hunt, 1991) as shown in Proposition 4.41. Formalization and central-
ization can both be means to coordinate activities, but high centralization
and high formalization may be more than is required. The internal process
climate is quite demanding for mechanisms to maintain order and control
in the traditional sense. Although the span of control can be wide as given
in Proposition 4.42, this notion is reemphasized in Proposition 4.43 with a
call for rules and procedures for coordination and control (Hunt, 1991). The
media richness, Proposition 4.44, can be low (Hunt, 1991) with a relatively
134 Chapter 4. Organizational Climate

moderate amount of information required. Given the climate and the

atmosphere and the organization itself, the incentives should be procedural
based for the individual as stated in Proposition 4.45. An individual should
have a well-definedjob and responsibility with rather narrow limits and the
reward question is whether the individual does the task appropriately or
not. The internal process climate requires well structured and well
documented information processing to be effective and efficient.

Table 4.9. Propositions 4.38-4.45

4.38. If the climate is internal process, then the configuration should be a

bureaucracy or functional.

4.39. If the climate is internal process, then the complexity should be high
with both high vertical and horizontal differentiation.

4.40. If the climate is internal process, then the formalization should be

high.

4.41. If the climate is internal process, then the centralization should be

medium or high.

4.42. If the climate is internal process, then the span of control should be
wide.

4.43. If the climate is internal process, then the coordination and control
should be by rules and procedures.

4.44. If the climate is internal process, then the media richness should be
low with a moderate amount of information.

4.45. If the climate is internal process, then the incentives should be

procedures based with an individual orientation.

4.6.7 Describing the Rational Goal Climate

The rational climate is closer to the internal process climate than to the
group and developmental climates although they are different. The main
difference is the score on the resistance to change. The rules are given in
Propositions 4.46-5.52. The rational goal climate is also structured with an
4.6 Climate as a Contingency 135

emphasis on planning, productivity, and efficiency (Quinn and Kimberley,

1984).

Description Climate Structure

Low trust Rational Divisional Configuration

High conflict goal High complexity
Medium to low morale Medium formalization
Low equitable rewards Medium centralization
Medium to low change Medium span of control
resistance Coordination and control through
Low leader credibility planning and meetings
High level of scapegoating High amount of information and
medium media richness
Incentives based on results for
both groups and individuals

Figure 4.6. Rational Climate: Description and Effect

Information processing in the rational goal climate is similar to internal

process climate but with a greater emphasis on environmental/external
information. The low level of trust, high conflict, etc. lead to a private view
of information, and sharing and exchange of information does not occur
spontaneously. The rational goal climate is a very competitive environment
to work in. It is not to be expected that the employees will be loyal to the
organization in the sense that high turnover can be expected. With the low
resistance to change many reorganizations at the personnel level may be
expected with a very tough competition for the prestigious jobs.

Table 4.10. Propositions 4.46-4.52

4.46. If the level oftrust in the organization is low, then the climate is likely
to be a rational goal climate.

4.47. If the conflict in the organization is high, then the climate is likely to
be a rational goal climate.

4.48. If the employees' morale is low-medium, then the climate is likely to

be a rational goal climate.

4.49. If the organizational rewards are inequitable, then the climate is

likely to be a rational goal climate.
136 Chapter 4. Organizational Climate

4.50. If the resistance to change is low to medium, then the climate is likely
to be a rational goal climate.

4.51. If the leader credibility is low, then the climate is likely to be a

rational goal climate.

4.52. If the level of scapegoating is high, then the climate is likely to be a

rational goal climate.

4.6.8 Rational Goal Climate Effects on Structure

The rational goal climate implications for the structure are similar to those
of the internal process climate. This is not surprising as the climate
descriptions themselves were in the same spirit. Nonetheless, there is a
variation in emphasis and tone. The rational goal climate calls for a less
"tight" structure and has a more external focus to its orientation. To begin,
a divisional configuration is recommended in Proposition 4.53. The more
external focus is compatible with the divisional structure. The organization
should use the competitive environment. A very competitive environment
may be very harmful for the functional configuration (Burton and Obel,
1988). Complexity should be high (Proposition 4.54) with high specializa-
tion of tasks. However, formalization (Proposition 4.55) and centralization
(Proposition 4.56) should be medium (Zammuto and Krakower, 1991, Hunt,
1991 and Quinn and Rohrbaugh, 1983) where they should be high for the
internal process climate. This is a less "tight" mechanism of organization
and is compatible with the greater external focus here. The span of control
should be medium as given in Proposition 4.57. Coordination and control
should use more process in planning and meetings than for the internal
process climate (Proposition 4.58). Media richness should be a high amount
with a medium richness of the information as suggested in Proposition
4.59. Finally, the incentives should be more results-oriented for both the
individual (Hunt, 1991) and the group-Proposition 4.60. This, too, is very
compatible with the more external competitive orientation of the rational
goal climate and fits very well with the divisional configuration recom-
mended here.
Here again, the rational goal climate with its limited information
processing requires a higher level of information processing within the
structure of the organization. The divisional configuration mitigates some
need for information. The organization is structured to gather, communi-
cate, and interpret the requisite information to manage. High complexity
divides the tasks which must be coordinated and controlled with medium
formalization, medium to high centralization, and an emphasis on planning
4.7 Managing the Climate 137

and meetings. In brief, the rational goal climate requires that the organiza-
tion itself procures the requisite information.

Table 4.11. Propositions 4.53-4.60

4.53. If the climate is rational goal, then the configuration should be

divisional.

4.54. If the climate is rational goal, then the complexity should be high.

4.55. If the climate is rational goal, then the formalization should be

medium.

4.56. If the climate is rational goal, then the centralization should be

medium.

4.57. If the climate is rational goal, then the span of control should be
medium.

4.58. If the climate is rational goal, then the coordination and control
should be via planning and meetings.

4.59. If the climate is rational goal, then the media richness should be
medium with a large amount of information.

4.60. If the climate is rational goal, then the incentives should be results
based with a group and individual orientation.

4.7 Managing the Climate

We argued in chapter 1 that fit is important. To obtain fit both sides in the
model in Figure 1.3 should be aligned. This means that to obtain a
contingency fit with structure and climate, you may change the structure
or you may try to adjust the climate.
To change the perception of the climate in an organization quickly may
not be that easy. Generally the climate is rather stable (Campbell et ai,
1970). It may take a long time to build trust, morale, and a perception that
rewards are given in an equitable way, but it may be very easy to destroy.
Management may try to build a special psychological climate, by
instituting a number of rituals and by showing over time that they are open
138 Chapter 4. Organizational Climate

and can be trusted. A special climate may also be instituted by hiring

people with special views and values. Theory X people will normally show
less trust and more resistance to change than do Theory Y people; see
Chapter 3.
The choices of strategy, environment, and the organizational design
itself provide feedback to the people employed in the organization and may
affect their perception of the working climate and thus affect the way
information is processed. For example, the anti Microsoft movement both
in government and on the Internet is said to affect the perception of the
Microsoft employees that Microsoft is a "great" place to work. This may
affect the work morale at Microsoft. Again a proper alignment is called for.

4.8 The Climate at SAS

Throughout its many years of steady growth, SAS had developed into a
proud and respected organization-well known for its high quality and good
service. SAS's reputation for high quality was widely attributed to the
management's constant focus on new technologies, new aircrafts and new
engines.
SAS was reputed for the loyalty of its employees. In return. SAS
treasured its employees by offering most of them life-long employment
relationships, and a wage level generally higher than that of most
competitors.
With the appointment ofCarlzon in 1980, many of these changes were
initiated. Carlzon's first goal was to change the strategy of SAS into that
of a more market-oriented strategy. Since SAS could not compete on size
nor prices, employees were considered to be SAS's best means for differenti-
ation. By offering a better service to customers, SAS would create
preferences in the customers minds. As employees were the ones in daily
contact with customers, the implementation ofCarlzon's vision necessitated
the co-operation of all employees.
To create a higher team spirit as well as heighten the motivation of
employees, Carlzon made a big effort to align the corporate climate with the
new strategy.
In this way, the management formulated common ideas and values that
SAS should stand for in the future. The basic message was that SAS now
had to create a service climate, meaning an increased focus on customers
wish and demand.
Efforts were made to heighten the co-operation across functional areas
such as cabin crew, cleaning-staff and technical staff. An amelioration of
the SAS service could not be achieved if the various functional groups could
not depend on each others to fill out their jobs to their uttermost. This was
4.8 The Climate at BAS 139

necessary as all areas were interdependent and thus deeply dependent on

each other.
Especially front line employees, who were most radically affected by the
changes, were enthusiastic. A team spirit large enough to evoke the
admiration of most competitors was created.
Unions and middle managers were harder to convince, however. The
changes brought along an obliteration of occupational demarcations and a
reduction in the number of staff which the unions obviously did not favor.
Middle managers were not too enthusiastic either, as they had been
deprived of much of their earlier authority.
Nevertheless, SAS created solid financial growth, and it was hard to
convince anybody that Carlzon's strategy was not a success. Indeed,
Carlzon was loved and admired by most employees as well as by most
management consultants, and even competitors. The success of SAS was
very much accredited to his personality.
Over time however Carlzon's entrepreneurial ideas got out of hand. In
his effort to enlarge the company, SAS made investments in areas that
were not related to its core products. The management slowly started to
lose sight of vision. As the earnings went down-so did the employees'
admiration for Carlzon.
Employees were still excited about their new working environment but
started criticizing Carlzon for his efforts to change the company via
reductions and cutbacks. One of the many criticisms of Carlzon seemed to
be that he had broken one of SAS's non written laws by not recruiting
insiders for the company's management positions. This was a problem, as
most employees believed that part of the reason for SAS's failing success
was that the new managers from the outside did not have the same
expertise in deciding what was good for SAS, as many of the senior
employees would have had. Certain barriers were thus created between the
management group and the employees. The deregulation worsened this
relationship, as employees felt that the management was not handling this
threat well enough-they found it hard to believe that the deregulation could
seriously be threatening SAS, considering the company's long history of
success.
The company suffered from the slowly increasing mistrust in Carlzon
and his management group, and it was difficult for Carlzon to make any of
the needed rationalizations without losing even more of the employees'
faith, something which he could not afford.
Stenberg, who was appointed in April 1994, did not conceal the fact that
his major goal was to reduce the costs of SAS, and that this would
necessitate major cutbacks and rationalizations.
Although many of Stenberg's rationalizations have led to employee
frustrations, nobody can deny the fact that Stenberg has managed to get
140 Chapter 4. Organizational Climate

SAS back on its feet financially. Stenberg is respected for this, but in a less
impassioned way than Carlzon used to be.
SAS today is probably a less symbiotic company than it used to be. The
company has been haunted by personnel strikes, particularly among the
cabin crew, who are dissatisfied with their new working conditions as well
as the many firings of their colleagues. The focus on costs has therefore in
some periods been at the expense of customers' wishes, threatening to
deteriorate the company's competitiveness. Another problem has been
employee discontent with not being let in on the management's plans.
Jan Stenberg follows another personnel policy than Carlzon, and
generally does not use as much open dialogue to communicate his plans and
visions to the employees. Although he does not use as much time communi-
catingwith employees, his overall sense of purpose is nevertheless probably
very clear to most of the employees; Jan Stenberg wants results! SAS has
been going through a number of changes in the organizational climate.
First, it seems that there was a high level oftrust, a little conflict, equitable
rewards, but a high resistance to change. This would come from Proposi-
tions 4.1, 4.2, 4.3, 4.4 and 4.5 lead towards a group climate. Carlzon with
his entrepreneurial spirit worked hard on changing the resistance to
change creating a developmental climate.
When problems started to emerge the level of trust fell, conflicts were
more common, as well as lack of credibility ofCarlzon and the management
group. Additionally, a resistance to major change in some sense reappeared.
This drove the climate towards an internal process climate.
Stenberg from a very different point of departure had the same ideas
with respect to Carlzon: to change the resistance to change. Coming from
an internal process climate he moved towards a rational goal climate.
Using the descriptions by Hoijberg and Petrock(1993) SAS has gone from
being a friendly place to work, passing by an entrepreneurial phase, to a
tough demanding place to work where the leaders are hard drivers,
producers, and results-oriented.

4.9 Summary
The organizational climate is an important factor in determining the design
of the organization. Climate is one of several factors, including the
management preference, size, environment, technology, and strategy which
influences an appropriate choice of the organizational design. In Figure 1.3,
the organization's climate then is one of the left-hand factors in determin-
ing the organization's structure on the right-hand side. E.g., the climate
helps set the appropriate level of centralization-a high level of trust
suggests a less formalized and more decentralized organization.
4.9 Summary 141

The organization's climate and its culture are intertwined in every day
managerial usage and in some ofthe literature. Yet, they are quite distinct
concepts; climate is a prevailing condition or atmosphere in an organiza-
tion, where the culture is pattern of behavior including commonly held
beliefs and social forms. The climate is more enduring and given, where the
organizational structure is embedded in the culture. Climate is more given;
culture can be learned and transferred from one individual to another. In
terms of design, climate is a contingency on the left hand side of our
multidimensional model in Figure 1, culture is emergent in the choice ofthe
right hand side design elements.
The organization's climate is a measure of the organization-not a
measure of the individuals in the organization. Nonetheless, we look to the
individuals for our information about the climate. Climate is then
measured in terms of trust, conflict, morale, equity of rewards, resistance
to change, leader credibility and the level of scapegoating. Individuals
"know" what the climate is and they are our source of information. These
measures are reliable and valid, i.e., individuals canjudge these issues and
respond in a consistent and meaningful way.
The competing values approach can then be utilized to summarize and
categorize the climate into four types: group, rational goal, internal process
and developmental. The seven measures above can be mapped into the four
climate categories as suggested in Figure 4.2. E.g., a high level of trust is
an element of a group climate, which in tum suggests a more decentralized
organization.
Each of the four climate types is consistent with a different organiza-
tional structure. The group climate not only indicates greater decentraliza-
tion, but also less formalization and greater use of meetings for coordina-
tion. A less trusting climate should be less decentralized and more
formalized in its structure, with a greater emphasis on fixed assignments
to achieve coordination.
The support for these recommendations comes in part from the
literature, but also from an understanding ofthe behavior of the organiza-
tion and how it uses information. A high trust organization tends to have
a more open and free flow of information and thus, coordination can be
obtained with a good deal oflocalized decision making or decentralization.
In the less trusting organization, information is less shared and more
guarded and localized and thus, coordination is realized within more
formalized structures and more centralized decision making.
The climate-structure relations are more complex and multidimensional
going beyond trust and decentralization. Climate is not only trust, but
includes six other dimensions; the organizational design incorporates not
only centralization, formalization and means for coordination, but the
configuration, rewards, etc. as well as presented in Figure 4.2. Further,
142 Chapter 4. Organizational Climate

climate is complemented by other contingencies in determining an

appropriate design as given in Figure 1. The multidimensional contingency
model is now more complete with the incorporation of climate as a factor in
the organization's design.
In this chapter, we have incorporated climate as a contingency in the
multidimensional contingency model of organizational design. Climate
gives us more important information to help determine an appropriate
design which is effective and efficient.

References
Bluedorn, Allen C. and Earl F. Lundgren. 1993. "A Culture-Match Perspective for
Strategic Change." Research In Organizational Change And Development, 7,
137-179. I

Burton, Richard M. And B!6rge Obe1.1988. "Opportunism, Incentives, arid the

M-form Hypothesis: A Laboratory Study." Journal of Economic Behavior and
Organization, 10,99-119.
Burton, Richard M., J!6rgen Lauridsen, and B!6rge Obel. 1998. An Empirical
Analysis of Climate in an Organizational Context, Working Paper.
Cameron, Kim S. and Sarah J. Freeman. 1991. "Cultural Congruence, Strength, and
Type: Relationships to Effectiveness." Research in Organizational Change and
Development, 5, 23-58.
Campbell, J., M.D. Dunnette, E.E. Lawler III, and Karl E. Weick. 1970. Managerial
Behavior, Performance, and Effectiveness. McGraw-Hill.
Cooper, Randolph B. & Robert E. Quinn. 1993. "Implications of the Competing
Values Framework for Management Information Systems." Human Resource
Management, 32(1), John Wiley & Sons, 175-201.
Delery, John E. and Harold D. Doty. 1996. "Modes of Theorizing in Strategic Human
Resource Management: Tests of Universalistic, Contingency; and Configura-
tional Performance Predictions." Academy of Management Journal, 36(4),
802-835.
Denison, Daniel R. 1990. Corporate Culture and Organizational Effectiveness. Wiley.
Denison, Daniel R. and Gretchen M. Spreitzer. 1991. "Organizational Culture and
Organizational Development: A Competing Values Approach." Research in
Organizational change and Development, 5,1-21.
Denison, Daniel R.. 1996. "What is the difference between organizational culture
and organizational climate? A Native's point of view on a Decade of Paradigm
Wars." Academy of Management Review, 21(3), 619-654.
Giek, DonaldG. & Patricia 1. Lees. 1993. "On Massive Change: Using the
Competing Values Framework to Organize the Educational Efforts of the
Human Resource function in New York State Government." Human Resource
Management, 32(1), John Wiley & Sons, 9-28.
Glick, William H. 1985. "Conceptualizing and Measuring Organizational and
Psychological Climate: Pitfalls in Multilevel Research." Academy ofManagement
Review, 10(3),601-616.
References 143

Hooijberg, Robert & Frank Petrock. 1993. "On Cultural Change: Using the
competing Values Framework to Help Leaders Execute a Transformational
Strategy." Human Resource Management, 32(1), John Wiley & Sons, 29-50.
James, Lawrence R and Allan P. Jones. 1974. "Organizational Climate: A Review
of Theory and Research." Psychological Bulletin, 81(12) 1096-1112.
Koys, Daniel J. and Thomas A DeCotiis. 1991. "Inductive Measures of Psychological
Climate." Human Relations, 44(3), 265-285.
Litwin, George H. and Robert A. Stringer Jr. 1968. Motivation and Organizational
Climate. Harvard University.
Poole, Marshall Scott and Robert D. McPhee. 1983. "A Structural Analysis of
Organizational Climate." Linda L. Putman end Michael E. Pacahowsky (eds)
Communication and Organizations. An InterpretiveApproach, Sage Publications.
Poole, Marshall Scott. 1985. "Communication and Organizational Climate: Review,
Critique, and a New Perspective." Robert D. McPhee and Phillip K. Tompkins
(eds) Organizational Communications: Traditional themes and New Directions,
Sage Publications.
Quinn, RE., & J. Rohrbaugh. 1983. "A spatial model of effectiveness criteria:
Towards a competing values approach to organizational analysis." Management
Science 29,363-377.
Quinn, Robert E. And John R Kimberly. 1984. "Paradox, Planning, and Persever-
ance: Guidelines for Managerial Practice." In John R Kimberly and Robert R.
Quinn (eds), Managing Organizational Transitions, Dow-Jones-Irwin, 295-314.
Quinn, Robert E. And Gretchen Spreitzer. 1991. "The Psychometrics of the
Competing Values Culture Instrument and an Analysis of the Impact of
Organizational Culture on Quality of Life." Research in Organizational Change
and Development, 5, 115-142, JAI Press, inc.
Rousseau, Denise M. 1988. "The Construction of Climate in Organizational
Research." C. L. Cooper and I. T. Roberson (eds), International Review of
Industrial and Organizational Psychology, 3, John Wiley & Sons, 139-158.
SAS. 1994. SAS Annual Reports. Stockholm: SAS.
SAS. 1995. Annual Report. Stockholm: SAS.
SAS. 1996. MiljfJregnskab. Stockholm: SAS
SAS. 1996. Annual Report. Stockholm: SAS
SAS. 1996. De ffJrste 50 dr. Stockholm: SAS
SAS. 1996. ScandinatianAirlines 1946-1996. Stockholm: SAS
SAS. 1997. Deldrsrapport 1: 1997. Stockholm: SAS
Scheider, Benamin (ed.), 1990, Organizational Climate and Culture, San Francisco,
Jossey-Bass.
Schein, Edgar H. 1992. Organizational Culture and Leadership, San Francisco,
Jossey-Bass. (Second edition).
Siehl, C. And J. Martin. 1990. "Organizational culture: A Key to Financial
Performance." B. Schneider eds. Organizational Climate and Culture. Jossey-
Bass.
Tagiuri and George H. Litwin. 1968. "Organizational Climate." Harvard University.
Tosi, Henry. 1992. The Environment/Organization/Person Contingency Model:A
Meso Approach to the Study of Organization. JAI Press.
144 Chapter 4. Organizational Climate

Yeung, Arthur KO., J. Wayne Brockband, and David O. Ulrich. 1991. "Organiza-
tional Cultures and Human Resource Practices: An Empirical Assessment."
Research in Organizational Change and Development, 5, 59-81.
Zammuto, Raymond F. And Jack Y. Krakower. 1991. "Quantitative and Qualitative
Studies in Organizational culture." Research in Organizational Change and
Development, 5, 83-114.
CHAPTER 5

Size and Skill Capabilities

5.1 Introduction
The Institute of Applied Computer Science (introduced in Chapter 3) was
negotiating a contract with a European Union research agency. This
contract would be the largest single contract in the history of the company.
The contract will require Applied Computer Science to employ about forty
people and to be the coordinating unit for a number of research teams
located in other European countries. Benny Mortensen was not worried
about the scientific and technological aspects of the new project but
expressed concerns about running a company that would be twice its
current size. He realized that he no longer would be as heavily involved in
every project and that more traveling and time spent on recruiting would
leave him less time to have detailed knowledge of how projects are doing.
To a large extent, the success of the company depended on his ability to
utilize resources efficiently, thus cutting cost and time, by using knowledge
from one project to improve on another project.
The growth ofthe Institute ofApplied Computer Science generated some
management concerns. The manager was concerned that quality and cost
control could not be maintained as he became less involved in particular
projects. His management style and personality have been important for
the success of Applied Computer Science. Its growth may require a
structural change, but Benny Mortensen probably will try to develop a
structure that fits not only the company's particular size and growth
pattern but also his own style and needs. Furthermore, the working
conditions for employees of Applied Computer Science will change, too. It
is important that there is a fit between working conditions and the
individuals employed by the organization.
It is obvious that the increase in size changes the management and
organizational problems of this company. As the size increases, the CEO
may have to increase the company's level of decentralization. Centralized
decision making has been the main source of coordination of the activities.
Now Benny Mortensen needs to think about other ways to ensure
coordination. Project cost and quality were also supervised by the CEO.
During the startup many administrative procedures were carried out by
Benny Mortensen or other employees. The Institute of Applied Computer
Science does not have people in administrative functions, but with growth,
146 Chapter 5. Size and Skill Capabilities

administrative positions may have to be created. This will allow Benny

Mortensen to use his information-processing capacity on coordination and
decision making. With increased decentralization and specialization,
formalized rules or incentive mechanisms may also be changed.
The small company-with its low formalization and high centraliza-
tion-had a competitive advantage by being flexible and fast in its decision
making. Introducing a high degree of formalization may be appropriate
with respect to cost and quality but may hamper flexibility. As the company
grows, Benny Mortensen should reassess the organizational structure.
In recent years, size has entered the discussion in two ways. First, in
many industries, e.g. pharmaceuticals, and banking, a merger wave have
been seen creating new organizations much larger than before. Second,
many have preached "small is beautiful" with the result of downsizing and
outsourcing.
Generally, size is an important contingency for organizational design.
The size paradigm has been around for some time (Spencer, 1898). Yet
there have been a number of controversies about size and its effect on the
organization.

5.2 A Literature Review 1

5.2.1 Size: A Variety of Perceptions
Organizational size has been of interest to social scientists for a long time.
Over ninety years ago it was asserted that the effect of size "is a character
of social bodies, as of living bodies, that while they increase in size they
increase in structure" (Spencer, 1898, vol. 2, bk. 2, p. 440). More recently,
sociologists and organizational theorists (e.g., Blau, 1970; Blau and
Schoenherr, 1971; Hickson, Pugh, and Pheysey, 1969; Meyer, 1972; Pugh,
Hickson, Hinings, and Turner, 1969; Kimberly, 1976; Slater, 1985) have
investigated the relation between size and structure-that is, size as an
imperative factor explaining the organizational structure. In these and
other works, size is operationalized in a variety of ways: number of
employees, number of products or services, total sales, number of divisions,
and so forth. One path, in particular, seems to be well traveled-to
investigate the effect of administrative intensity or headquarters' burden.
Child (1973) states that "more studies have probably been carried out on
the proportion of employees occupying administrative or supportive roles
than on any other single aspect of organization structures" (p. 328).
An information processing model of the firm provides yet another
perspective, suggesting that the size of the firm is limited due to the
5.2 A Literature Review 147

costliness of coordinating the activities within an administrative structure

(Williamson, 1975). This is in general agreement with Starbuck (1965), who
suggests that, beyond some cost-optimal point, "managerial problems
become inordinately complex as size increases, producing progressively
higher production costs" (p. 459). Even here, size is not well defined. The
absence of a common view on how to measure size (or, more specifically, on
the use of differing measures or dimensions of the broadly defined "size"
construct) has made findings from size and structure hypothesis tests
somewhat equivocal. (')
Of the many debates in the literature, one is particularly illustrative.
Donaldson (1982), testing prior work, states that organizational diversity
(of products, particularly) is more highly associated with structure
(divisionalization, in particular) than is size (whether measured in sales,
assets, or number of employees). Grinyer (1982, p. 339), responding to
Donaldson, suggests that "more complex divisional structures tend to be
more bureaucratic," using "decentralized operating decisions" (see also
Grinyer and Ardekani, 1980). Furthermore, Grinyer suggests that while
Donaldson's hypotheses may hold for product-based divisionalization, they
do not hold for divisionalization based on other attributes, such as
geographic dispersion or some minimum size as a threshold for divisionali-
zation property. In short, Grinyer suggests that "it would be a mistake ...
to take too simplistic a view on this issue. Growth, size, diversification and
divisionalization are all strongly interconnected" (pp. 342-343). Child
(1982), commenting on the Donaldson-Grinyer debate, suggests that size
expressed as a threshold below which divisionalization is inappropriate and
above which appropriate is also an oversimplification. Clearly, a lack of
specificity with regard to the size construct makes equivocality a basic
problem (if not a hallmark) of such research.

5.2.2 The Measurement of Size

Inquiry into the concept of size has been constrained by problems involving
the operationalization of its measurement. Kimberly (1976) states that
such studies have been conducted in a theoretical wasteland since
hypotheses have typically been addressed apart from available theories of
complex organization. Slater (1985, p. 161) finds that most organizational
theorists "think of organizational size mainly in personnel terms," that is,
by counting organizational members.
Kimberly (1976), on the other hand, suggests that other aspects or
measures of organizational size (that is, other than number of members)
may be useful and that attention should be paid to the physical, fiscal,
input, and output dimensions as well. Kimberly concluded his 1976 work
148 Chapter 5. Size and Skill Capabilities

with three points. First, the ambiguous status of size (in the organizational
literature) is due to its being too global a measure to permit a clear
specification of its organizational role; size has numerous aspects, the
theoretical and empirical aspects of which must be made more specific.
Second, various aspects of size may fulfill differing causal or indicative roles
in various types of organizations. Finally, effective organizational conceptu-
alization depends on conducting inquiries with a dynamic (rather than a
static) perspective.
The various measures oforganizational size are not interchangeable due
to dependence on differing conceptual referents. However, many research-
ers fall back on total member counts since such counts are likely to be
highly correlated to a number of differently conceived nonpersonnel
measures (e.g., Anderson and Warkov, 1961; Hawley, Boland, and Boland,
1965; Pugh, Hickson, Hinings, and Turner, 1969). In other words, personnel
count is often considered a surrogate for other measures of organizational
size. There are, of course, other appropriate measures of size.
Ask any CEO, IIHow large is your company?1I The usual respond is, lIour
revenue is 100 million. 1I Revenues or sales is very often the size measure of
interest. In a market economy, revenue is a natural way to think about size.
But there are many others: assets, profits, countries, and the number of
employees. All of these measures of size are reasonable and are the correct
measure, but for different purposes.

5.2.3 Size as Imperative

While size itself has generated much interest, size as an organizational
determinant has also been of widespread interest. The global use of size as
an organizational measure is the hallmark of theorists who support a IIsize
imperative II-that is, organizational size as a determinant of differentiation
(Blau, 1970) and organizational structure (e.g., Blau and Schoenherr, 1971;
Hickson, Pugh, and Pheysey, 1969; Meyer, 1972). By imperative, we mean
a state or quality that IInecessarily aff'ect(s) organizations in a certain wayll
(Robey, 1982, p. 211). Size, in this research, was seen to be positively
related to increased differentiation, specialization, and formalization. These
assertions are subject to some criticism (Aldrich, 1972; Hall, Haas, and
Johnson, 1967; Mayhew, Levinger, McPherson, and James, 1972). Results
from these critical works are mixed, at best, and adherence to the size
imperative has been debated extensively (Donaldson, 1982).
5.2 A Literature Review 149

5.2.4 An Information.Processing Perspective

on Size
Size as an imperative for organizational structure has not been a major
concern in information processing models of organization, but one central
question does regard the limitation offirm size. Williamson (1975, p. 117)
summarizes the proposition that "the argument ... comes down to this: The
distinctive powers of internal organization are impaired and transactional
diseconomies are incurred as firm size and the degree ofvertical integration
are progressively extended, organizational form held constant." In other
words, increasing size (however measured) and high levels of centralized
decision making are posited to have a detrimental effect on the quality of
a firm's operations and, by extension, its outcomes. Earlier, Caplow (1957)
emphasized the problems of communication that result from increases in
the number of employees; as employees increase linearly, then the number
of communication networks increases exponentially.
The information-processing requirements of an organization follow "who
talks to whom about what." More simply, we can count who can talk to
whom by examining the number of possible communication links in an
organization.
Let us begin with a two person organization and then increase the
number of people. Pictures helps.

------2

3L------~2

~----""7I2

5~-f___\_-_jf 2

4 "'------~ 3

For a two-person organization, there is one link between the two. For three
people, there are three possible links. For four, there are six. For five, the
150 Chapter 5. Size and Skill Capabilities

number grows to 10, and the pattern continues. It is clear that the number
of possible links (1,3,6,10, ... ) is going up faster than the number of
individuals (1,2,3,4 ... ). The general formula for the number ofpossible links
is: (~). For N equal to 10, there are 45 possible links. For N equal to 100,
there are 4950 links.
For organizations with a very large number of people, the number of
links becomes almost unimaginable and it is impossible for everyone to talk
with everyone else. Network organization, without some reduction of the
communication possibilities, will quickly become too complex. We therefore
invent mechanisms so that an individual can talk with only a few individu-
als in an organized way to accomplish the overall organizational task.
The simplest way to decrease the number oflinks is to put someone in
charge and have everyone talk to that one person, but not to each other,
i.e., create a hierarchy. The simple organizational configuration, presented
in chapter 2, is such an organization. Even so, it works only for small
organizations as the ''boss'' becomes overwhelmed with the information; we
call it "information overload." This approach can be replicated and then we
have hierarchies with more than one level, i.e., there are bosses of bosses.
A hierarchy can grow and it can have a large number of people.
Hierarchies are efficient in reducing the need for information processing in
an organization. Hierarchies limit the number of information links in an
organization. But if there is a need for communication between two
individuals who are far removed in the organization, a hierarchy can be an
impediment to efficiency and effectiveness as the organizational tasks are
not done or at least, not done well. Hence, we also seek a few selected non
hierarchical liaison mechanisms to link individuals across the hierarchy:
Matrix organization, committees, task forces, ad hoc structures, to name a
few. But we introduce only a few of the many possible links. The issue is to
recognize and introduce the few which are needed to process information
for organizational efficiency and effectiveness.
New information technology changes the way people in an organization
can communicate. A simple e-mail system allows one to broadcast to many
people across hierarchical boundaries. Voice mail is another device that lets
people communicate without the constraint of being present at the same
time. Modern information technology force one to revisit the notion of size.
Brynjolfsson et al (1994) suggests that there is a positive correlation
between information technology investments and smaller firm sizes.
Information technology may reduce the need for bosses and may increase
the number of people a boss can handle. Brynjolfsson (1994) found that an
increase of information technology reduces the vertical differentiation in
firms.
Large firms are, however, seen as proportionately more costly to manage
than small firms. Size here is a general measure of the firm's activities that
5.3 The Effects of Size and Skill Capability on Organizational Structure 151

could be related to the number of products and processes. It may also be

positively correlated with the numbers of employees-but here the underly-
ing notion of size, has to do with transactional requirements. This does not
suggest a "one best way"; information processing models of organization
occur in a number of forms. In Galbraith's (1976) view, the central
managerial problem is to reduce uncertainty. Two primary strategies are
to reduce the need for information or to increase the capacity to process the
required information. Size does not enter the model explicitly. Yet it is clear
that more people, more products, and more processes increase the
complexity of organizational decision making. This might well increase
uncertainty, which is defined as the difference between the information
required and that available to manage the organization. Using common
dimensions of size, it seems reasonable that size increases uncertainty,
which leads to an increased need for coordination, which then influences
the choice of the best information processing structure. This is an inferred
size imperative-that is, that size can influence structure through size's
effect on information processing demands as well as the information-
processing capacity of the organization.

5.3 The Effects of Size and Skill Capa-

bility on Organizational Structure
5.3.1 Measuring Size and Skill Capability for
Design Purposes
The controversy about the effect of size on organizational structure is not
diminished by our desire to measure the size ofthe organization for design
purposes. Previous studies have established some positive and negative
correlations between various size measures and design variables in organi-
zational structure. But even those that agree on the correlations have very
different views about what a small, medium, or large organization is.
Robbins (1990, p. 61) defines "a large organization as one having approxi-
mately two thousand or more employees." Carillo and Kopelman (1991)
define small organizations as having fewer than ten employees, medium-
size organizations as having eleven to nineteen employees, and the large
organization as having twenty employees or more. Smith, Guthrie, and
Chen (1989) labeled organizations with less than thirty employees small,
those with thirty to 300 employees medium, and those with more than 300
employees large. Miller, Glick, Wang, and Huber (1991) use 475 employees
as a threshold between small and large organizations. Some authors do not
152 Chapter 5. Size and Skill Capabilities

categorize organizations as small, medium, and large but map number of

employees directly into, for example, a vertical differentiation measure. In
this case a functional relationship between number of employees and
number of levels in the organization is established. The mapping can be
linear or nonlinear. For example, Marsh and Mannari (1989) use a log
transformation. In our validation process we found that the results of our
expert system were particularly sensitive to the definition of size, (Baligh,
Burton, and Obel, 1996). The reason is that size in itself has an effect but
also that size turns out to be a moderator of the effect of other contingen-
cies.
From our information processing perspective, we find that a size
measurement is related to the number of people in the organization.
However, the information processing capacity depends on the type of people
in the organization. One problem with comparing various empirical results
is that some studies deal with a very narrow homogenous set of organiza-
tions (Carillo and Kopelman, 1991), while others have a much more
heterogenous set of organizations and do meta analyses (Miller, 1987).
In our search for an appropriate measure of size that meets our criteria
as being both practical and simple, we have used a measure that is the
number of employees adjusted for their level of professionalization. This
measures the information-processing demand on the organization design.
There are a number of reasons for this choice. First, organizations differ
and employ different types of people. We find the task they perform is less
important, but that the way they handle information and make decisions
is important. For a given level of size, the greater the degree of profes-
sionalization, the higher the possible level of decentralization, organiza-
tional complexity (Scott, 1992, p. 255), and formalized coordination
mechanisms (Scott, 1992, p. 253). This is also related to the span of control.
Despite the fact that the evidence is somewhat conflicting, it seems that the
more complex the tasks that a person can do, the smaller the span of
control of his superior (Scott, 1992, p. 255). Higher levels ofprofessionaliza-
tion thus seem to increase formalization between units, but at the same
time formalization within a unit may decrease. Therefore, if the span of
control is small, a smaller number of employees will lead to a higher level
of vertical differentiation, and we account for this in our compound
measure. This argument may be moderated by the use of information
technology as with some types of information technology comes is an
increasing demand for professionalization.
Skill capability or level of professionalization is measured on a five-point
scale. The proportion of employees that hold advanced degrees or have
many years of specialized training is measured on a scale from 1 to 5 with
o to 10 percent equal to 1, 11 to 20 percent equal to 2, 21 to 50 percent
equal to 3, 51 to 75 percent equal to 4 and 76 to 100 percent equal to 5. This
5.3 The Effects of Size and Skill Capability on Organizational Structure 153

scale is similar to the scales used by Hage (1965) and Miller, Glick, Wang,
and Huber (1991).
The adjusted-size measure is calculated as the number of employees
times the professionalization score. The adjusted-size measure is then
mapped into small, medium, and large according to the scale shown in
Table 5.1. The relationship between number of employees, professionali-
zation, and size is shown in Table 5.2.

Table 5.1. Size Categorization

Adjusted size Measure Size Category

< 100 Small

101 - 500 Medium with less impact
501 - 1000 Medium
1001 - 2000 Large with less impact
2001 < Large

Table 5.2. Thresholds for Categorizing the Adjusted-Size Measure,

Number of People Employed

Professionalization
Size measure 1 2 3 4 5

Small <100 <50 <33 <25 <20

Medium with less impact 101-500 51-250 34-166 26-125 21-100
Medium 505-1000 251-500 167-333 126-250 101-200

LarJ:~e with less impact 1001-2000 501-1000 334-666 251-500 201-400

Large 2001< 1001< 667< 501< 401<

An organization with fewer than twenty people will always be categorized

as small. One with twenty to 100 employees will be categorized as small or
medium with less impact. This means that if an organization has between
twenty-one and 100 employees with a high degree ofprofessionalization, it
is a relatively small organization but without the restrictions on the
organizational structure that come from being very small. However, once
it passes 100 employees it is very likely that the organization cannot be
seen as a small organization but has some properties of a larger organi-
zation. This is supported by empirical research that shows that companies
with forty to 100 employees face transition problems from being a very
154 Chapter 5. Size and Skill Capabilities

small to being a larger organization (Boje, Madsen, Obel, and Sf/lrensen,

1990). The same transition issues become important when moving from a
medium to a large organization.
Returning to our introductory example, Benny Mortensen is rightly
concerned that an increase in number of employees-most of whom are
professionals-is a cause for concern; the institute's size does change from
small to a medium-size organization.
The size issue is not completely resolved by our adjusted measure, but
we have found that it corresponds well with the literature and has worked
well with respect to our validation process. The adjusted size fits the
measure mentioned by Robbins (1990); in his treatment of size and
discussion of technology, the size measure can be related to manufacturing
companies with employees with a low level of professionalization. This is
Woodward's (1965) mass-production organization. At the other end of the
scale, the organizations studied by Carillo and Kopelman (1991) were
metropolitan branches of a financial service company, a type of organiza-
tion supposedly staffed with employees with a high degree of professionali-
zation.
We have found that the size measure has worked well independently of
the type of organization: service or production, or labor intensive or less
labor intensive. It has worked well in different industries with different
kinds of technologies.

5.3.2 Size as a Contingency

Size affects the structure of the organization in at least two ways-in the
effect of size that directly can be attributed to the adjusted-size measure,
and in the effect of size on other contingencies. Only the first is be discussed
in this chapter. The second kind of effect was presented in Chapters 3 and
4 and also is discussed in subsequent chapters.
As illustrated in the first section of this chapter, the change in size can
have a significant effect on appropriate organizational structure. Simple
information arguments based on bounded rationality assumptions lead to
the conclusion that an increase in size would lead to a higher degree of
organizational complexity and a higher degree of decentralization. A
control-based argument suggests a need for higher degree offormalization.
However, there are qualifications. In owner-controlled organizations the
increase in size has a much lower effect than in other types of organiza-
tions. Pondy (1969) argued that a goal for owner-run organizations is to
maintain control, and structural changes that yield lesser control are not
likely to take place. Geeraerts (1984), studying 126 Dutch companies,
showed that the correlation between centralization, formalization, and
5.3 The Effects of Size and Skill Capability on Organizational Structure 155

complexity was much lower in owner-run companies than in professional-

run companies.
The effect of an increase in size seems also to be higher for small
organizations than for large organizations, leading to a nonlinear relation-
ship between size and structure. We incorporate this nonlinearity by an
adjusted-certainty factor. Size not only affects complexity, formalization,
and centralization, but also has an effect on configuration. Small organiza-
tions tend to be most efficient, if they have a simple or ad hoc configuration.
Additionally, for some configurations like the divisional, the organization
should not be small. With this general background, we consider the size
propositions.

5.3.2.1 Size Effects on Complexity

Proposition 5.1 states that a large, not public organization should have a
high degree of organizational complexity ( Table 5.3). Miller (1987) shows
twenty-seven empirical studies including his metaanalysis, which support
this proposition. As the size increases, specialization of task is both
observed and reasonable. The economics of specialization can be realized.
particularly when the increase in size is due to an increase in professionali-
zation, specialization should increase.

Table 5.3. Size Effects on Complexity, Propositions 5.1-5.4

5.1. If the organization is large and not public, then complexity should be
high.

5.2. If the organization is large and public, then complexity should be

medium.

5.3. If the organization is of medium size, then complexity should be

medium.

5.4. If the organization is small, then complexity should be low.

Proposition 5.2 modifies the first proposition for ownership-namely, public

organizations have only medium complexity. The relationship between size
and complexity seems to be less strong in public organizations than in
private organizations (Blau, 1970).
156 Chapter 5. Size and Skill Capabilities

Propositions 5.3 and 5.4 follow the general argument. As size decreases,
there is less complexity and less specification of tasks. These propositions
state that an increase in size increases the complexity of the organization.
The increase in specialization may be the case both with respect to line and
staff. Generally, as small organizations grow, so does their need for
specialized staffs. When the organization is small, marketing, sales, and
accounting functions may be done by the manager, perhaps with some help.
When the organization grows, so do the supporting activities leading to
marketing departments, accounting, and personnel departments. At the
line dimension, growth may introduce functional specialization-such as in
production. Both of these arguments lead to higher vertical and horizontal
differentiation and in general to higher organizational complexity.

5.3.2.2 Size Effects on Centralization

Perhaps the most intuitive proposition of all is that a large organization
should be decentralized. A straightforward argument follows from an
information processing view-namely, there is much information to process
in a large organization and the decentralization of decision making is
desirable, even necessary to avoid top management overload and delay. The
increase in information processing may come either from an increase in the
number of employees or from an increase in professionalization. Proposi-
tions 5.5 and 5.6 state that large organizations should be decentral-
ized-private organizations less so than those that are not private (Table
5.4).
Privately owned organizations run most efficiently if the owner is
actively involved in decision making. Goals and missions are then clear,
and activities are coordinated. The relationship is empirically supported by
Geeraerts (1984).
Propositions 5.7, 5.8, and 5.9 are similar in concept. Medium-size
organizations should have medium centralization, with privately owned
organization having higher centralization. Small organizations should more
likely be centralized.
Throughout, the argument is that larger organizations should be more
decentralized; privately owned organizations are less extreme than
nonprivate, but both follow the same pattern. However, Miller (1987) finds
the relationship between size and centralization is not clear and significant.
Yet from a normative view, it is a very reasonable proposition that follows
from our information processing point of view and the notion of bounded
rationality. As size increases, so does the demand for information process-
ing and with the individual's limited information processing capacity,
delegation can be one reasonable response.
5.3 The Effects of Size and Skill Capability on Organizational Structure 157

Table 5.4. Size Effects on Centralization, Propositions 5.5-5.6

5.5. If the organization is large and not private, then centralization should
be low.

5.6. If the organization is a large private organization, then centralization

may have to be high but it also could be medium.

5.7. If the organization is of medium size and private, then centralization

may have to be high but it also could be medium.

5.B. If the organization is of medium size and not private, then centraliza-
tion should be medium but it also may be high.

5.9. If the organization is small, then it is most likely that it should have
high centralization but it also could be medium.

The argument is also related to the previous propositions on complexity.

Size increases complexity and thus by itself creates greater information-
processing demands. With the formation of specialized units, it may be
virtually impossible for management to be on top of every detail. If
management wants to make all decisions, these are delayed and not
coordinated, and problems arise. Then he or she has to delegate, and
decentralization increases.

5.3.2.3 Size Effects on Formalization

The relationship between size and formalization seems to be straight-

forward. Miller's (1987) metaanalysis supports Propositions 5.10, 5.11, and
5.12 (Table 5.5). Beginning at the end, there are two arguments for
Proposition 5.12. As size increases, decentralization increases, and
management loses control. An increase in formalization will enhance
control and ensure coordination. Additionally, as size increases, it is more
likely that the same activity has to be carried out more than once. It may
therefore be cost efficient to find the best way to carry out the activity and
then write a rule. Formalization thus decreases the amount of information
to be processed, overcoming the increased need due to the increase in size.
Propositions 5.10 and 5.11 suggest that the formalization should
decrease for smaller organizations. Here a cost argument explains the logic.
158 Chapter 5. Size and Skill Capabilities

Small organizations can easily be coordinated by a centralized decision

making. The cost of introducing formalization does not payoff.

Table 5.5. Size Effects on Formalization, Proposition 5.10-5.14

5.lD. If the organization is small, then formalization should be low.

5.11. If the organization is of medium size, then formalization should be

medium.

5.12. If the organization is large, then formalization should be high.

5.13. If formalization is high and professionalization is high, then formal-

ization between units should be high and formalization within units
should be medium.

5.14. Ifformalization is medium and professionalization is high, then for-

malization between units may be medium or high, and formalization
within units may be medium or low.

However, it is not enough to increase formalization when needed; the

formalization has to be the right type. If the rule is wrong, it may be better
to be without a rule.
Propositions 5.13 and 5.14 moderate the formalization effect for
professionalization. Professional organizations are less formalized at the
task level but remain equally formalized between organizational units.
Generally, formalization increases standardization, which can also be
obtained via professionalization. Professionals are trained to behave in a
standardized way; certified nurses and accountants, for example, are
trained to care for patients and prepare balance sheets, respectively. From
an information-processing point of view, professionals with skills can
process more information than less skilled employees.

5.3.2.4 Size Effects on Configuration

It generally is agreed that ifthe organization is small, a simple structure
is appropriate. In Mintzberg's (1979) discussion of his five configurations,
the only configuration that fits the small organization is the simple one.
From an information processing point of view, the simple organization is
the least costly configuration that can handle the information processing
5.3 The Effects of Size and Skill Capability on Organizational Structure 159

requirements of the small organization. Proposition 5.15 is supported by

the study by Geeraerts (1984).
Proposition 5.16 is widely agreed on (Daft, 1992, Mintzberg, 1979;
Robbins, 1990), as it follows the notion that an organization needs a
minimum size to have a more complex organizational configuration. The
size has to cause a minimum information-processing demand to make the
more costly complex configurations appropriate. Propositions 5.15 and 5.16
are consistent with ideas about the evolution of an organization. The
organization starts small, then grows by exploiting its markets, and finally,
continues to grow via diversification. In that process the configuration
changes from a simple configuration to a functional configuration and ends
with a multidivisional configuration in the last stage.

Table 5.6. Size Effects on Configuration, Propositions 5.15-5.16

5.15. If the organization is small, then it should be either an ad hoc or a

simple configuration.

5.16. If the organization is large or of medium size, then the organization

can be either a matrix, functional, divisional, machine bureaucracy,
or a professional bureaucracy.

The simple configuration with centralized decision making and loose

organization is an appropriate organizational design, when the organization
is small, young, and in its formative stage. It can respond quickly and foster
innovation usually centered with the founder. When the organization
obtains success and starts to grow, it usually moves into some kind of mass
production. Formalization and specialization are necessary to obtain
economies of scale and consistent level of quality. When local markets are
exploited, then the organization starts to diversify to other markets. This
forms the basis for divisionalizing the sales department. The next phase is
reached when diversification is based on products. When that happens, the
conditions for multidivisional configurations are satisfied. Throughout this
process the organization grows in size, and organizational changes are
made to meet the demand for information processing capacity.
160 Chapter 5. Size and Skill Capabilities

5.4 Managing Size

The main proposition in Chapter 1 is that a fit between contingency factors
and organizational structure is necessary to obtain optimal performance.
If there is a misfit, either the contingency factor value can be changed or
the organizational structure can be changed.
How can size be changed? Most organizations have either an implicit or
explicit goal that the organization should grow. The goal is most often
stated in terms of profit or other financial terms but also can include opera-
tions. Generally, as was pointed out in Section 5.2, there is a high
correlation between financial size measures and the measure of size.
Organizations work hard to change their size.
The change in size can happen in an evolutionary way or can happen in
big jumps. As the Institute of Applied Computer Science example in Section
5.1 shows, a big contract can suddenly change a company's size. Size also
can increase by mergers and acquisitions. In Europe, there is a big merger
wave. Different kinds of mergers lead to an increase in size but also to a
number of other changes that affect organizational structure. This is not a
new story.
So far we have discussed only increases in size, but size also may
decrease. Due to retrenchments in various industries, many organizations
suddenly have shrunk. It happens for small organizations, and in some
cases they die during the process. It happens for large corporations as well.
IBM, UNISYS, and SEARS are examples of organizations that have
decreased their size.
Additionally, introduction of new information technology may reduce or
change the number of type of employees in an organization. Many
companies have tried to eliminate complete layers of middle management
due to a more automated and efficient information and production
technology. Modern technology have also made certain types of employees
obsolete-mostly people with low or no education-while increased demands
for new types of employees-mostly people with a high level of education or
training.
The effect on the organization by an increase in size is not necessarily
the opposite of the effect of a decrease (Weitzel and Jonsson, 1989), and the
management process is quite different. However, there are many controver-
sies about these differences (McKinley, 1993), and it is an area that is not
much researched.
It may be that the process of downsizing an organization is not the
reverse of growth from the point of view of organizational design-but
should it be? Most research has focused on descriptive theory and has not
had a normative point of view. From a normative point of view, there may
5.5 The Size of SAS 161

be no difference between the most effective and efficient organizations, but

from an organizational change point of view, the difference may be big.

5.5 The Size of SAS

When SAS was founded in 1954, it could fly to sixty-seven cities (Buraas,
1972, pp. 106-109). By 1990, SAS had eighty-five destinations and had
alliances with 291 destinations from Copenhagen, 181 destinations from
Oslo, and 254 destinations from Stockholm with no more than one stopover
(SAS, 1990, p. 24). SAS had 20.000 employees.
With the appointment of Jan Stenberg as SAS President and CEO in
April 1994, followed a new era in SAS's history. The overlying idea for
Stenberg's new strategy program, was a renewed focus on SAS's core
business: the airline. The goal was to create a cost-effective company. To
reach this goal, an extensive rationalization program was initialized,
aiming at a total cost reduction of SEK 2.9 billion. Shortly after his
appointment, 2,930 employees were fired, primarily pilots and cabin crew.
As discussed earlier, there are numerous size measures. Kimberly (1976)
suggested physical, fiscal, input, and output dimensions as well. Yet, in
many, if not most organizations, these measures are correlated with
number of employees, and an excellent surrogate measure for organiza-
tional size is number of employees. From an information processing view
having more employees can increase a company's capacity to process
information, but they can also create a demand for information processing
as more individuals must be coordinated and controlled. The later view
drives many of our arguments. For SAS, the large number of employees
creates a need for a large capacity to process information.
Briefly, the size ofSAS is large. By whatever measure, including number
of employees, it has always been large, and large size has certain implica-
tions for design. Propositions 5.1, 5.5, 5.12, and 5.16 indicate the fit
congruencies.
SAS - a large, non public corporation-should have high complexity, low
centralization, and high formalization. The large size is compatible with
numerous· configurations-all of which can be found within SAS and its
numerous subunits. In general, SAS has a good fit for its large size. It has
high complexity, high formalization. Despite recent tendencies to increase
centralization, the company is still relatively decentralized on an opera-
tional basis. The company also remains relatively high on complexity and
formalization. These are appropriate means of coordination.
162 Chapter 5. Size and Skill Capabilities

5.6 Summary
There are a number of size measures. The number of employees is the best
measure of size for designing the appropriate organization. In order to
consider the information processing requirements of the organization, this
measure should be modified for the professionalization level of the
organization-that is, greater professionalization yields effectively more
employees than nonprofessional.
The size imperative indicates that a large organization should have high
complexity, high formalization, and low centralization. A small organization
should have low complexity, low formalization, and high or medium
centralization. The size implication for the organizational configuration is
less definite. The size of the organization is an important determinant of
the appropriate organizational design.

Notes
1. This section is based on Burton, Minton, and Obel (1991).

References
Aldrich, H. E. 1972. "Technology and Organization Structure: Re-examination of the
Findings ofthe Aston Group." Administrative Science Quarterly, 26-43.
Anderson, T., and S. Warkov. 1961. "Organizational Size and Functional Complex-
ity: A Study of Administration in Hospitals." American Sociological Review,
23-28.
Baligh, Helmy H., Richard M. Burton, and B0rge Obel. 1996. "Organizational
Consultant: Creating a Usable Theory for Organizational Design." Manuscript.
Blau, P. M. 1970. "A Formal Theory of Differentiation in Organizations" American
Sociological Review, 201-218.
Blau, P. M., and R. A, Schoenherr. 1971. The Structure of Organizations. New York:
Basic Books.
Boje, Per, Tage Koed Madsen, B0rge Obel, and Christen S0rensen. 1990. De
erhvervsmressige konsekvenser og muligheder for Fyn set i lyset af den faste
Storebreltsforbindelse og EF's indre Marked. Odense, Denmark: Den Fynske
Fond For Erhvers0konomisk Forskning, (in Danish).
Brynjolfsson, Erik. 1994. "Information assets, technology, and organization."
Management Science, 40(12), 1645-1662.
References 163
Brynjolfsson, Erik, Thomas W. Malone, Vijay Gurbaxani, Ajit Kambil. 1994. "Does
information technology lead to smaller firms?" Management Science, 40(12),
1628-1644.
Buraas, Anders. 1972. Fly over Fly. Oslo, Norway: Gyldendal Norsk Forlag NS.
Burton, Richard M., John W. Minton, and BI/lrge Obel. 1991. "Organization Size and
Efficiency: An Information Processing View." The Scandinavian Journal of
Management, 7(2), 79-93.
Caplow, T. 1957 "Organizational Size." Administrative Science Quarterly, 484-505.
Carillo, Phoebe, M., and Richard E. Kopelman. 1991. "Organization Structure and
Productivity." Group and Organization Studies, 16(1),44-59.
Child, J. 1973 "Parkinson's Progress: Accounting for the Number of Specialists in
Organizations." Administrative Science Quarterly, 328-346.
Child, J. 1982. "Discussion Note: Divisionalization and Size: A Comment on the
Donaldson! Grinyer Debate." Organization Studies, 251-353.
Daft, Richard L. 1992. Organization Theory and Design (4th ed.). St Paul, MN: West
(2nd ed., 1986, 3rd ed., 1989).
Donaldson, L. 1982. "Divisionalization and Size: A Theoretical and Empirical
Critique." Organization Studies, 321-337.
Galbraith, J. R. 1976. "Organization Design: An Information Processing View."
Interfaces, 28-36.
Geeraerts, G. 1984. "The Effect of Ownership on the Organization Structure in
Small Firms." Administrative Science Quarterly, 29, 232-37.
Grinyer, P. H. 1982 "Discussion Note: Divisionalization and Size: A Rejoinder."
Organization Studies, 339-350.
Grinyer, P. H., and M. Ardekani, 1980. "Dimensions of Organizational Structure:
a Critical Replication" Academy of Management Journal, 405-421.
Hage, Jerald. 1965. "An Axiomatic Theory of Organizations. "Administrative Science
Quarterly, 10,366-76.
Hall, R. H., J. E. Haas, and Johnson, N. J. 1967. "Organizational Size, Complexity
and Formalization." American Sociological Review, 903-912.
Hawley, A., W. Boland, and M. Boland. 1965. "Population Size and Administration
in Institutions of Higher Education." American Sociological Review, 252-255.
Hickson, D. J., D. S. Pugh, and D. C. Pheysey. 1969. "Operations Technology and
Organization Structure: An Empirical Appraisal." Administrative Science
Quarterly, 14,378-397.
Kimberly, J. R. 1976. "Organization Size and the Structuralist Perspective: A
Review, Critique and Proposal." Administrative Science Quarterly, 571-597.
Marsh, Robert M., and Hiroshi Mannari. 1989. "The Size Imperative? Longitudinal
Tests." Organizational Studies, 10(1),83-95.
Mayhew, B. H., R. L. Levinger, J. M McPherson, and T. F. James. 1972. "System
Size and Structural Differentiation in Formal Organizations: A Baseline
Generator for Two Major Theoretical Propositions." American Sociological
Review, 629-633.
McKinley, William. 1993. "Organizational Decline and Adaptation; Theoretical Con-
troversies." Organizational Science, 4(1), 1-9.
Meyer, M. W. 1972. "Size and the Structure of Organizations: A Casual Analysis."
American Sociological Review, 434-440.
164 Chapter 5. Size and Skill Capabilities

Miller, Chet, William H. Glick, Yau-de Wang, and George P. Huber. 1991.
"Understanding Technology-Structure Relationships: Theory Development and
Meta-analytic Theory Testing." Academy of Management Journal, 34(2),
370-399.
Miller, George A. 1987. "Meta-analyses and the Culture Free Hypothesis."
Organizational Studies, 8(4), 309-326.
Mintzberg, Henry. 1979. The Structuring of Organizations. Englewood Cliffs, NJ:
Prentice-Hall.
Pondy, Louis R. 1969. "Effects of Size, Complexity, and Ownership on Administra-
tive Intensity." Administrative Science Quarterly, 14,47-60.
Pugh, D. S., D. J. Hickson, C. R. Hinings, and C. Turner. 1969. "The Context of
Organization Structures." Administrative Science Quarterly, 91-114.
Robbins, Stephen P. 1990. Organization Theory, Structure, Design, and Application.
Englewood Cliffs, NJ: Prentice-Hall.
Robey, D. 1982 Designing Organizations: A Macro Perspective. Homewood, IL: Irwin.
SAS. 1960-1991. SAS Annual Reports. Stockholm: SAS.
SAS. 1994. SAS Annual Reports. Stockholm: SAS.
SAS. 1995. Annual Report. Stockholm: SAS.
SAS. 1996. MiljfJregnskab. Stockholm: SAS
SAS. 1996. Annual Report. Stockholm: SAS
SAS. 1996. De ffJrste 50 ar. Stockholm: SAS
SAS. 1996. ScandinatianAirlines 1946-1996. Stockholm: SAS
SAS. 1997. Delarsrapport 1: 1997. Stockholm: SAS
Scott, W. Richard. 1992. Organizations, Rational, Natural, and Open Systems.
Englewood Cliffs, NJ: Prentice-Hall.
Slater, R. 0.1985. "Organization Size and Differentiation." In S. B. Bacharach and
S. M. Mitchell (Eds.), Research in the Sociology of Organizations (vol. 4) (pp.
127-180). Greenwich, CT: JAI Press.
Smith, Ken G., James P. Guthrie, and Ming-Jer Chen. 1989. "Strategy, Size and
Performance." Organization Studies, 10(1),63-81.
Spencer, H. 1898. Principles of Sociology. New York: Appleton.
Starbuck, W. H. 1965. "Organizational Growth and Development." In J. G. March
(Ed.), Handbook of Organizations (pp. 451-433). Chicago, IL: Rand McNally,
Weitzel, William, and Ellen Jonsson. 1989. "Decline in Organizations: A Literature
Integration and Extension." Administrative Science Quarterly, 34, 91-109.
Williamson, O. E. 1975. Markets and Hierarchies: Analysis and Antitrust Implica-
tions. New York: Free Press.
Woodward, Joan. 1965. Industrial Organization, Theory and Practice. Oxford:
Oxford University Press.
CHAPTER 6

The Environment

6.1 Introduction
On Monday, January 13, 1992, the Danish business newspaper B0rsen
published a story about Samsonite's new distribution system for its
European market. Samsonite produces luggage in Belgium for the
European market. In each European country it has sold its products
through a national company that had exclusive rights to import Sam-
sonite's products. The national company then sold Samsonite products to
stores. In Denmark, the firm Bon Gout had held the contract with
Samsonite for twenty-three years, but Samsonite canceled the contract to
sell directly to stores from its headquarters in Belgium. It developed an on-
line order system that enables it to sell to all countries in the European
Union from Belgium, and allow it to take advantage of changes within the
new EU single market.
For both Samsonite and Bon Gout the environment is changing.
Deregulation within the EU led Samsonite to change the way it sells its
products. It is likely that its headquarters in Belgium has been reorganized
to fit both the new information technology and a new type of customers. For
Bon Gout a well-established relationship with a supplier has changed to a
hostile environment. Not only has Samsonite canceled the exclusive-rights
contract, but it also has stopped supplying to Bon Gout and is working hard
to transfer individual stores to the on-line purchasing system. If Bon Gout
does not choose a strategy and organizational structure that copes with this
new situation, it may realize severe problems. For both companies the
environment has a great effect on strategy and organizational structure.
The changes in the EU are one set of deregulations. Deregulation is a
fact for many industries throughout Europe and North America. Tradition-
ally, airlines, railroads, telecommunications, and banking, among other
industries, have been closely regulated. The deregulation movement trans-
cends national boundaries and political ideologies. The nature of the
regulatory environment has changed and that change calls for adopting
new business strategies and new organizational configurations and
capitalizing on new opportunities.
A basic thesis is that the change in regulation requires new business
strategies, which in turn require different organizational configurations
166 Chapter 6. The Environment

than were efficacious in a regulated environment. In a word, deregulation

requires customer-oriented strategies and an organizational configuration
that can respond efficiently to customer preferences in a timely manner.
Such innovative requisites seem self-evident, yet there are numerous
pitfalls as well as windows of opportunity (Burton and Obel, 1986).
Deregulation implies changes in the environment. More generally, a
primary characteristic of the environment is degree of uncertainty, which
may arise in numerous ways. For Bon Gout uncertainty has changed and
increased. It used to have a stable relationship with its supplier and its
customers. Now the supply side has changed, but it also faces a new
competitor. Additionally, although the market for Samsonite luggage has
been fairly domestic, the new competitor has a more international profile.
Given the degree of uncertainty, one can then indicate an appropriate
internal organizational configuration, if the firm is to be effective and
efficient and, in the long run, viable.
Contingency theory (Lawrence and Lorsch, 1967), open systems theory
(Thompson, 1967), and strategy and structure arguments (Chandler, 1962),
among others, follow the common theme that there must be a fit between
the environment and the organization. Each theory or model has its own
features, and the implications of uncertainty suggest different specific
adjustments. Understanding and fitting with the environment require an
information processing capacity commensurate with the uncertainty in the
environment (Galbraith, 1973). Environment is very important for the
organization: despite numerous controversies, this much is agreed. Whether
it is the environmental imperative of contingency theory (Lawrence and
Lorsch, 1967, Duncan, 1979) or the environmental determinism of popula-
tion ecology (Hannan and Freeman, 1977), the organization's environment
must be considered. From a contingency theory viewpoint, the environment
must be reckoned with and adapted to. The environment-structure
imperative indicates that an organization should design its structure in
relation to its environment. Here Bon Gout should react and redesign its
structure. Even within this framework, there is considerable controversy.
What does one mean by environment? How does one describe and measure
the environment? With the environment defined, described, and measured,
what are the implications for the organization's structure, and how should
the organization operate? Besides theoretical interest, managers want
answers and recommendations about what they should do. In this chapter,
the goal is to examine these issues with an eye to developing design recom-
mendations.
6.2 Background and the Literature 167

6.2 Background and the Literature

The concept, description, and measurement of an organization's environ-
ment remains difficult and is a basis for continuing controversy (Milliken,
1987). Additionally, the effect on the organization is not fully agreed on
(Koberg and Vngson, 1987). We review the most important contributions
to our understanding of the environment-structure relationship.

6.2.1 Measures of the Environment

In the literature the environment has been described using many
descriptors and has been categorized in many different ways (Scott, 1992,
and Buchko, 1994). For the contingency theory of organization, one or two
variables related to an uncertainty measure have been the norm in most
empirical investigations. However, despite the fact that some measure of
uncertainty has been used, it is not clear whether they measure the same
thing or how they are related. Lawrence and Lorsch (1967) and Duncan
(1972) are classic examples and are discussed below.

2 3

Instability
(rate of change)

Unpredictability
Ignorance of data
and cause/effects
Uncertainty
Numbers of variables
(h omogenous· heterogeneous)

Complexity
Interdependence
of variables

Figure 6.1. Description ofthe Environment

Source: Lawrence, (1981).

Lawrence and Lorsch (1967), in their now classic study, employed an

environmental measure of uncertainty. Uncertainty was a perceptual
measurement ofthe clarity ofinformation, certainty of causal relationships,
168 Chapter 6. The Environment

and time spans of definitive feedback (Lawrence and Lorsch, 1967, p. 29).
Lawrence (1981, p. 316) (Figure 6.1) maps the instability (rate of change),
ignorance of data and cause and effects into unpredictability. He also maps
the number of variables, homogeneity or heterogeneity, and interdepen-
dency of variables into complexity. Unpredictability and complexity are
then mapped into uncertainty. These environmental mappings are thus
aggregate measure of more basic descriptors.
Duncan (1972) developed a two-dimensional environmental measure to
capture the notion of environmental uncertainty: environmental change or
dynamism (unstable to stable) and environmental complexity (simple to
complex). The environmental change dimension encompasses the possibility
that the environment can change and can change in unpredictable ways.
This fits well our intuitive notion of uncertainty; we simply may not
know what will happen. Environmental complexity is another aspect of the
environment-namely, the number of elements in the environment that are
important to the organization. Duncan's two measures of environmental
uncertainty are clearly aggregate measures; each captures a different
aspect of the environment: change and complexity.
Downey, Hellriegel, and Slocum (1975) critiqued Duncan's environmen-
tal uncertainty measures and compared them with the measure used by
Lawrence and Lorsch. Downey, Hellriegel, and Slocum found problems with
Duncan's measure. In the original study, Duncan (1972) directly summed
the subscales without standardization, creating a biased weighing. The bias
made perceived complexity dominate and "only allow a minute weighing for
the ability to assign probabilities" (p. 628). Downey, Hellriegel, and Slocum
(1975) found, when subscales were standardized, that environmental
change was the more important aspect of environmental uncertainty.
Comparing the two measures of environmental uncertainty by Duncan and
Lawrence and Lorsch, Downey, Hellriegel, and Slocum conclude that the
measures actually were two different instruments. Downey, Hellriegel, and
Slocum also investigated the Lawrence and Lorsch measure and found like
Tosi, Aldag, and Stoney (1973), the uncertainty measure deficient in terms
of validity and reliability.
The importance of the above discussion becomes clear when we discuss
the environment and structure relationship because Duncan and Lawrence
and Lorsch obtain different results with respect to the environment and
structure relationship. We argue that the difference is a result of their
different uncertainty instruments.
Criticism aside, the two measures, environmental change and environ-
mental complexity, do capture important dimensions of the environment.
Despite methodological criticisms of these classic studies, the organiza-
tion's environment remains an important contingency. Downey, Hellriegel,
and Slocum (1975, p. 628) conclude that "Beyond the methodological
6.2 Background and the Literature 169

adequacy of specific research instruments, the inconsistent results obtained

in this study between Lawrence and Lorsch's and Duncan's uncertainty
measures raise even more serious questions. Uncertainty concepts as
presently used in organizational theory involve much ambiguity. This does
not mean all contingency theory need be restricted to one meaning for
uncertainty. Moreover, it does not mean that contingency theory must wait
for the development of the one meaning of uncertainty."
The research developments of the last few years confirm their assess-
ment. We have not settled on the measure of environment uncertainty.
Indeed, there are many others. Contingency theory continues to be a
well-accepted paradigm where the environment, however measured, is
important. Other environmental measures have been created for particular
purposes.
Bourgeois and Eisenhardt (1988, p. 816) defined a high-velocity
environment as one where "changes in demand, competition, and technol-
ogy are so rapid and discontinuous that information is often inaccurate,
unavailable, or obsolete." This definition incorporates both the reality of the
environment (rapid, discontinuous) and what is known (accurate, availabil-
ity, timely) about the environment. In other frameworks, high velocity could
be characterized as dynamic and uncertain. Eisenhardt (1989) summarized
the implications for management. High-velocity environments call for fast
decisions. Fast decision-makers use more information, not less; they
develop more alternatives, not less. And further, fast decisions based on
this pattern lead to superior performance.
Tung (1979) used a three-dimensional typology to measure the environ-
ment: rate of change, complexity, and routineness. She showed that these
three measures could be mapped into a measure of perceived uncertainty.
Milliken (1987) also used a three-dimensional measure. He divided the
environmental uncertainty into state, effect, and response uncertainty.
State uncertainty means that one does not understand how components of
the environment might be changing. Effect uncertainty is the inability to
predict whether a change in the environment will have an effect on the
organization, while response uncertainty is the lack of understanding of
what response is appropriate for a given change in the environment.
Daft and Lengel (1986) in their analysis of the informational require-
ments in an organization used environmental uncertainty and equivocality.
Equivocality is the "existence of multiple and conflicting interpretations
about an organization." High equivocality means confusion and lack of
understanding. Equivocality is a "measure of the organization's ignorance
of whether a variable exists in the space" (Daft and Lengel, 1986, p. 567).
Much earlier, Ashby (1956) utilized variety as a concept to describe and
measure the environment and the capacity of an organism to adjust.
Variety is defined as "the number of distinct elements" (p. 126). The
170 Chapter 6. The Environment

environment can be described by a finite list of variables that can take on

discrete or continuous values depending on the particular variable. The size
of an organization's environment depends on the number of variables and
their possible values. An environment with four variables, each of two
values (yes, no) has 24 or sixteen possible states. An organization's
environment can be rather large. Ashby's notion of variety of the environ-
ment and the organization (or organism) leads to the famous law of
requisite variety: "only variety can destroy variety" (p. 207).
Kuhn (1986, p. 3) further interpreted the law of requisite variety:
"Environmental disturbances can be kept from causing performance
deviations only when the environment's variety is exceeded (or equalled) by
the firm's variety, which in turn, is exceeded (or equalled) by the manage-
ment's decision-making capacity, or variety. To block any disruptions
caused by the uncontrolled (i.e., the environment's) variables, then, the
controlled (i.e., the firm's) variables must encompass sufficient latitudes."
Intuitively, the organization must be able to adjust to its environment
in its many possible states. Although very appealing, the law of requisite
variety requires careful attention to operationalize and to derive implica-
tions for organizational theory.
The problem of defining, describing, and measuring an environment is
difficult and remains contentious. In organizational theory, the tendency
toward one-dimensional measures, like the Lawrence and Lorsch's and
Duncan's uncertainty measures, has been criticized. Indeed, it would be a
marvel to develop a one-dimensional measure for such a multidimensional
reality. At the other extreme, Ashby's variety measure is quite unwieldy.
Somewhere, there is a balance of simplicity and complexity that meets
our purpose for a design-oriented contingency theory. As we argue in the
next section, it is a small list of three environmental characteristics or
attributes; one environmental measure is not enough. However, we do not
know how to deal with a large number of characteristics. Lawrence (1981,
p. 317) called for such a balance: "The key problem is whether we can treat
this idea as a unitary dimension of environment for the sake of simplicity,
or whether we must treat it as a set of separate but related factors for the
sake of accuracy. The factors and terms, that have been discussed as
related to uncertainty, are complexity, number of variables, predictability,
ignorance of relevant data and cause-and-effect relations, interdependence
of variables, and rate of change."
So far we have treated the environment as a single body. Lawrence and
Lorsch (1967) recognized that different parts of the organization may be
facing different types of environments. They focused on the market, science,
and technoeconomic environmental sectors. Fink, Jenks, and Willits (1983)
discuss multiple environments of economics, technological, political-legal,
social-cultural, and external physical environments. Fink, Jenks, and
6.2 Background and the Literature 171
Willits (1983, pp. 278-284) discuss static attributes of complexity, routine-
ness, interconnectedness, and remoteness, as well as dynamic attributes of
rate of change and predictability of change. They present contingency
relations between these environmental factors and the organization's
structure, goals, and rewards, among others (pp. 284-290). It is a very
elaborate set of contingency relations concerning the design of the
organization.
The two environmental dimensions of complexity and uncertainty
(Robbins, 1990, p. 219) and complexity and change (Daft, 1992, p. 52) are
popular textbook concepts. They remain intuitively appealing, if not
methodologically rigorous, and can be operationalized as important con-
tingencies for the organization's structure. The debate on the proper
measure of the environment continues, however (Buchko, 1994).
One additional measure has to be mentioned here-degree of environ-
mental hostility. Bon Gout was threatened by the actions of Samsonite and
had to react appropriately.
Hostility is not uncertainty. Extreme hostility implies that someone or
something threatens the existence of the organization. Hostile environ-
ments are characterized by precarious settings, intense competition, harsh,
overwhelming business climates, and the relative lack of exploitable
opportunities. In 1990, a Danish hairdresser announced publicly that use
of the combined shampoo and conditioner Wash-and-Go caused people to
lose their hair. The news media jumped on the story, and sales on the
Danish market dropped from a 24 percent market share to 3 percent
basically over-night. Wash-and-Go is sold worldwide, so the situation was
a real threat to Procter & Gamble. Koed Madsen and M,dler Jensen (1992)
and M011er Jensen (1993) categorized the environmental situation for
Procter & Gamble in Denmark as a hostile one. Low hostility implies
benign environment (Covin and Slevin, 1989). As we argue later, the
organizational proper response to high hostility is different from its
response to high uncertainty (Robbins, 1990).

6.2.2 The Environment-Structure Relationship

Environmental-organizational relations are dominated by three perspec-
tives: the environment is dominant, there must be a good fit between the
environment and the organization, and there are organizational principles
that operate for any environment. Population ecology in organizational
theory is an example of an environment-dominant argument. The environ-
mental-organizational fit thesis captures the contingency theory, open
systems, and other approaches. We will review briefly the first two themes
172 Chapter 6. The Environment

and then concentrate on implementation of the environmental-organiza-

tional fit themes.
Population ecology considers the environment as dominant. It is the
"environment which optimizes" (Hannan and Freeman, 1977, p. 939).
Through competition, the environment separates the successes from the
failures (Pfeffer, 1982, p. 186). The organization must be linked with its
environment in order to survive. In the Bon Gout case, EU deregulation
may imply that it may not be feasible for a luggage manufacturer to sell
through national agents that have exclusive rights. The environment may
optimize so there is no room for local agents. The environment, with its
changes in information technology and regulation, determines who will
survive and prosper and who will die. It is less clear how one can develop
approaches for viability-if it is, indeed, possible. The environment selects;
whether organizations can adapt consciously is an open question.
Yet, the organization has two choices of design: specialism and
generalism. Hannan and Freeman (p.953) offer insight on the choice, "when
the environment changes rapidly among quite different states, the cost of
generalism is high. Since the demands in the different states are dissimilar,
considerable structural management is required of generalists. But since
the environment changes rapidly, these organizations will spend most of
their time and energies adjusting structures. It is apparently better under
such conditions to adopt a specialized structure and 'ride out' the adverse
environments." Of course, this choice depends upon a number of factors:
how long will the "storm" last; how fast can the structural adjustment be
made and what is its cost; and what is the opportunity cost of having a
nonoptimal structure. Some organizations will fail in the storm. Population
ecology focuses on selection by the environment, where the contingency
approach focuses on the adaptation of the organization to the environment.
Environmental-organizational relations and necessity for fit are explicit
in much of modem organizational theory. The organization's structure is
contingent on the environment and the technology, where the organiza-
tion's size is also a consideration. Lawrence and Lorsch (1967) argue that
increased environmental uncertainty leads to increased organizational
differentiation, making organizational integration more difficult. They
define differentiation to mean that the organization has departments that
are different in both tasks and orientation. This relates to horizontal
specialization and our measure of organizational complexity. Lawrence and
Lorsch studied three well-defined industries that they categorized as
ranging from low to high uncertainty. In their study Lawrence and Lorsch
found that increased uncertainty in the environment required increased
differentiation in the organizational structure for the organization to be
efficient. That then requires increased integration to make the different
departments work in cooperation. Integration devices typically include
6.2 Background and the Literature 173

rules and procedures, configurational plans, the authority ofthe hierarchy,

and decision-making committees.
Duncan (1979) begins with an environmental classification along a
static-dynamic dimension and a simple-complex dimension. Functional,
divisional (decentralization), and matrix models are then analyzed for
strengths and weaknesses. Finally, the environmental-organizational
relations for fit are developed. For example, a simple static environment
calls for a functional organization. Galbraith (1973) focuses on the firm's
information processing choices as alternatives to cope with environmental
uncertainty (Chapter 2). In this stream of research, environmental
uncertainty requires the firm to adapt.
As the uncertainty of the environment increases, so does the demand for
information processing. A company in a very uncertain competitive market
situation has to monitor the market and make quick changes, and doing
that right requires a large amount of information. An alternative approach
is to insulate the organization from its uncertain environment. The classic
approach is to create a buffer inventory to absorb environmental shocks.
For example, a firm may keep extra raw materials inventories on hand,
decouple its sequential processes with intermediate inventories, and use
finished goods inventories to coordinate with customers. More generally,
slack resources (Cyert and March, 1963) can be created in excessive cash
balances, number of employees, and capital equipment. Buffer inventories
and slack resources reduce the need to process appropriate information
quickly and efficiently.
The open systems theory also emphasizes the importance of environ-
mental uncertainty on the organization's structure (Thompson, 1967,
proposition 6.2C, p. 72): "When the range of task-environment is large or
unpredictable, the responsible organizational component must achieve the
necessary adaptation by monitoring the environment and planning re-
sponses, and this calls for localized units." Localized units are similar to the
idea of self-contained units in Galbraith (1973) information processing
framework.
Burns and Stalker (1961) developed two types of organizations-the
mechanistic and the organic. The mechanistic structure was characterized
by high complexity, formalization, and centralization, while the organic
structure was characterized by low formalization and decentralization.
Additionally, the task definitions were flexible, leading to low differentia-
tion and low complexity. Burns and Stalker believed that the organic
structure was needed to be efficient ifthe environment was turbulent. Ifthe
environment was stable, then the mechanistic structure was the best.
Compared to the views of Burns and Stalker, Lawrence and Lorsch give
almost opposite advice on how organizations should respond to environ-
mental uncertainty. Burns and Stalker recommend that the proper
174 Chapter 6. The Environment

response to uncertainty is low organizational complexity, while Lawrence

and Lorsch recommend a high degree of horizontal differentiation-a high
organizational complexity. The empirical studies of the environmental-
organizational structure relationships have resulted in mixed results
(Tung, 1979; Koberg and Ungson, 1987). Miller (1992) finds that "attempts
to achieve fit with the environment uncertainty can prevent or destroy
internal complementaries." Basically, he reaches his conclusion by
combining various results from the literature.
A careful review of the many studies and arguments reveals that many
of the different results may be the result of different use of the term
environmental uncertainty and that describing the environment using a
single measure may be too simple: a proper balance has to be obtained, but
there are many pitfalls. Jurkovich (1974) developed five measures of the
environment, each measured on a high-low scale that leads to sixty-four
different environmental categories. Such a measure is both too simple and
too complex. It is too simple because a high-low score usually is too gross,
and it may be too complex with respect to the number of dimensions. In the
next section we put the various pieces together and develop a three-
dimensional measure of the state of the environment for design purposes.

6.3 Describing the Environment: Equi-

vocality, Uncertainty, Complexity,
and Hostility
Environment has been defined, described, and measured in many
ways-from unidimensional uncertainty to multidimensional measures. We
search for a balance: the definition should be reasonably precise and useful
but not totally descriptive and detailed beyond utility. The goal is to specify
the environment in measurement terms that can be of help in the design
of an organization-that is, to assess environment in a contingency design
framework.
We propose that an organization's environment be described and
measured in four dimensions in a four-variable list of characteristics or
attributes. Those environmental characteristics are: equivocality, uncer-
tainty, environmental complexity, and hostility. Equivocality is the
"existence of multiple and conflicting interpretations" (Daft and Lengel,
1986). Uncertainty is related to earlier discussed measures of uncertainty,
but is more restrictive. It is uncertainty of all specific parameter values
such as prices, cost and so on. Environmental Complexity is similar to
earlier measures ofthe number offactors in the environment that affect the
6.3 Describing the Environment: Equivocality, Uncertainty,Complexity, 175
and Hostility

organization. Hostility is the level of competition and how malevolent the

environment is.
Equivocality means confusion and lack of understanding. Equivocality
is a "measure of the organization's ignorance of whether a variable exists
in the space" (Daft and Lengel, 1986, p. 567). We use the notion that
equivocality is the organization's ignorance about whether the variable
exists in the environment. That is, the variable is ill specified and unknown
to the organization, or its importance for the organization is not known.
Referring back to figure 6.1 and Lawrence and Lorsch's environmental
scheme, equivocality incorporates the "ignorance of data and cause/effects"
to the ignorance about what the important variables may be. Of course, the
organization must have some information about its environment for
equivocality to exist-namely, the organization must be aware of its
ignorance. Equivocality is related to the concept of the agenda of the
organization (Arrow, 1974). Ifthe agenda of the organization is known and
set, then equivocality is low. If the agenda is unknown or not set, then
equivocality is high.
Deregulation, as discussed previously, may introduce equivocality.
Under regulation, the dimensions on which to compete are known but after
deregulation it may not be clear how competition will evolve. EU deregula-
tion and the actions taken by Samsonite increased the equivocality of Bon
Gout's environment. In the telecommunication industry, competition has
been based on price and service, and competitors have been other telecom-
munication firms. The entrance of digital switches enabled the computer
industry to enter the telecommunications market. In Canada the telecom-
munication industry also has been challenged by TV cable companies, and
it is not clear what the multimedia business will do in the future and how
wireless communication will evolve. Denmark is experiencing similar
changes. The degree of equivocality has increased. During the last decade
the telecommunication industry has faced various degrees of equivocality,
and it remains rather high.
The equivocality measure is related to many ofthe measures discussed
above. For example, the routineness measure used in Tung (1979) and
Jurkovich (1974) and the effect uncertainty in Milliken (1987) are closely
related to our equivocality measure. As Downey, Hellriegel, and Slocum
(1975) have pointed out, it is also closely related to Duncan's measure of
uncertainty. Additionally, Burns and Stalker's measure of turbulence may
be closer to equivocality than to uncertainty.
Uncertainty is our second environmental variable. Uncertainty is not
knowing the value of an environmental variable or descriptors. Before the
move by Samsonite, Bon Gout knew what suitcases to order and sell. But
it might not have known how many it could sell at a given price. Likewise
176 Chapter 6. The Environment

a telecommunication company may not know how many calls it has to

handle at certain points in time. In both situations the companies face
uncertainty. They do not know the exact value of a particular variable that
is important to the company. This definition is close to the common sense
definition and resembles the Lawrence and Lorsch (1967) definition of
uncertainty, but more directly to the instability of the environment.
Equivocality and uncertainty are conceptually distinct, yet the
distinction is not always easy. Basically uncertainty requires that at least
in principle a probability measure can be specified. Additionally, uncer-
tainty is in most cases related to issues that the organization has experi-
enced previously. Fluctuation in demand of the organization's goods and
services is a good example.
When something very new may happen-new regulations, new technol-
ogy, and so on, this is related to equivocality. In this case, a probability
measure may not be the right way to describe the lack of certain informa-
tion. Equivocality is related to something the organization has not
experienced before.
The environmental complexity is the number of variables in the
environment and their interdependency-Bon Gout sales, for example,
depend not only on prices but also on fashion. This is a more complex
environment than one in which sales are dependent only on price. Sales
also may depend on actions taken by competitors, and the variables may be
related. The competitors' actions may depend on the price asked by Bon
Gout. Managing interdependent variables is more complex than managing
independent variables. Therefore, an interdependent environment is more
complex than one in which variables vary independently. The environmen-
tal complexity measure has been used by a number of authors including
Duncan (1972), Tung (1979), and Jurkovich (1974).
Hostility is a measure of how benign or malevolent the environment is.
Hostility can vary from a supporting environment to one that is predatory
and out to destroy the organization. Natural causes, e.g. hurricane or flood,
can be the cause ofthis extreme level of hostility, or it can also result from
the actions of others, e.g. a hostile takeover or sabotage. Bon Gout had been
operating in a relatively benign environment, but the hostility quickly
increased and it became threatened for its very existence; it lost its product
and image of quality which threatened its very existence. For high hostility,
someone or something is manipulating the environment to deteriorate the
performance.
The environment can be more precisely described as a list of descriptors
or variables {Xl' X 2 , '" Xn} and its effect on the organization can be modeled
by f(Xl' X 2 , ' " Xn ). Then we say the following:

• Complexity is higher as n increases and the l( 's are interdependent.

6.8 Describing the Environment: Equivocality, Uncertainty, Complexity, 177
and Hostility

• Uncertainty increases as the value of a variable ~ (or variables) is less

well known (that is, the variance of Xi increases).
• Equivocality increases as the organization is more ignorant about what
are the important variables in its environment, (that is, the organization
does not know on which set of variables the f-function is defined nor the
function itself).
• Hostility increases as a change in one of the variables changes to
threaten the organization's performance and perhaps its existence; it is
a discontinuity.

The four measures were chosen because they can be related to a vast
literature on empirical studies as well as fit well with our information-
processing view of the organization. The greater the environmental com-
plexity, the greater the information processing demand on the organization.
The organization has to monitor more issues and assess the effect on the
organization.
Increased uncertainty also increases the demand for information-
processing capacity, but in a different way. While increased complexity
increases the number ofvariables that the organization has to monitor and
react to, uncertainty relates to the frequency that the variables have to be
monitored and adapted to.
Equivocality relates to the missing information that may hold important
contingencies.
Hostility captures whether the environment is supportive or whether the
environmental is threatening and potentially destructive.
An increase in each of the environmental dimensions increases the
demand for information processing capacity but in different ways. Each
environmental characteristic presents a different information challenge, as
summarized in Figure 6.2. Greater environmental complexity increases the
amount of information to process, as there are more issues of importance
to the organization. In a competitive market, price may be the only impor-
tant environment parameter; in a fashion market, many parameters are
important. Greater uncertainty requires greater capability to forecast or
adjust to the changing environment. This does not necessarily increase the
amount of information, but does require a different organizational
response. It must either project what will happen or adjust quickly to the
environment. The first is forecasting; the second is adapting to feedback.
Many organizations use a combination of both: for example, a firm with
uncertain sales will forecast and also adjust quickly to actual sales. Greater
equivocality requires broader scanning of the environment for heretofore
unknown and unimportant environmental parameters. Product and service
quality, as defined by the customer, continues to generate new parameters
of importance. We frequently do not know what the customer is saying.
178 Chapter 6. The Environment

Environmental characteristics Information processing implication

Greater complexity More environmental parameters to

monitor

Parameters can take on a greater

Greater uncertainty
range of values and change quickly

Unknown issues to scan and attempt

Greater equivocality
to monitor

Greater hostility A change in relevance of what information

is important; new information emerges

Figure 6.2. Information Processing Implications of Environmental

Characteristics

The four environmental characteristics are general attributes, not the

detailed list of~ 's. It is a great simplification. Larger characteristic spaces
are possible. Environmental complexity could be split into two measures:
one is n, the number of variables, and another is their interdependency. Of
course, each bivariable interdependency or correlation could be a more
detailed list of characteristics. Although possible and more precise, there
is a cost. Smaller characteristics spaces are also possible. Environmental
uncertainty, as a perceptual measure, has been applied, as discussed above.
Two-dimensional measures include (1) uncertainty and complexity and (2)
uncertainty and equivocality. The rationale for an environmental descrip-
tion or list of characteristics must be justified by its application. Next, we
present environmental contingency propositions that rely on the four
environmental characteristics defined above.

The Environment as Five Forces1

The organization's environment can be described in many ways, using different frameworks.
Michael Porter's five forces model is a well known and widely applied in strategic analysis. It
focuses on the industry structure and the competitive nature of the environment: buyers,
suppliers, substitutes, potential entrants and rivalry among existing firms. More specifically, the
firm's competitive situation is analyzed as:
6.8 Describing the Environment: Equivocality, Uncertainty,Complexity, 179
and Hostility

• bargaining power of buyers,

• bargaining power of suppliers,
• threat of substitute products or services,
• threat of new entrants
and,
• rivalry among existing firms.

The buyers' power increases as buyers purchase large quantities, purchase standardized
products for which they are ready substitutes and can switch easily, etc. and generally, can
then demand lower costs, higher quality products and better service.
The suppliers' power is the other side of the firm. Supplier power is high when the suppliers
can demand higher prices and deliver lower quality or service. Basically, when the firm has
limited alternatives or only alternatives which are very costly. Suppliers have greater power
when there are few, if any substitutes, the product or service is unique or the customer is not
very important to the supplier.
The threat of substitutes by other firms is increased when other products or services are
ready substitutes and are functionally equivalent.
The threat of new entrants is high when the cost of entry is low and the market is attractive
in price and/or size. This threat is less when there are economies of scale, product differentia-
tion, high capital entry costs, unique and specific industry knowledge which is difficult to
transfer.
Greater rivalry among existing firms exists when products are similar, high fixed costs, slow
growth, overcapacity and only a very few can be winners in a game of survival.
The five forces describe the market structure and embed other dimensions as uncertainty
and risk and other environmental characteristics. These additional environmental measures
provide a richer understanding of the organization's environment. The five forces-four
environmental dimensions crosswalk matrix provides a way to develop more depth about the
environment. In the table below, the five forces are given down the left-hand side and our
environmental measures are shown along the top. The matrix entries are then the crosswalk
from the five forces of the industry to our environmental measures. Eg., the uncertainty in the
environment is composed of the uncertainty from each of the five forces; the uncertainty about
the threat of new entrants should be reflected in the uncertainty measure.

Uncertainty Equivocality Complexity Hostility

Bargaining of
power of buyers

Bargaining
power of
suppliers

Threat of substi-
tute products or
services

Threat of new
entrants

Rivalry among
existing firms
180 Chapter 6. The Environment

For Bon Gout, an industry structure five forces analysis would highlight the threat from the
supplier, Samsonite. Samsonite intended to forward integrate its distribution chain and
eliminate the local retailer. The other four forces must be considered, but their importance is
so great here. The buyers' have considerable power and can go to other stores and buy other
brands of luggage. Substitutes for luggage depend upon the scope; there are numerous
substitutes for suitcases to include bags, sacks, etc. New entrants are possible as the cost of
entry is relatively low in retailing. However, Bon GoUt and Samsonite have considerable
reputation. The rivalry depends upon the local situation for Bon Gout. There is ample supply
of luggage from around the world and ready availability in the local market; the rivalry is
reasonably intense.
To do the crosswalk from the five forces to the environmental dimensions for Bon Gout, we
would begin with the supplier threat and focus on the hostility dimension, ie, the supplier-
hostility cell. It is clearly high and perhaps, extreme. The hostility from other forces is minor in
comparison. Looking now to the organizational complexity, there are a large number of buyers
which can be differentiated by buyer group according to income, age, etc. The number of
suppliers may not be large, but there are more now than in the past. New entrants are possible.
And the rivalry suggests that there is price, advertising, promotions, etc. which help make the
market competition. Overall, the environment is rather complex and we judge the level to be
high. The environmental uncertainty is judged to be medium; the variation in the five forces may
be increasing, but is not high. We can put reasonable bounds around the uncertainty. The
equivocality is more difficult as it is essentially a measure of what we do not know. There may
be new entrants in the shadows, but we know the buyers and the substitutes for luggage. The
equivocality seems low. However, before Samsonite's surprise move, the equivocality was high,
but Bon Gout did not know it. That is the essence of equivocality.
As illustrated above, the crosswalk from the five forces to the environmental dimensions can
logical and systematic, but it cannot be reduced to an easy method or simple technique. There
is no formula as the influences of the five forces will be weighed for the particular situation. We
began with the supplier hostility and it is very important in this case, but this situation is not the
norm for most organizations. Further, even if we could find a formula, we want to think through
each situation and understand it fully.
Porter's five forces industry structure provides an alternative way to describe the environment
and a help to developing the environmental measures for our organizational design approach.

6.4 Environment as a Contingency

Factor
Our interest in describing the environment is to indicate the effect of the
environment on the organizational structure and design. The environment
as a contingency is well established (Lawrence and Lorsch, 1967; Duncan,
1979; Galbraith, 1973; among others). Yet there remains controversy about
the exact nature ofthose influences, as discussed above.
As we take our point of departure from the information processing point
of view of the organization, we try to sort out the controversies by using
information processing arguments to find the relationships between
contingency factors and organizational design parameters. In the previous
section we established how our four measures of the environment affect the
demand for information-processing capacity.
6.4 Environment as a Contingency Factor 181

The organizational design response is to create an organization with an

information processing capacity to match the demand or to change the
demand by manipulating the environment. There are numerous ways to
create such a match. If there are many repeated tasks, it is advantageous
to write a rule about how to perform the task the best way. If, however,
there is a great uncertainty about what to do, then it may not be possible
to write such a rule, but instead management needs to monitor the results
and create an incentive system that will lead the organization in the right
direction. The fit between the decision-making process and the control and
incentive system is, therefore, important.
In this section and the next, we develop from this perspective a number
of propositions, which are summarized in Tables 6.1, 6.2, 6.3, and 6.4.
These propositions relate the organizational environment to the organiza-
tional structure: formalization, complexity, centralization, configuration,
media richness, coordination and control, and incentives. The goal is to
recommend an appropriate design.
In Table 6.1, the environmental characteristics of equivocality, com-
plexity, and uncertainty are given a high or low value, yielding 23 or eight
environmental possibilities. In practice, a finer scale must be used as is the
case in Organizational Consultant. For the sake of simplicity, we limit
ourselves here to a high and low discussion. For each environmental situa-
tion, we give recommendations on formalization, organizational complexity,
and centralization. These propositions are derived and supported by
considerable research (Burns and Stalker, 1961; Tung, 1979; Lawrence and
Lorsch, 1967; and Duncan, 1979). Hostility propositions are included in
Table 6.2.

Table 6.1. Environmental Effects on Formalization, Organizational

Complexity, and Centralization

a. LOW EQUIVOCALITY
ENVIRONMENTAL
COMPLEXITY
Proposition 6.2 Proposition 6.4
High Formalization: high Formalization: medium
Organizational Organizational
complexity: medium complexity: high
Centralization: medium Centralization: low

Proposition 6.1 Proposition 6.3

Low Formalization: high Formalization: medium
Organizational Organizational
complexity: medium complexity: high
Centralization: high Centralization: medium
Low Hlgh
UNCERTAINTY
182 Chapter 6. The Environment

h. HIGH EQUIVOCALITY
ENVIRONMENTAL
COMPLEXITY

Proposition 6.8 Proposition 6.6

High Formalization: medium Formalization: low
Organizational Organizational
complexity: medium complexity: low
Centralization: low Centralization: low

Proposition 6.7 Proposition 6.5

Low Formalization: medium Formalization: low
Organizational Organizational
complexity: medium complexity: low
Centralization: high Centralization: high
Low High
UNCERTAINTY

Table 6.1 is divided into two parts: part a shows the complexity and
uncertainty effects for low equivocality; part b shows the complexity and
uncertainty effects for high equivocality.
In Proposition 6.1, the environment is quite simple: equivocality is low
so we know what is important. The low complexity indicates that only a few
variables describe the environment. Low uncertainty suggests that we
know what the value of each variable is-that is the variance is small. The
information processing requirements from the environment are small.
Returning to the Bon Gout example the corporation would be in this
environment, for example, if it sold only Samsonite luggage in Denmark
paying Samsonite in Danish currency at a fixed price and if it made an
agreement with the stores it serves that they would order and pay for
suitcases well in advance of Bon Gout's ordering from Samsonite.
The question is then how the organization should be structured: its
formalization, organizational complexity, and centralization. For this
environmental situation, formalization should be high. There are few
variables in the environment, all variables are known, and their values can
be predicted with a high degree of certainty. in such a situation using a
standard procedure is efficient (Burns and Stalker, 1961 Duncan, 1979).
High formalization indicates that the organization's procedures are usually
written down and codified. There are standard ways of doing things, and
these procedures are well known. For this situation, these procedural rules
mayor may not be complicated, and there can be many or few ofthem. Also
the organizational complexity (that is, vertical and horizontal differentia-
tion) should be medium. There is no need for an elaborate reporting and
control hierarchy, or high vertical differentiation but there may be a
requirement for work specialization and high horizontal differentiation
(Burns and Stalker, 1961). Horizontal specialization may be on the task
level, but very likely horizontal differentiation on the department level will
6.4 Environment as a Contingency Factor 183

be medium to low (Lawrence and Lorsch, 1967). The organizational

complexity recommendation will be modified by the size ofthe organization.
Small organizations generally tend to have lower organizational complexity
than large firms. Similar modifications should be made with respect to all
propositions in this chapter.
Centralization should be high for this situation. Top management has
the capacity to gather, interpret, process information and make decisions,
most of which are current operations decisions. There are few strategic
concerns. However, the recommended high centralization can be questioned
on two fronts. First, with high formalization, is high centralization an
overkill for such a simple situation? Clearly, there are tradeoff's between
formalization and centralization (Robbins, 1990, p. 112-113). So a balance
must be achieved. Second, for a large organization, centralized decision
making and control may be too demanding for top management, who simply
do not have the time to do everything required. Rather, top management
should develop more formalization and be more decentralized in its
operations. Overall, this simple environmental situation calls for high
formalization of procedures, and centralization also can be high if it does
not create an overkill of management or if the organization is not too large
to be centralized. Again, similar considerations have to be made for all the
propositions presented in this chapter.
We now consider an environmental variation in which the organizational
complexity is high, and equivocality and uncertainty remain low. This is the
upper left corner of Table 6.1 and is expressed in Table 6.2 as Proposition
6.2. For Bon Gout this would be its situation if it imported from many
suppliers and conditions remained the same as above.

Table 6.2. Environmental Effects on Formalization, Complexity and

Centralization, Propositions 6.1-6.9

6.1. If the environment has low equivocality, low complexity and low
uncertainty, then the formalization should be high, organizational
complexity should be medium, and centralization should be high.

6.2. If the environment has low equivocality, high complexity and low
uncertainty, then formalization should be high, organizational
complexity medium and centralization should be medium.

6.3. If the environment has low equivocality, low complexity and high
uncertainty, then formalization should be medium, organizational
complexity should be high, and centralization should be medium.
184 Chapter 6. The Environment

6.4. If the environment has low equivocality, high complexity and high
uncertainty, then formalization should medium, organizational
complexity high and centralization low.

6.5. If the environment has high equivocality, low complexity and high
uncertainty, then formalization should be low, organizational
complexity should be low and centralization should be high.

6.6. If the environment has high equivocality, high complexity and high
uncertainty, then formalization should be low, organizational
complexity should be low, and centralization should be low.

6.7. If the environment has high equivocality, low complexity and low
uncertainty, then formalization should be medium, organizational
complexity should be medium and centralization should be high.

6.B. If the environment has high equivocality, high complexity and low
uncertainty, then formalization should be medium, organizational
complexity should be medium and centralization should be low.

6.9. If the hostility is extreme, then formalization should be low, organiza-

tional complexity should be low, and centralization should be very
high. If the hostility is high, then the centralization and organiza-
tional complexity should be lowered from its level determined by other
factors.

Proposition 6.2 indicates the higher complexity decreases the recommended

centralization. The argument for high formalization is similar to that for
Proposition 6.1. With a high environmental complexity, the environment
has many important variables to consider; however, the variables are well
known and their values are relatively fixed as equivocality is low and
uncertainty is low. Here the argument for high formalization is even
stronger than for Proposition 6.1. It is efficient to develop standard
procedures to take care of the many variables and adjust operating
decisions as required. Bon Gout could develop a standard ordering and
delivery system that would apply to all suppliers, or it could tailor this
system to individual suppliers making sure they were handled correctly
each time. Through analysis and experience, formalized procedures can be
developed and codified. Procedures also can lead to well-conceived
coordinated activities within the organization.
6.4 Environment as a Contingency Factor 185

Here again, the organizational complexity should be medium; there is

no need for an elaborate hierarchy or work specialization (Lawrence and
Lorsch, 1967).
Centralization becomes problematic. There are arguments for low and
high centralization. Unless the organization is small, the high complexity
of the environment makes lower centralization the most efficient solution.
A bounded rationality argument suggests that a top manager cannot
monitor all the external factors. With a stable environment control of
operations is mail1tained through high formalization. We suggest that the
weight of the argument is to recommend medium centralization. Top
management can be overwhelmed with making enumerable decisions for
which formalized procedures are possible; this is particularly relevant for
the large organization. The risk is that top management will become
overloaded and the timeliness of actions may suffer (Duncan, 1979).
Returning to Table 6.1, in the lower right corner of part a, we change
environmental uncertainty to high and complexity to low, and equivocality
remains low. Uncertainty is higher, but complexity is lower. For Bon Gout
this would be a situation where it would not have pre ordering from the
stores and the amount it could sell was uncertain. Comparing Proposition
6.3 with Proposition 6.1, we see that by changing the environmental
uncertainty from low to high, the formalization, organizational complexity,
and centralization all change. Clearly, we suggest uncertainty has a
dominate influence on the organization's structure. With a high uncertainty
and low complexity and low equivocality, there are only a few important
elements in the environment (low complexity), and they are well known
(low equivocality), but their values are not. It is a relatively simple
environment for which the numerical values are difficult to predict. So
there is a need to react quickly to events as they become known: generally,
quick reactions are important.
Medium formalization suggests a relatively small number of written
rules. The rules apply to procedures to deal with the uncertainty. In this
case social formalization using professionals may be appropriate.
High organizational complexity provides the organization with a
capacity to react on many levels. The situation is difficult but well known.
The use of specialists is appropriate to deal with these difficulties. Since the
number of variables in the environment is low, it may be possible to
coordinate the activities via centralized decision making. However, with
high uncertainty, a highly centralized organization could suffer quickly
from information overload, and there also could be a misfit if a social
formalization is used. Therefore, a medium centralization would be a proper
response to the environmental requirements.
186 Chapter 6. The Environment

We now turn to the upper right corner of Table 6.1. The environment has
high complexity, high uncertainty, and low equivocality. For Bon Gout this
could mean no preordering and multiple suppliers.
For Proposition 6.4, high uncertainty is a primary issue; coupled now
with high environmental complexity, the arguments above are even
stronger. High formalization and written rules are even more likely to get
in the way of needed adjustments. Thus, formalization should be medium
and related to procedures and policies. High environmental complexity and
high uncertainty require that the organization should react and adjust
quickly to the many factors in the environment as their values become
known. The force of the argument is increased, as the highly complex
environment has even more elements that require quick response. A highly
centralized organization is even more likely to suffer from information
overload, and low centralization is appropriate. A high degree of differentia-
tion is required (Lawrence and Lorsch, 1967), which implies a high degree
of organizational complexity.
In these four propositions, we have kept equivocality low and varied
environmental complexity and uncertainty. Now, we consider the lower half
of Table 6.1, where the equivocality is high-that is, the environmental
parameters are not well known and there is considerable ambiguity.
If the environmental equivocality is high, the organization does not
know which variables in the environment are or will be important for the
organization. The particular set of variables may not be known, but if a
given set is realized to be the important set, the number and the associated
uncertainty may be high or low. If the equivocality is high and the
uncertainty is low, one appropriate response by the organization could be
scenario planning. Ifthe uncertainty is high, scenario planning is difficult,
and the organization has to be ready to adapt quickly. Looking back at the
literature, one may argue that Lawrence and Lorsch (1967) did their study
under low equivocality, while the Burns and Stalker (1961) definition of
turbulence is close to the high equivocable situation. Therefore, their
arguments about turbulence fall under the category of high equivocality.
These observations lead us to four propositions for a high equivocality
environment.
Referring to the lower right corner of Table 6.1, not much is known
about the environment, except that its complexity is low-that is, there are
likely to be only a few important parameters in the environment. This is
much the situation for Bon Gout described in the introduction, if Bon Gout
only had Samsonite as the single supplier. High formalization with an
elaboration of rules is not likely to work well. High organizational
complexity with a high horizontal differentiation and tasks specialization,
a high vertical differentiation with a tall hierarchy, and slow response is
not likely to respond appropriately to the environment as it becomes
6.4 Environment as a Contingency Factor 187

known. Since the set of variables that will be important is not known, no
rules can be developed. Thus, the formalization and organizational
complexity should be low. However, actions must be taken and taken
quickly. Since the environment has low complexity, a centralized head can
deal with the information and react and adapt to the environment in a
timely manner. In fact, this type of environment may develop into a hostile
environment. In brief, it is a rather informal organization with a central-
ized head who reads the environment and directs the organization.
The situation for Proposition 6.6 is similar except that the environmen-
tal complexity is now high. Multiple suppliers that terminated their
contract would put Bon Gout in this situation.
As for Proposition 6.5, formalization remains low, organizational
complexity is low, but in contrast, centralization is now low. With the
complex environment, there is a lot of information to read, interpret, and
translate into actions. Given that this should be done quickly, a centralized
head is likely to become overloaded, resulting in costly delays. Conse-
quently, a decentralized organization is most likely to be able to cope with
the large amount of data. But if a centralized head can handle the
information appropriately and quickly with appropriate decision-support
systems, a centralized head is possible. However, given the equivocality of
the environment, this decision-support system must be able to handle
nonstandard situations and not just resolve high uncertainty, which may
be handled with large numerical data processing. Given today's information
technology for equivocal situations, decentralization is more appropriate.
In this situation both centralization and formalization are low. This
means that Proposition 6.6 does not recommend any means for integration.
An appropriate incentive system has to be in place to coordinate the various
activities. This issue will be treated in more detail in the next section.
Proposition 6.7 considers the low complexity, low uncertainty situation.
Low uncertainty and low complexity make it likely that a central head can
cope with the situation where the organization knows that only a few
factors will affect the organization. However, it does not know which ones
but only that each possible set offactors has a well-known value. This is a
rare situation, but it is possible. Some public organizations may not know
their future domain but know only that legislators in the future will decide
the domain from a well-specified set of possibilities. This is an example of
the situation specified in Proposition 6.7.
The argument for Proposition 6.8 follows that of Proposition 6.7, but
here a bounded rationality argument decreases the centralization from high
to low for the higher complexity situation.
Looking back at the empirical studies in the literature, one could argue
that Lawrence and Lorsch (1967) conducted their study in a situation
where all their companies experienced a low equivocality. Our propositions
188 Chapter 6. The Environment

are consistent with their results. Going from Propositions 6.1 and 6.2 to 6.4
and 6.3, there is an increase in organizational complexity due to an increase
in environmental uncertainty. Lawrence and Lorsch (1967, pp. 96-97)
found that the required intensity ofthe integration for the three industries
studied was almost the same. However, the type of integration was
different. In the low environmental uncertainty case, the integrating
devices were more related to the hierarchy of authority, and routine devices
while in the high uncertainty case were less routine and less centralized.
This is all consistent with Table 6.1.
Burns and Stalker's (1961) high-turbulence situation can be described
in Proposition 6.6. Their stable environment situation would be in either
6.1 or 6.2. Our propositions are consistent with their arguments as well.
Comparing Proposition 6.1 with 6.6, there is an increase in turbulence.
Formalization, centralization, and complexity all move in the right
direction.
In many of the empirical studies, both environmental measures and
organizational structures are lumped into a single measure. It is easily seen
that putting high and low equivocality measures into the uncertainty
measure could produce almost any result. For example, comparing 6.2 with
6.4 would result in a positive correlation between an aggregate measure of
environmental uncertainty and organizational complexity. A negative
correlation can be found by comparing 6.2 with 6.6. Comparing 6.7 with 6.3
would not give a significant difference for formalization, while a negative
correlation can be found comparing 6.2 with 6.4. The above observations
may explain the nonconclusive results obtained by Pennings (1987) and
Koberg and Ungson (1987), for example, and also the internal misfit
relations discussed by Miller (1992).
Tung (1979) aggregated her organizational measures into one. She
mapped specialization, standardization, formalization, participation into
mechanistic and organic structures. Despite the fact that standardization
and formalization are not identical measures, they both relate to our
measure offormalization. She therefore put extra weights on the formaliza-
tion measure. Tung used environmental measures very similar to our three
measures. Comparing her results (Tung, 1979, p. 689) with our proposition,
we find that our results are not inconsistent with her results.
The general results in the literature have proposed that the relation-
ships are of a monotone type (Schoonhoven, 1981). This may not be the case
and could be the reason for some of the conflicting results. The general
results also show that environmental equivocality and uncertainty effect
particularly formalization and organizational complexity, while environ-
mental complexity affects centralization.
It is, however, generally agreed that ifthe environment reaches a level
that threatens the life of the organization, then the organization should
6.4 Environment as a Contingency Factor 189

have a centralized organization regardless of the complexity of the

environment (Mintzberg, 1979). An organic structure with low formaliza-
tion and low organizational complexity may also be appropriate (Covin and
Slevin, 1989). Here, the environment is said to be extremely hostile and
management must intervene directly and decisively in the details of the
situation. This is stated in proposition 6.9 for extreme hostility and for less
extreme, but high hostility. Generally speaking, if the environmental
hostility is extreme, then it overrules, maybe only temporarily, other
factors of the appropriate organizational design. Hostility is then a
discontinuity of the environment and suggests quick and dramatic
organizational changes, if only for a short time. For Bon Gout, the
Samsonite change created a high to extreme hostile environment and
threatened its very existence-at least, its existence as Bon Gout had
known. Management had to intervene directly in many decisions that it
otherwise would not have been involved in. In the Wash & Go case
described previously, members of the central management of Procter &
Gamble came from the United States to deal with the Danish problem.
When the problem was solved, regular organizational decision procedures
were reestablished. In the united states, Johnson and Johnson have
experienced similar hostility when the over the counter drug Tylenol was
tampered with. A few individuals died. J&J immediately withdrew its
entire inventory and reestablished the brand. Only top management could
take such immediate and far reaching decisions. As we shall argue later,
the organization should establish a simple configuration, if only tempo-
rarily.
Extreme hostility calls for extreme management action, but the response
is likely to be costly and inefficient. Top priority is focus on the threat and
act on it. Many, perhaps thousands of other issues are not considered and
many other issues are poorly dealt with. It is unlikely that Bon Gout dealt
well with the normal business decisions during the crisis. We know that the
Tylenol crisis cost millions of dollars. In order to deal effectively with the
crisis, efficiency is not an issue. In hostile situations, we could also say the
risk distribution has a high probability oflarge losses. March (1994) argues
that individuals act as risk seekers, where in more normal circumstances
the individual would be more risk averse.
For a highly competitive business environment, Richardson (1996) found
that successful apparel firms decreased the level of vertical integration for
a rapid response required by shortened product life cycles. This is
consistent with the later part of proposition 6.8, which indicates that the
organizational complexity should be lowered for a high environmental
hostility.
190 Chapter 6. The Environment

6.5 Environmental Effects on Configura-

tion, Coordination, Media Richness,
and Incentives
The propositions presented in the previous section were based on the
organizational theory literature, a number of empirical results and argued
from the information-requirement perspective. When environmental com-
plexity, equivocality, and uncertainty increase, so does the requirement for
information processing capacity. Each of the environmental dimensions
requires different means to deal with the increased information processing.
This was, to a certain extent, dealt with in the previous section: formaliza-
tion, for example, is one means to increase information processing capacity
in the case with low scores on all of environmental dimensions, as was also
discussed in Chapter 2.
Here, we consider the effect of the environment on other structural
elements: organizational configuration, coordination mechanisms, media
richness, and incentives. Organizational configuration is the fundamental
organizational architecture-that is, functional, divisional, simple, matrix,
and ad hoc, as described in Chapter 2. Coordination mechanisms include
operating rules and procedures, integrating and liaison activities, group
meetings, and planning. There are many choices, and the choice of a coordi-
nation mechanism depends on the particular environmental situation.
Media richness indicates the form, amount, and kind of information. Daft
(1992, p. 286) defines information richness as the information-carrying
capacity of data. He orders media encounters from the richest medium to
the least rich: (1) face-to-face, (2) telephone and other personal electronic
media (3) letters, notes, and memos, and (4) bulletins, computer reports,
and data reports. Face-to-face conversations go two ways and involve many
clues that convey feelings, attitudes, and interpretations. In contrast, long
computer printouts of numeric data lack interpretation. A richer medium
can deal with more complex information. For media richness recommenda-
tions, we categorize richness as high, medium, or low and give a recom-
mendation about the amount or quantity of information. Incentives can
take on many variations, but two fundamental dimensions are: (1)
individual or group and (2) procedure oriented or results based (that is,
whether procedures were applied correctly or whether results turn out
well). The challenge is to select an incentive consistent with organizational
goals as well as with other organizational design parameters. Here we limit
discussion to procedure- or results-oriented situations only. In Table 6.3,
the environmental effects of equivocality, complexity, and uncertainty are
summarized for organizational configuration, coordination, media richness,
6.5 Environmental Effects on Configuration, Coordination, 191
Media Richness, and Incentives

and incentives. It has the same format as Table 6.1: low equivocality is
shown in the upper half, and high equivocality is shown in the lower half.
Table 6.4 states the propositions formally.
The arguments used to justify the propositions are based on information
processing capacity. The conclusion on the configurations follows results by
Galbraith (1973), Duncan (1979), and Mintzberg (1979). They also follow
general recommendations provided in textbooks like Daft (1992) and
Robbins (1990). The results derived on the coordination and control and
liaison devices are consistent with the results in Mintzberg (1979). In
particular this section's treatment of the information processing require-
ment and the concept of media richness are developed from Daft and Lengel
(1986). Some empirical support for the media-richness arguments can be
found in Rice (1992). The results on incentives are derived from the
discussion on planning in Mintzberg (1979), Burton and Obel (1988),
Eisenhardt (1975), and Ouchi (1979). The design ofincentives is also closely
related to the design of control mechanisms.

Table 6.3. Environmental Effects on Configuration, Coordination, Media

Richness, and Incentives

a. LOW EQUIVOCALITY
ENVIRONMENTAL
COMPLEXITY

Proposition 6.11 Proposition 6.13

High Configuration: functional, divisional Configuration: functional with liai-
Coordination: rules and procedures son activities, divisional
Media richness: low richness, mod- Coordination: integrators
erate amount of information Media richness: medium richness,
Incentives: procedural large amount of information
Incentives: results

Low Proposition 6.10 Proposition 6.12

Configuration: simple, functional Configuration: simple, divisional
Coordination: direct supervision, Coordination: direct supervision,
planning planning and forecasting
Media richness: medium richness, Media richness: medium richness,
small amount of information moderate amount of information
Incentives: procedural Incentives: results
Low
UNCERTAINTY
192 Chapter 6. The Environment

h. HIGH EQUIVOCALITY
ENVIRONMENTAL
COMPLEXITY

High Proposition 6.17 Proposition 6.15

Configuration: matrix Configuration: matrix, ad hoc
Coordination: planning, integrators, Coordination: integrator, group
group meetings meetings
Media richness: high richness, mod- Media richness: high richness,
erate amount large amount
Incentives: results Incentives: results

Proposition 6.16 Proposition 6.14

Low Configuration: simple, functional Configuration: simple, ad hoc
Coordination: direct supervision, Coordination: direct supervision,
planning. group meetings
Media richness: high richness, small Media richness: High richness,
amount small amount
Incentives: results Incentives: results
Low
UNCERTAINTY

We begin with Proposition 6.10 in Table 6.3 with the lower left corner of
part a. This is the simplest environment, with low equivocality, low com-
plexity, and low uncertainty, which suggests a simple or functional configu-
ration, direct supervision and planning for coordination, moderate media
richness, small amount of information, and procedural incentives. This fits
well with the conclusion from Proposition 6.1, with its high formalization
and medium organizational complexity. Centralization is high, although it
could be modified by size. If size is small, a simple configuration is
appropriate. If size is large, a functional configuration is appropriate.

Table 6.4. Environmental Effects on Organizational Configuration, Coordi-

nation, Media Richness, and Incentives, Propositions 6.10-6.18

6.10. If the environment has low equivocality, low complexity, and low
uncertainty, then the organizational configuration should be simple
or functional, media richness should be medium with a small amount
of information, coordination and control should be direct supervision
and planning, and incentives should be procedural based.

6.11. If the environment has low equivocality, high complexity, and low
uncertainty, then the organizational configuration should be func-
tional or divisional; media richness should be low with a moderate
amount of information, coordination and control should be rules and
procedures, and incentives should be procedural based.
6.5 Environmental Effects on Configuration, Coordination, 193
Media Richness, and Incentives

6.12. If the environment has low equivocality, low complexity, and high un-
certainty, then the organizational configuration should be simple or
divisional, media richness should be medium with a medium amount
of information, coordination and control should be based on direct
supervision planning and forecasting, and incentives should be results
based.

6.13. If the environment has low equivocality, high complexity, and high
uncertainty, then the organizational configuration should be func-
tional with liaison activities or divisional, media richness should be
moderate with a large amount of information, coordination and
control should be integrators, and incentives should be results based.

6.14. If the environment has high equivocality, low complexity, and high
uncertainty, then the organizational configuration should be simple
or ad hoc, media richness should be high with a small amount of
information, coordination and control should be via direct supervision
and group meetings, and incentives should be results based.

6.15. If the environment has high equivocality, high complexity, and high
uncertainty, then the organizational configuration should be matrix
or ad hoc, media richness should be high with a large amount of
information, coordination and control should be via integrators and
group meetings, and incentives should be results based.

6.16. If the environment has high equivocality, low complexity, and low
uncertainty, then the organizational configuration should be simple
or functional, media richness should be high with a small amount of
information, coordination and control should be via direct supervision
and integrators, and incentives should be results based.

6.17. If the environment has high equivocality, high complexity, and low
uncertainty, then the organizational configuration should matrix or
ad hoc, media richness should be high with a moderate amount of
information, coordination and control should be via planning, inte-
grators, and group meetings, and incentives should be results based.

6.18. If the hostility is extreme, then organizational configuration should be

simple. If the hostility is high, the organizational configuration could
be simple, but other factors should also be considered.
194 Chapter 6. The Environment

The relatively simple environment described in Proposition 6.10 is well

known and changes little: a simple or functional organization is appropri-
ate. The infonnation media need not be rich nor provide a large amount of
infonnation (Daft and Weick, 1984). In Table 6.3 two aspects of media
richness are given-(l) degree of richness and (2) amount of information.
Direct supervision with some planning is appropriate. Incentives can be
procedure based and based on implementation of the rules of high
formalization. It is appropriate to see that rules are followed and imple-
mented.
This situation provides a baseline. All other environments are less
simple and generally call for a more elaborate organization that can deal
with the need for more infonnation processing and complex decisions to
resolve.
Moving to the upper left corner, Proposition 6.11, the environment now
has greater complexity.
We previously argued that centralization should decrease, moving the
configuration toward a functional configuration. The functional configura-
tion pennits some task specialization to deal with the more complex
environment. Similarly, an alternative argument could be made for a
divisional product organization, if the environment partitions itself into a
relatively independent number of products and the internal technology is
similarly partitionable (Lawrence and Lorsch, 1967). Although arguments
exist for alternatives, a functional organization is most likely to be able to
deal with this environment in an efficient manner.
Due to increased complexity, a larger amount of infonnation must be
considered, although it need not be rich for this low-uncertainty and low-
equivocality environment. The coordination and control can be based on
rules and procedures. The best incentives remain procedure based as the
formalization is high and further procedure based incentives fit well with
a functional organization where results are difficult to evaluate equitably
for the separate departments. Thus, the incentives should focus on
perfonning activities well.
In Table 6.3, we now consider the lower right corner of part a. The
uncertainty is now high with low equivocality and low environmental
complexity. For this situation we have medium fonnalization and central-
ization with organizational complexity high.
Proposition 6.12 provides a basis for controversy and debate. There are
a number of alternative arguments for different recommendations. The
higher uncertainty definitely requires reading, understanding, and reacting
to a larger amount of information. There are a number of alternative ways
to accomplish the same end (Galbraith, 1974). Here, the environment has
low equivocality and low complexity, but it is difficult to predict with high
uncertainty. Thus, it becomes important to react to the environment as it
6.5 Environmental Effects on Configuration, Coordination, 195
Media Richness, and Incentives

becomes known. A simple organization with an excellent information-

support system can handle the uncertainty. However, if the size of
organization is large, a divisional organization is more likely to be
appropriate to deal with the internal complications of coordination and
control due to its high organizational complexity. These two recommenda-
tions are very different.
With high uncertainty, a moderate amount of information will be
required. But with low equivocality and low complexity, there is not likely
to be a need for rich information. We know what we need to know.
Coordination and control should be obtained via planning and forecasting
with direct supervision.
Results are clearly important. With a simpler functional organization
and low formalization, as argued in Proposition 6.3, there are fewer rules.
Thus, a procedure-based incentive may not meet organizational goals.
Incentives could be based on the results.
We now add high environmental complexity to this situation. It is more
difficult to cope with this environment, and we shall see some changes for
Proposition 6.13.
The functional organization is not likely to work well without some
mechanisms to tie it together quickly and deal with new information and
situations, and liaison activities are mandatory. These mechanisms could
range from rather informal information sharing to more permanent project,
product, or customer managers. In any event, an unaugmented functional
organization is likely to suffer from information overload with decisions not
being made in a timely fashion. The low degree of centralization and
medium formalization require a more elaborate coordination mechanism.
A divisional organization could be appropriate here, but two additional
conditions must be met. First, environmental complexity must be mostly in
terms of a large number of parameters, but they must be partitionable
along the lines of products or customers. If the environment is highly
interdependent, a divisional separation is likely to lend conflicts among the
divisions over customers, suppliers, and so on. Second, the technology must
be partitionable along divisional lines, so that each division can manage
relatively independent internal operations. Without these additional
conditions, the divisional configuration will lead to numerous conflictual
situations. The functional organization with liaison activities requires a
high level of managerial skill to cope with the environment, and the
required coordination will require integrators.
Media richness is moderate with a large amount of information required
to cope with environmental complexity and uncertainty. Incentives remain
results based, as we argued for proposition 6.12. Procedures are likely to be
changing to meet the need for good results. Therefore, the focus should be
196 Chapter 6. The Environment

on the results. To obtain the desired results incentives cannot be based on

the procedures.
We now turn to part b of Table 6.2, where equivocality is high. We begin
with Proposition 6.14; high equivocality, low complexity, and high
uncertainty demand a high level of information processing. Not much is
known about the environment, except that only a few parameters are
important-that is, the environmental complexity is low. However, much is
unknown: what are these important parameters and their values? Thus,
there is a need to read and interpret for this environment and to take
action in a timely manner. Formalization and organizational complexity
should be low with a high degree of centralization.
A simple organization with a focus on the environment is appropriate.
The supporting information system must be broadbrush in its scope but
also capable of focusing in on important issues. Top management must then
determine what to do and give directions. This is consistent with a highly
centralized organization, as argued in Proposition 6.5. Despite the fact it is
centralized, many people should be involved in interpreting information.
Group meetings and direct supervision are a means for coordination and
control.
An alternative organizational configuration would be an ad hoc
organization in which everyone is be responsible and responsive to
environment changes. Of course, a more decentralized organization must
be consistent with an ad hoc configuration. This would be the case if other
contingency factors drive the organization toward a more decentralized
organization than that prescribed in Proposition 6.5.
Media richness must be high: information will come from many sources
and in many formats. However, there will not be a lot of it as the environ-
ment is not complex. Generally, high equivocality requires high media
richness (Daft and Lengel, 1986). Kurke and Aldrich (1983) found that
managers prefer verbal media over written contact when the environment
is dynamic.
The incentives should be results based. Formalization (Proposition 6.5)
is low, and there are few rules and procedures. The goal is to succeed
regardless ofthe uncertainty and equivocality, and there must be consider-
able improvisation. Incentives on results will encourage individuals to take
appropriate actions.
Now let us consider the most complicated situation, Proposition 6.15.
Environmental complexity is now high, as is equivocality and uncertainty
in the upper right corner of part b of Table 6.3. This is the least understood
of all environments.
High environmental complexity generates additional requirements for
information processing. A simple configuration of organization is not likely
to be able to cope. A matrix organization or ad hoc organization is more
6.5 Environmental Effects on Configuration, Coordination, 197
Media Richness, and Incentives

likely to be successful. A matrix organization is recommended with caution,

because it is costly and difficult to manage. However, simpler configura-
tions such as a functional or divisional configuration are too inflexible for
high equivocality, uncertainty, and complexity. Formalization, organiza-
tional complexity, and centralization are all recommended to be low. This
leads toward the ad hoc configuration. However, if organizational complex-
ity has to increase, then the matrix is the next choice.
Coordination should be based on integrators and group meetings. Media
richness should be high, with a large amount of information. This is the
highest information-processing requirement ofall the situations. Incentives
must be results based as argued in Proposition 6.13 above.
We now consider modifying the environment for low uncertainty,
Proposition 6.17. This situation is very similar to the previous situation
except that the low uncertainty decreases the amount of information that
must be processed. The most significant change is that since uncertainty
is low, it is possible to do scenario planning, and some formalization related
to process will be possible. Integrators and group meetings are still needed
to coordinate the various activities. An ad hoc configuration is too costly
with the decrease in environmental uncertainty.
Finally, we tum to Proposition 6.16 where equivocality remains high
and complexity and uncertainty are low. This may not be an unlikely
situation, particularly fornonpublic organizations, as discussed in previous
sections. There are only a few environmental parameters, but we do not
know what they are. For example, a hot dog vendor may not know whether
it is the weather or the taste of the hot dog that generates low sales.
In comparison with Proposition 6.17, reduction of complexity increases
centralization in Proposition 6.7 and makes a simple structure possible. For
large organizations a functional configuration may be appropriate. It also
decreases the amount of information even further. Direct supervision and
some planning will be appropriate. Media richness must be high to cope
with the equivocality, but a small amount of information is sufficient.
Finally, incentives must be results based. The propositions are sununarized
in Tables 6.2 and 6.4. Additionally, an extra rule is added in each of these
tables relating to the condition of environmental hostility.
Proposition 6.16 states that an extremely hostile environment calls for
a simple configuration. Together with proposition 6.10, decision-making is
centralized and the normal structure is suspended. Here we have a crisis
which threatens the very existence of the organization. These threats
emerge from either natural or competitive actions. The information is
focused and not necessarily large. For Bon Gout, Samsonite changed its po-
licy. For Procter and Gamble, Wash and Go was said to cause one's hair to
fall out. For Johnson and Johnson, Tylenol had been laced with poison. The
198 Chapter 6. The Environment

implications are huge, but the base information is very focused. Decision-
making is a matter of judgement, not more information. Of course, there
are millions of minor issues, but they are indeed minor and can be ignored
for the large decision. These crisis situations require fast decisions,
judgement, but not large amounts of information.
In this chapter, we have related the environment of an organization to
the structure of the organization. The environment was described in terms
of equivocality, complexity, and uncertainty. The organizational structure
was considered in seven ways: formalization, complexity, centralization,
organizational configuration, media richness, coordination and control, and
incentives. The structure must match the environment, but the structure
itself must be internally consistent. For example, consider the situation
with low equivocality, low uncertainty, and low complexity. From Tables 6.1
and 6.3 it follows that the organizational structure that fits this environ-
ment is a structure with high formalization, medium organizational
complexity, and a high centralization. The configuration should be simple
or functional with some planning. Coordination and control are based on
direct supervision. Richness of the media is medium with a small amount
of information. Incentives should be based on procedure because the
centralized decision making has all the responsibility about results. All the
dimensions have to fit as they do.

Incentives: Procedural or Results Based

Organizational incentives directly affect every individual in a very personal and continuing way.
Thus, the choice of the incentive scheme is a major managerial choice with very important
consequences for the individual's own performance and also for the performance of the entire
organization. In section 6.5, we developed a number of rules conceming the appropriate choice
of incentives. Incentives were simply categorized as: procedural- and results-based with
emphasis either on the individual or the group. Briefly, a procedural-based incentive asks the
individual to do the job according to the organizational rules and norms. A results-based
incentive asks the individual to do the job according to desired results. For the most part,
organizations have procedural-based incentives for most individuals most of the time. That is,
if one does hislher job within the organizational norms, he or she will receive an hourly
wage/salary with some increase for a job well done. Results-based incentives are not widely
used, even though there is everyday discussion and a vast theoretical literature exists.
Incentives include: bonuses, merit salary increases, stock options, risk-sharing contracts, to
name a few. Results-based incentives inherently involve some risk for the individual.
For the six propositions in section 6.5, there were only two situations where procedural-
based incentives were recommended-when the equivocality is low, the uncertainty is low and
the environmental complexity can be either high or low (Propositions 6.10 and 6.11). Here it is
possible to devise good operational rules and norms which can be implemented to obtain the
organization's goals and obtain good performance. We know what the environment will be and
can plan ahead to deal with it. For all other situations, when the equivocality is high or the
uncertainty is high, the results-based incentive scheme is recommended. Here, the
environment is not known in advance and we argue that the individual should have the incentive
to adjust to the actual environment in order to meet the organization's goals and obtain good
performance. However, if the performance is not realized, then the individual will not receive
the high results-based rewards. Hence, the individual is sharing the risk with the organization.
6.6 Environmental Effects on Configuration, Coordination, 199
Media Richness, and Incentives

If we live in a world of uncertainty and equivocality and results are then desirable, why are
they not used more widely? There are a number of reasons. First, results-based incentives are
difficult and costly to administer. It is difficult to construct individual performance measures for
each individual in an organization. The performance measure must be established; it must be
monitored and measured; it must be converted into a reward for the individual. Second,
performance depends upon a group or team, or even the whole organization and is not easily
partitioned into individual effort. Indeed, organizations should have large positive externalities
among the individuals, that is one imperative for the organization itself. Of course, group
incentives are possible and reasonable. Then a sharing rule for individuals in the group must
be devised and accepted by the group members to be effective. Groups can also suffer from
the free rider-the individual who accepts the rewards of the group, but does not do his/her fair
share in the eyes of the group. Groups incentives can be very effective, but again they are
difficult and costly to administer.
Some, perhaps many individuals do not want to accept the risk of a results-based incentive.
There is a long tradition in industry that the business risk should be taken by the organization
and not by the individual. Simply, "why should I receive less for a job well done if management
cannot determine a way to be profitable." This is a philosophical statement about the nature of
the relation between the individual and the employer. From a more pragmatic view, the
individual may be risk averse and would rather accept a known (lower than average) reward
than incur the risk and uncertainty of a results-based incentive scheme.
For many of these reasons, results-based incentives are not widely used. Despite their
appeal, they are costly and difficult to administer and many individuals do not want to accept
the risk.
Incentives must meet the contingency fit as discussed above, but must also meet the design
fit conditions. That is, the incentives must fit with the organization configuration and
organizational properties. E.g., a divisional organization is well matched with results-based
incentives at the divisional level, Le., divisional profits. Indeed, the results-based incentives can
be a strong reason for moving from a procedural based functional configuration to a divisional
form. Results-based incentives for a functional configuration and procedural-based incentives
for a divisional configuration are likely to lead to inefficiencies and are not recommended. (See
the propositions in section 6.5.)
Design fit requires that a change in the organization should be viewed in its entirety. To
change only one design element and leave the others unchanged will likely create a design
misfit. As discussed above, a change from a functional configuration without a change in
incentives creates a misfit. We observe that this is a common managerial error. Not infrequent-
ly, management will change the organizational configuration and not change the incentive
scheme, the coordination mechanisms, the rules and norms of formalization and the level of
centralization. The organization is then left to its own devices to make these later changes
without rationale, plan or a sense of expectation for the individuals-it can be a trying
experience.
Incentives can playa central role in matching the individual efforts and expectations with the
organizational goals and performance. For low equivocality and uncertainty, procedural-based
incentives are appropriate. For high equivocality or high uncertainty, results-based incentives
are recommended as a means to motivate the individual to take appropriate actions to achieve
the organization's goals. Despite the logic for results-based incentives, there are also good
reasons for not utilizing incentive based rules: difficult and costly to administer, philosophical
separation of responsibilities, risk aversion by the individual, to name a few. However, the
design fit criteria indicates that inefficiencies can arise when the incentives are misaligned with
the other organizational design configurations and properties.

The other extreme situation with high equivocality, high uncertainty, and
high complexity is best met with low formalization, low organizational
complexity, and low centralization. The configuration should be matrix or
ad hoc. Coordination should use integrators and group meetings. Media
200 Chapter 6. The Environment

richness should be high and have the capacity to process large amounts of
data. Since procedures cannot be established, incentives should be based
on results.
The propositions are nonnative, by which we mean that they are our best
recommendation to managers for practice. The support for the proposition
is a synthesis: the literature in organizational theory is rich and empirically
abundant, our validation of Organizational Consultant (knowledge base)
led us to reexamine what we know, ask new questions, and propose
tentative answers, and finally, the propositions must be intuitively
appealing and arguable to the practicing manager and student.

6.6 Operationalizations of the Environ-

mental Measures
We have proposed a four-dimensional measure of the environment. Even
though such a measure is more detailed than the ones used in many
studies, they are still rather aggregate. Two issues are dealt with in this
section. The first one is whether our measures are objective measures, and
the second issue is what the environment really is.
The most objective dimension of the four is complexity. It is relatively
easy to enumerate the important factors in the environment that affect the
organization. Disagreements may arise over the type of effect and the
intensity of the effect, but usually a rough count can be agreed on. The
situation is quite different with respect to the other three dimensions
(Tung, 1979, Downey, Hellriegel, and Slocum, 1975). Both a high degree of
uncertainty and a high degree of equivocality may be attributed to lack of
infonnation or the randonmess of some events. If the uncertainty can be
attributed to lack ofinfonnation, the degree ofuncertainty and equivocality
depends on the precise infonnation the organization has. One organization
may consider a situation very uncertain and equivocal, while another
organization in exactly the same environment considers its environmental
uncertainty and equivocality very low. Hostility is both an issue of
perception and the reality of whether someone or something is out to get
you. The environment is enacted (Weick, 1969). In the telecommunication
industry, one company may be on top of the newest technology develop-
ments, have good contact with the political scene, and based on this, have
a very good notion about what the future brings. Another company may not
have the capacity to do the same and may find the future uncertain and
difficult to understand and predict. A number of means to reduce the
uncertainty and equivocality are discussed in the next section.
6.6 Operationalizations of the Environmental Measures 201

The environment consists of many different parts. The industry,

including its size and competitions, is a major part of the organization's
environment. Actual customers are also included in the environment.
Suppliers of resources for the organization are important as well. This
includes suppliers of raw materials, services, and the labor market.
Financial resources have to be taken into consideration including the stock
market, banks, and private investors (see the earlier capsule on the Five
Forces).
The evolution of the techniques of production is a part of an organiza-
tion's environment. The political and general economical factors are a part
too, including regulation, taxes, services, and the political system in
general. Additionally, inflation rates, exchange rates, and other important
economic factors have to be considered.
Finally, more general sociological factors should be accounted for.
Values, beliefs, education, religion, work ethic, and special current trends
such as environmental movements and women's rights are important.
Briefly defined, the environment is everything that is outside the
organization. This relates to the unit of analysis, as was discussed in
Chapters 1 and 2 and a later capsule on new forms and strategy in Chapter
8. Some of the factors will be more important than others and more or less
uncertain for a particular organization.

Hypercompetition and Flexibility

Hypercompetition, or Schumpeterian competition, goes beyond the normal static competition
to include a number of elements: dynamics, rapid change and adaptation, short product life
cycles, discontinuities, creative destruction and general Darwinian behavior. D'Aveni(1994, p.
154) describes hypercompetition as "an environment characterized by intense and rapid
competitive moves," and the "behavior is the process of continuously generating new
competitive advantages and destroying, obsoleting, or neutralizing the opponent's competitive
advantage, thereby creating disequilibrium ..." High technology, telecommunications, software,
aircraft, healthcare and biotechnology are a few of the hypercompetitive industries.
Hypercompetition is widespread and is not limited to a select number of industries. Unlike
extreme hostility, hypercompetition does not end and can be thought of as a continuing
situation-a continuing state of disequilibrium, perhaps a new type of equilibrium.
What is the appropriate organizational response to hypercompetition? D'Aveni (1994, p.
172-176) offers a new7S's framework: superior stakeholder satisfaction, strategiC soothsaying,
speed, surprise, shifting the rules, signaling, and simultaneous and sequential strategic thrusts.
Later (p.196), he recasts speed and surprise capability for disruption. The capability for
disruption is most closely related to the organizational design; although speed and surprise are
organizational characteristics of potential behavior than organizational design itself. That is, for
a hypercompetitive environment, we want an organization which acts with speed and surprise
to create disruptions and temporary advantages for the organization. The organizational design
rules to realize speed and surprise are largely yet to be developed; however, D'Aveni offers a
number of rich and detailed examples including examples from the film industry.
Volberda (1996) argues that a flexible organizational form is required for a hypercompetitive
environment-a contingency rule that states, "if the environment is hypercompetitive, then the
organizational form should be flexible." A flexible organization involves a paradox. "Flexibility
is the degree to which an organization has a variety of managerial capabilities and the speed
202 Chapter 6. The Environment

at which they can be activated, to increase the control capacity of management and improve
the controllability of the organization" (p. 361). Ashby's (1956) law of requisite variety is
fundamental; the variety of the potential actions by the organization must be as large as the
variety of the disturbances in the environment. Volberda then offers three propositions for the
rigid form, the planned form and the flexible form-the later Proposition 3 (p. 366-367) being of
most interest here:

In a fundamentally unpredictable environment, which may also be dynamic and complex

(hypercompetitive), the optimal form employs a broad flexibility mix dominated by structural
and strategic flexibility and has a non routine technology, an organic structure, and an
innovative culture. The intelligence-gathering and information-processing aspects of
metaflexibility are directed toward enhancing the receptiveness to new environments.

He argues that this form is a good balance between change and the need for preservation.
In brief, a hypercompetitive environment requires a flexible organizational form.
Social network forms and cluster forms are examples of flexible forms. The social network
form works well for the biotech hypercompetition. It is characterized by redeployable technology
and organic structure, but with strong professional ties between actors. The cluster form gives
strategic range of action, but with structural control and preservation.
Hypercompetition and flexible form call for a new vocabulary for the change and its rapidity
that we are experiencing for the new millennium. It is interesting to look for similarities of
concepts and juxtapose definitions and design rules.
In our terms a hypercompetitive environment could be categorized as: high equivocality, high
uncertainty, high environmental complexity and high competitiveness. The technology is
nonroutine. For this environment and technology, we would recommend: low formalization, low
organizational complexity and low centralization (Propositions 6.6 and 7.1) and matrix or ad hoc
form, coordination by integrators and meetings, high media richness and a large amount of
environmental information and results-based incentives (Propositions 6.15 and 7.22).
We suggest that these recommendations are not unlike the Volberda recommendations:
structural and strategic flexibility, organic structure, innovative culture, intelligence-gathering
and information-processing aspects of metaflexibility.
Hypercompetition and nonroutine technology are situational fit statements in our framework
and it is a very reasonable requirement.
Hypercompetition and flexible organization are new words and now a part of managerial
vocabulary. They embed the essence of time: change is more rapid and the environment is less
benign. Yet, it remains a research question whether hypercompetition is a new strategy para-
digm and flexibility a new organizational form, or whether these are particular environments and
organizational requirements which have been incorporated, albeit as end points in existing
theory. It may be some time for this research agenda to develop and for answers to emerge.

6.7 Managing the Environment

When Bon Gout learned that Samsonite had canceled its contract it took
action. First, it managed to obtain an appropriate inventory. Second, it
sued Samsonite and started to react to the fierce competition from
Samsonite-in the stores. This shows that actions can be taken to change
the value of all four measurements of the environment.
We argue that there has to be a fit between environment and organiza-
tional structure. The fit can be obtained either by adjusting the structure
or by changing the environment. Ifhigh uncertainty and equivocality come
6.7 Managing the Environment 203

from lack of information, actions that provide better information will reduce
the scores. Market analysis, participation in technical conferences, and so
on are all means to increase knowledge about the environment. The
organization may even change its domain, or it may have to. For Bon Gout
there may exist no other choice than to find something else to do or die.
Buffering may reduce the effect of uncertainties. Bon Gout managed to
obtain an inventory. That is a buffering strategy. If the demand for the
output is very irregular, the organization may have to even out the
demand. Price differentiation and rationing are strategies that help in that
direction. Advertising is perhaps the best-known way to try to influence the
environment. Vertical integration and lobbying are further examples, but
many more exist.
Some strategies may be available to some organizations and some may
not. It may depend on size (large organizations can do more advertising
than small organizations), culture (in some cultures bribing is customary,
but in other settings it is not), or legal setting (public organizations are
usually restricted in such actions).
For the organization it is important to analyze the source ofequivocality,
uncertainty, environmental complexity, and hostility. Table 6.5 shows a
chart to begin. For each of the thirty-two entries in Table 6.5, equivocality,
uncertainty, complexity, hostility are to be accessed. A complete analysis
then will provide a complete picture on the four dimensions. Interdepen-
dence between the various sectors has to be taken into account. For
example, decisions with respect to the tax laws may affect economic factors
like the inflation.

Table 6.5. Environmental Sectors and Dimensions

Environmental
Sector Equivocality Uncertainty Complexity Hostility

Market and industry

Raw material
Human resources

Financial resources
Technology
Economic system
Political system
Social system
204 Chapter 6. The Environment

The organization then has to establish a fit between the organization and
the environment. The adaptation of the organization to the environment
was the subject of previous sections. Fit also can be obtained by changing
the environment. However, the manipulation depends on which entry in
Table 6.5 can be changed and how it depends on other sectors. The market
and industry can be changed by a domain change, vertical integration,
advertising, mergers, market analysis, and so on. Contracting and buffering
strategies are useful with respect to suppliers. Negotiations with the union,
educational activities, and recruitment are issues in the human resources
sector. Decisions about leverage, choice of bank, and timing of issuing new
stocks are choices that affect the equivocality, uncertainty, and complexity
of the environment. Investment in research and development can reduce
equivocality and uncertainty in the technology sector.
Not all factors can be manipulated easily: general economic factors
cannot. Organizations have to respond properly to inflation rates and
unemployment rates without having many possibilities to affect these
factors. Economic factors may be affected through lobbying, the general
term for activities that seek to influence political decisions.
Organizations are part of a general social system. It is interesting to
contrast the evolution of organizational designs with the evolution of
society. The more democratic, less centralized organizations reflect similar
trends in our societies. For international organizations, social and cultural
factors are particularly important. One organizational design may work in
one country, but not in another. The global organization has a choice of
which activities to do in which country.
The organization has its choices. It can adapt the organization to its
environment, or it can try to manipulate the environment to fit the
organization.

6.8 The SAS Environment

Until the beginning of the 1970s, the market for international air traffic
was heavily regulated and based solely on bilateral international agree-
ments. These agreements controlled which routes each airline company
could fly and how much capacity each company could offer. The forum for
international price coordination was the International Air Transport
Association (lATA), founded in 1945. lATA had members all over the world,
and consequently the price agreed on by lATA became the world market
price. When a price had to be decided, airline companies produced
calculations of how cheaply they could sell their travels, and based on these
calculations a mutually acceptable price was negotiated.
6.8 The BAS Environment 205

Finally, prices had to be approved by the national governments, which

was normally a mere formality. Like so many other airline companies, SAS
could consider the prices fixed between two lATA meetings and conse-
quently price competition was excluded. Governmental control of routes and
lATA-fixed prices were further reasons that the environment that SAS
encountered in the beginning of the 1950s was very stable. The market
showed constant growth with room for everyone. Oil prices were increasing
only slightly, and most airline companies had a fair profit every year. This
was the situation until the beginning ofthe 1970s when the first oil crisis
changed the whole market situation.
Suddenly oil prices rose explosively, and consequently total costs of
airline companies rose dramatically, and there was nothing the airlines
could do about it. The oil crisis resulted in a general weakening ofthe entire
world market, and where the market for air travel had been steadily
growing it suddenly stagnated. Many airline companies found that their
assumption that the market growth of the 1960s would continue had led
them to purchase planes that were no longer necessary. Furthermore, jet
airplanes were now introduced on the market with larger airplanes. All in
all, it resulted in too many seats for too few passengers, so the companies
lowered prices to fill their planes. At the same time the US government
began to liberalize the American airline policy. This meant increased
competition on the already keen market for trans-Atlantic air travel. Oil
crises and liberalization in the United States resulted in an increasingly
noticeable overcapacity. All were eager to get new customers, but where
would they find them? The only possibility was to "steal" customers from
other airline companies, and thereby the competition in Europe began in
earnest.
During the 1980's competition became increasingly severe. Both
European and overseas airline companies began to prepare themselves for
the open borders of the European Union in 1992. The airline business had
been protected because each country had given airline companies monopo-
lies on their routes. Through their organization lATA, the airline companies
had tried to avoid potential competitive situations. But it was quite clear
to everyone that EU's open borders would enable anyone to enter and
capture new markets. No one knew for sure what the new situation would
look like and which factors would be decisive. Therefore, it was important
to get as firmly established as possible in the market before 1993. In this
connection SAS had its own special problems. The fact that SAS was
situated on the periphery ofthe EU was not a favorable starting point. The
three large European airline companies-British Airways, Air France, and
Lufthansa-all had a larger home market, and Air France and Lufthansa,
in particular, were situated centrally in Europe.
206 Chapter 6. The Environment

The rigid, bureaucratic organization that SAS had established during its
time of prosperity was not at all able to succeed in the new more unstable
and complex environment. An obvious response to the crisis was to make
all sections within SAS more cost aware, and economies were made through
the end of 1970s. But as these efforts failed to produce results, management
realized that SAS's structure was obsolete. SAS needed a new structure
that could engage successfully with the new environment.
As mentioned earlier, Jan Carlzon was appointed manager of the part
ofthe new structure called the Airline Company SAS. This began a new era
for SAS. Carlzon's first task was to adapt the strategy to the new market
situation. He changed the company's strategy from being product oriented
to being market oriented. Furthermore, by giving front-line personnel more
responsibility and allowing them to discover alternative solutions (within
certain defined limits), SAS was able to react more quickly to changes in
the market. Envisioning increased competition from outsider companies
due to the deregulation of the industry, Carlzon stressed the importance of
SAS to establish a secondary home market. This should be achieved by co-
operation with another airline company. Carlzon's successor Stenberg later
followed up on Carlzon's idea, and in May 1996 SAS announced plans for
a strategic alliance with Lufthansa.
The free market competition resulting from the accomplishment of the
deregulation on April 1st, 1997, has further increased the importance of
SAS's competitiveness. Even on the Scandinavian market, in which SAS
has traditionally held a monopolistic status, the company has experienced
increased competition. As a means to take on competition, SAS has
expanded its number of flight offerings by making a number of strategic
alliances. Through these alliances SAS strives to cover a larger global
market. By sharing administrative costs with its allied partners, SAS is
furthermore hoping to reduce costs to an extent that will allow the company
to reduce prices and hence become more competitive. One of the most
important ofthese is the Star Alliance announced on May 1997 to initially
include: SAS, Lufthansa, Air Canada, THAI Airways International and
United Airlines. The core of this is a code share agreement, enabling all
member companies to offer more frequent departures to their customers.
Other environmental factors of importance to SAS are governmental
actions. In 1997, SAS's costs were increased when the European Commis-
sion's introduced a mineral oil directive, increasing environmental duty on
fuel for planes. Other governmental actions aiming at a reduction of the
allowed noise level of aircrafts and leakage of C02 has stressed the
importance of a replacement of SAS's existing aircraft fleet, as SAS may
otherwise be excluded from landing at some airports.
To sum up the situation, before the Mid-east oil crisis, SAS's environ-
ment could be described as low uncertainty, low complexity and low
6.9 Summary 207

equivocality-the simplest and most stable of all environments. Referring to

Table 6.1 and 6.4 and Propositions 6.1 and 6.10, we find that SAS should
have high formalization, medium organizational complexity, high central-
ization, a functional or a divisional configuration, coordination and control
by planning and direct supervision, medium rich information, and
procedural incentives. This is a reasonable description of the old SAS as
discussed in Chapter 2.
The late 1970s changed all of that. The SAS environment now had high
equivocality, high complexity, and high uncertainty-the least simple and
least stable environment. Again, referring to Tables 6.1 and 6.4 and
Propositions 6.6 and 6.15, many organizational changes were called for: low
formalization, low organizational complexity, low centralization, a matrix
or ad hoc configuration, coordination through integrators and group
meetings, high media richness, and results-oriented incentives. Carlzon
initiated a number of these changes-applying all to a certain extent.
Nonetheless, formalization and organizational complexity remained
relatively high.
Stenberg's appointment in 1994 has tended to recentralize the organiza-
tion while the formalization and organizational complexity remain
unchanged. The SAS organization can therefore be described as relatively
formalized, and complex with a low to medium level of centralization. The
effect of contingencies such as technology however have to be taken into
account also.

6.9 Summary
Unequivocal, simple, certain environments call for relatively uncomplicated
structures with an emphasis on formalization, rules, procedures, specializa-
tion, and rewards for doing a job well. As the environment has increased
variety, then the organization's structure must have increased variety. The
fundamental concern is to be able to process the requisite information-that
is, to read, interpret, and take appropriate actions in this more unknown
situation. Ashby's law of requisite variety posits that variety in the
organization must exceed variety in the environment for viability (Ashby,
1956). In design terms, if the environment has high variety, then the
organization should have high variety to cope. The elaboration of the
environment into three characteristic descriptions and the organizational
structure into seven elements is a way to give precision to the general
proposition and also incorporate what we know from the literature and
experiences.
208 Chapter 6. The Environment

Notes
1. Our thanks to Thorkild J0rgensen, Aarhus School of Business, who has
applied the five forces framework in his organizational design classes to
describe the environment and to Bo Eriksen, Odense University, who
suggested the outline for this capsule.

References
Arrow, Kenneth J. 1974. The Limits of Organization. New York: W.w. Norton.
Ashby, W. Ross. 1956. Introduction to Cybernetics. London: Methuen.
Buchko, Aaron, A. 1994. "Conceptualization and Measurement of Environmental
Uncertainty: An Assessment of the Miles and Show Perceived Environmental
Uncertainty Scale." Academy of Management Journal, 37(2), 410-425.
Bourgeois, J. and Kathleen Eisenhardt. 1988. "Strategic Decision Processes in High
Velocity Environments: Four cases in the Microcomputer Industry." Manage-
ment Science, 34, 816-835.
Burns, Tom, and G. M. Stalker. 1961. The Management of Innovation. London:
Tavistok.
Burton, R.M., and B. Obel. 1986. "Environmental-Organizational Relations: the
Effects of Deregulation." Thchnovation, 5(1), 23-34.
Burton, R.M., and B. Obel. 1988. "Opportunism, Incentives and the M-form
Hypothesis." Journal of Economic Behavior and Organization, 10, 99-119.
Chandler, Alfred D. 1962. Strategy and Structure: Chapters in the History of the
Industrial Enterprises. Cambridge, MA: MIT Press.
Covin, Jeffrey G., and Dennis P. Slevin. 1989. "Strategic Management of Small
Firms In Hostile and Benign Environments." Strategic Management Journal, 10,
75-87.
Cyert, R.M., and J. G. March. 1963. A Behavioral Theory of the Firm. Englewood
Cliffs, NJ: Prentice Hall.
Daft, Richard L. 1992. Organization Theory and Design (4th ed.). St. Paul, MN. (2nd
ed., 1986, 3rd ed., 1989).
Daft, Richard L., and Robert H. Lengel. 1986. "Organizational Information
Requirements, Media Richness and Structural Design." Management Science,
32(5), 554-571.
Daft, Richard L., and Karl E. Weick. 1984. "Toward a Model of Organizations an
Interpretation Systems." Academy of Management Review, 9, 254-295.
D'Aveni, Richard A. 1994. Hypercompetitive Rivalries: Competing in Highly
Dynamic Environments. New York: The Free Press
Downey, H. Kirk., Don Hellriegel, and John W. Slocum, Jr. 1975. "Environmental
Uncertainty: The Construct and its Application." Administrative Science
Quarterly, 20, 613-629.
References 209

Duncan, Robert B. 1972. "Characteristics of Organizational Environments and

Perceived Environmental Uncertainty." Administrative Science Quarterly, 17,
313-327.
Duncan, Robert. 1979. "What is the Right Organization Structure?" Organizational
Dynamics, Winter, 59-79.
Eisenhardt, Kathleen M. 1975. "Control: Organizational and Economic Approaches."
Management Science, 31(1), 134-149.
Eisenhardt, Kathleen M. 1989. "Making Fast Strategic Decisions in High-Velocity
Environments," Academy of Management, 32(3), 543-576.
Fink, Stephen L., R Stephen Jenks, and Robin D. Willits. 1983. Designing and
Managing Organizations. Homewood, IL: Irwin.
Galbraith, Jay R 1973. Designing Complex Organizations. Reading, MA: Addison-
Wesley.
Hannan, M.T., and J.H. Freeman. 1977. "The Population Ecology of Organizations."
American Journal of Sociology, 82, 929-964.
Jurkovich, Ray. 1974. "A Core Typology of Organizational Environments."
Administrative Science Quarterly, 19(3),380-394.
Koberg, Christine S. and Gerado R. Ungson, 1987. "The Effects of Environmental
Uncertainty and Dependence on Organizational Structure and Performance: A
Comparative Study." Journal of Management, 13(4),725-737.
Koed Madsen, Tage, and J.M. Jensen. 1992. "Analyse klassifikation og behandling
afnegative Rygter." Ledelse og ErhvervSf)konomi, 56(1) (in Danish).
Kuhn, Arthur J. 1986. Organizational Cybernetics and Business Policy: System
Design for Performance Control. University Park: Pennsylvania State University
Press.
Kurke, L.B. and H.E. Aldrich. 1983. "Mintzberg Was Right! A Replication and
Extension of the Nature of Managerial Work." Management Science, 29(8),
975-984.
Lawrence, Paul R 1981. "Organization and Environment Perspective." In Andrew
H. Van de Ven and William F. Joyce (Eds.), Perspectives on Organization Design
and Behavior (pp. 311-337). New York: Wiley.
Lawrence, Paul R, and Jay W. Lorsch. 1967. Organization and Environment.
Boston: Harvard Business Press.
March, James. 1994. "Three lectures on Efficiency and Adaptiveness in Organiza-
tions." Swedish School of Economics and Business Administration, Research
reports, No 32, Helsingfors.
Miller, Danny. 1992. "Environmental Fit Versus Internal Fit." Organizational
Science, 3(2),159-178.
Milliken, Frances J. 1987 "Three Types of Perceived Uncertainty About the
Environment: State, Effect, and Response Uncertainty." Academy of Manage-
ment Review, 12(1), 133-143.
Mintzberg, Henry. 1979. The Structuring of Organizations. Englewood Cliffs, NJ:
Prentice-Hall.
M9111er Jensen, Jan. 1993. "Strategic Framework for Analyzing Negative Rumors in
the Market Place: The Case of Wash & Go in Denmark." In M. Joseph Sirgy,
Kenneth D. Bahn and Tunc Erem (Eds.). Proceedings of the Sixth Bi-Annual
International Conference of the Academy of Marketing Science. Istanbul, (pp.
559-563).
210 Chapter 6. The Environment

Organization Science special issue on Hypercompetition, Jul-Aug 1996.7(4).

Ouchi, William G. 1979 "Design of Organizational Control Mechanisms." Manage-
ment Science, 25(9), 833-848.
Pennings, J.M. 1987 "Structural Contingency Theory: A Multivariate Test."
Organizational Studies, 8(3), 223-240.
Pfeffer, J. 1982. Organizations and Organization Theory. Boston: Pittman.
Porter, Michael E. 1979. "How competitive forces shape strategy." Harvard Business
Review, Mar-Ap, 137-145.
Porter, Michael E. 1980. Competitive Strategy: Techniquees for Analyzing Industries
and Competitors. New York: The Free Press.
Rice, Ronald E. 1992. "Task Analyzability, Use of New Media, and Effectiveness: A
Multi-site Exploration of Media Richness." Organizational Science, 3(4), 475-500.
Richardson, James. 1996. "Vertical Integration and Rapid Response in Fashion
Apparel." Organization Science, Jul-Aug, 7(4), 400-412.
Robbins, Stephen P. 1990. Organization Theory: Structure, Design, and Applica-
tions. Englewood Cliffs, NJ: Prentice-Hall.
SAS. 1975-991. SAS Annual Reports. Stockholm: SAS.
SAS. 1994. SAS Annual Reports. Stockholm: SAS.
SAS. 1995. Annual Report. Stockholm: SAS.
SAS. 1996. MiljfJregnskab. Stockholm: SAS
SAS. 1996. Annual Report. Stockholm: SAS
SAS. 1996. De ffJrste 50 dr. Stockholm: SAS
SAS. 1996. ScandinatianAirlines1946-1996. Stockholm: SAS
SAS. 1997. Deldrsrapport 1: 1997. Stockholm: SAS
Schoonhoven, Claudia Bird. 1981. "Problems with Contingence Theory: Testing
Assumptions Hidden within the Language of Contingency Theory." Administra-
tive Science Quarterly, 26, 349-377.
Scott, W. Richard. 1992. Organizations: Rational, Natural, and Open Systems.
Englewood Cliffs, NJ: Prentice-Hall.
Thompson, J.D. 1967. Organizations in Action. New York: McGraw-Hill.
Tosi, Henry, Ramon Aldag, and Ronald Stoney. 1973. "On the Measurement of the
Environment: An Assessment of the Lawrence and Lorsch Environmental
Uncertainty Scale." Administrative Science Quarterly, 18, 27-36.
Tung, Rosalie L. 1979. "Dimensions of Organizational Environments: An Explor-
atory Study of Their Impact on Organizational Structure." Academy of Manage-
ment Journal, 22(4), 672-693.
Volberda, Henk, W. 1996. "Toward the Flexible Form: How to Remain Vital in
Hypercompetitive Environments." Organization Science, 7(4), 359-374.
Weick, Karl E. 1969. The Social Psychology of Organizing. Reading, MA: Addison-
Wesley.
CHAPTER 7

Technology

7.1 Introduction
Med Electronic, Inc. is a medium-size company that specializes in electronic
apparatus that is used in the treatment of pain and in other electronic
devices used by hospitals. The machines have been custom made to the
particular needs of the user department or physician. Some basic compo-
nents are used in all its devices, but no two machines are similar. This has
caused problems when machines come in for repair because documentation
and specifications for a particular machine may be difficult to find. Med
Electronic's performance has been stable for some time, and the owners
have been pleased with its performance. The employees are either
engineers or highly trained technicians. Med Electronic has had a stable
share of the world market for its particular product.
Hospitals everywhere now face difficult economic times. Governmental
and private hospitals as well as health insurance companies are trying to
control costs. It is more difficult to maintain their usual revenues. On the
other hand, consumers spend more money on alternative health care. There
are increasing sales in do-it-yourself medical instruments, like blood
pressure measurers and pregnancy tests.
The planning committee of Med Electronic has proposed that the
company enter the consumer market. One pain-treatment machine that has
been produced in a number of different versions is considered to have
significant potential in the world market. It could be produced in high
quantities in a standard model, and the price could be relatively low. The
main problem would be to maintain high quality. The management thinks
favorably of this proposal, but has some concerns that the internal
structure of the corporation will not fit the new situation.
In the past, there had been no rules. The engineers were both designers
and salespersons. They dealt directly with hospitals and physicians. To
maintain quality, the production process has to be disciplined. Documenta-
tion has to be produced and maintained. Entering the consumer market
demands organizational changes. A separation of sales and production has
to be considered. Since the price is low, margins are low as well, which
requires cost control and efficient operations. Procedures should become
more formalized, and activities better coordinated. Design and documenta-
212 Chapter 7. Technology

tion have to be fixed. The technology of the company would change and
could cause a change in the organizational structure. Technology as a
contingency for the organization is the topic of this chapter.
In this chapter, we review the literature on the technology imperative;
i.e., the organization's technology affects the best choice of the organiza-
tional design. Information processing provides the framework for develop-
ingpropositions on the technology and design. We visit again the SAS and
examine the effect of its technology on the organizational design.

7.2 A Literature Review

Technology is the information, equipment, techniques, and processes
required to transform inputs into outputs (Robbins, 1990, p. 176). This very
general definition is widely accepted. The skill level and capacity of
employees help define the techniques and processes and must be compati-
ble with the equipment and physical plant.
The idea that the technology ofthe organization affects its structure was
a discovery of Joan Woodward (1965). Her study was the first to link
technology and structure.
Woodward, in her study of industry in the south ofEngland, categorized
technology into three types:

• Unit: Custom made and the work is non routine (ex: a craftshop or
job shop),
• Mass: Large batch or mass produced, usually standard products (ex:
an assembly line),
• Process: Highly controlled, standardized and continuous processing (ex:
a refinery or brewery).

The scale reflects increasing technological complexity. Woodward linked

these technologies with structure. The administrative component, or
nondirect worker component, increased with increased complexity.
From an information processing point of view, the categories may be
ordered mass, unit, and process. Mass production may require the lowest
information processing capacity, if proper formalization and coordination
mechanisms are used. Less formalization can be used in unit production,
and more complex information processing is needed to obtain coordination.
Ifthe organization is small, coordination can be obtained by centralization.
Professionalization also may be used. Process production with its automa-
tion requires complex information processing. The amount may, however,
be lower than for unit production for a given size of the organization.
Robbins (1990, p. 180) summarizes Woodward's findings (see Table 7.1).
7.2 A Literature Review 213

Table 7.1. Summary ofWoodward's Findings on the Relationship Between

Technological Complexity and Structure

Low ~
Technology
• High

Structural Unit Mass Process

Characteristic Production Production Production

Number of vertical levels 3 4 6

Supervisor's span of control 24 48 14
Manager per total employee ratio 1:23 1:16 1:8
Proportion of skilled workers HIGH LOW HIGH
Overall complexity LOW HIGH LOW
Formalization LOW HIGH LOW
Centralization LOW HIGH LOW

Briefly, as the technological complexity increases,

• Vertical differentiation increases,

• Span of control is curvilinear, with mass production having the widest
span of control,
• Manager per total employees ratio increases,
• Proportion of skilled workers is curvilinear, with mass production
requiring the least skill,
• Organizational complexity is curvilinear, with mass production being
the highest,
• Formalization is curvilinear, with mass production being the highest,
and
• Centralization is also curvelinear, with mass being the highest.

Complementary to the information processing arguments, an alternative

explanation begins with the nature of the work. Unit production is more
craft oriented: individual workers are not closely linked and can work
relatively independently. Hence, with close supervision, formalization and
centralization are not needed. Mass production is driven by the regularity
and interdependency of the worker and the machine or assembly line. The
machine itselfdrives the work and provides coordination of the in terdepen-
dency. Workers must conform to formalized work rules with high central-
ization; but a supervisor can supervise a large number of workers due to
the formalization and standardization.
Process production is yet a different issue: refineries and breweries run
themselves. The task is to keep them running smoothly and efficiently.
Control and maintenance of the process are major issues. These require
high skill, but there is a good deal of discretion about the exact timing of
214 Chapter 7. Technology

the work. Thus, the complexity, formalization, and centralization can be

low, but many managers are required in order to ensure the reliability of
the complex technology. Briefly, mass production is a clockwork for workers
driven by the machine; unit and process production permit greater
discretion for workers to accomplish tasks at their own pace.
Woodward's conclusions have been criticized for not taking size into
account. Usually, mass production organization will be larger than unit and
process organization. From an information-processing complexity point of
view, one may argue that the order is mass, unit, and process.Based on the
amount of information, it may be ordered unit, mass, and process. The
number of vertical levels may be related to the managers per total
employees ratio. However, span of control and skill level of employees are
linked to complexity of information. The more complex information, the
higher the demand for information processing capacity, and thus the higher
the number of vertical levels. This is a bounded rationality argument. The
relationship, however, is contingent on the skill level of employees in two
ways. If employees have a high skill level, they can process complex
information. Ifthe manager has a high skill level, he or she also can process
complex information. Therefore a high skill level will, ceteris paribus,
decrease the vertical differentiation. Modern information technology can
decrease the vertical information system and the number of vertical levels.
Computer information systems, e-mail, and so on increase information
capacity and eliminate the need for a large middle management. That is,
modern information systems can process information previously handled
by middle management. This is one explanation for the flattening occurring
in many organizations (See capsule on delayering later in this chapter).
With this view Woodward's results make perfect sense including the
implications for organizational design.
Perrow (1967, p.195) viewed technology differently: "By technology is
meant the actions that an individual performs upon an object with or
without the aid of tools or mechanical devices in order to make some change
in that object." He developed a two-dimensional scheme-task variability
and problem analyzability. Task variability could have few exceptions or
many exceptions. Problem analyzability could be well defined (a logical
analytical search process exists) or ill defined (no search program exists).
There are then four categories that are labeled routine, engineering, craft,
and nonroutine. See Figure 7.1, which places the categories on the two
dimensions.
Routine technologies are well defined and have few exceptions. At least
one of formalization and centralization is high, and perhaps both are. Tasks
are well defined and understood. Control is obtained through the applica-
tion of rules. Perrow (1967, p. 204) notes, "Given a routine technology, the
much maligned Weberian bureaucracy probably constitutes the socially
7.2 A Literature Review 215

optimum form of organizational structure." Standardized mass production

is a well known example.

/
/
/
/
III-Defined Craft Nonrouine
/
/
/
/
Problem Analyzability v
/
/
/
Well-Defined RoutinE! Engineering
/
/
/
/
/

Few Exceptions Many Exceptions

Task Variability

Figure 7.1. Perrow's Technology Dimensions

Source: Adapted from Perrow (1967, p_ 196)

Engineering is also well defined but involves many exceptions. Tasks can
vary in detail, but they are readily dealt with. Formalization is low, and
centralization high. Control requires communication through reports and
meetings. There is a good deal of task variability, which requires that
adjustments be made. Engineering problems tend to be in this category, as
well as accounting and perhaps medicine.
Craft contains few exceptions but is ill-defined. The tasks themselves
are difficult and involve few alternatives. Individual skill and experience
are quite important to deal with this technology. Formalization is not high,
and centralization is low. Initial training is quite important_ If coordination
is required, it usually is attained through meetings where complexity can
be dealt with. Ed May renovates old cars in Durham, North Carolina. He
works within a narrow set of tasks, but each car needs special treatment
and exactly what should be done is not known in advance. Ed has great
216 Chapter 7. Technology

skills, many years of experience, and Burton's sports FIAT just looks great
after the renovation.
Finally, nonroutine has many exceptions and is ill-defined. Highly
skilled professionals are needed. Formalization and centralization are low.
Coordination is obtained through meetings, and group norms provide the
controL Basic research is one example.
Here, technology can be reduced from routine to nonroutine-roughly
following diagonal in Figure 7.1. The information-processing demand in-
creases as we progress from routine to nonroutine. Routine technologies
require higher formalization and centralization than nonroutine technolo-
gies. Similarly, rules can be used for coordination for a routine technology
whereas the give and take of meetings is required for nonroutine technolo-
gies. This was the concern of the management ofMed Electronic mentioned
in the introduction. It planned a move from a nonroutine technology to a
routine technology along with an increase in the production quantity. To
process the required information and maintain high quality in standard
production, rules and standardized behavior is necessary. Centralized
decision making at the overall level will also be appropriate.
Perrow (1967, p. 207), in comparing his categories to Woodward's,
speculates that unit production is probably nonroutine production; mass
involves routine production, and process is not easy to categorize.
Scott (1998, p. 228-230) describes and measures technology along three
dimensions:

• Complexity, diversity: Number of different items requiring simultaneous

consideration,
• Uncertainty or unpredictability: Can the work be predicted in advance?
• Interdependence: Does a change in one item necessitate a change in a
different item?

As Scott (p. 257) summarizes: "Complexity, uncertainty, and interdepen-

dence are alike in at least one respect: each increases the amount of infor-
mation that must be processed during the course of a task performance.
Thus, as complexity, uncertainty, and interdependence increase, structural
modifications need to be made that will either (1) reduce the need for
information processing-for example, by lowering the level of interde-
pendence or by lowering performance standards-or (2) increase the capacity
of the information-processing system, by increasing the channel and node
capacity of the hierarchy or by legitimating lateral connections among
participants. "
In brief, greater variability in technology requires greater information
processing capacity. This increased capacity can come from highly skilled
professionals. But modem production technologies also may be used. CAD-
7.2 A Literature Review 217

CAM, flexible production methods, and robotics are all technologies that
can be introduced.
It is interesting to juxtapose Scott's three characteristics of technology
(complexity, interdependence and uncertainty) with our earlier environ-
mental characteristics (complexity, uncertainty and equivocality). We
suggest the correspondence shown in Table 7.2.

Table 7.2. Environmental-Technology Characteristics

Environmental Characteristics Technology Characteristics

Complexity Complexity, interdependence

Uncertainty Uncertainty
Equivocality Uncertainty

Environmental complexity definition captures the same aspects as

technology complexity and interdependence. Technology uncertainty covers
the concepts of both uncertainty and equivocality. Contrary, however, to the
measurement of the environment, the empirical research on technology-
structure relationships does not make it easy to separate results into the
two groups. The underlying concept behind the correspondence is informa-
tion processing: both the environment and technology require management
to cope with the requisite information to meet the organizational goals.
Galbraith (1974) argued that "the greater the uncertainty of the task, the
greater the amount of information that has to be processed." The task is
shorthand for the organizational goals of delivering a product or service in
an efficient manner. Uncertainty can arise from the environment, or
technology; whatever the origin of the uncertainty, management does deal
with it. Stated differently, it is the lack of information that is fundamen-
tal-not its origin in the environment or technology. Thus, at a more
abstract level, environmental and technological uncertainty are fundamen-
tally the same.
Technology also can be defined as the transformation of inputs into
outputs. This general definition also is very operational and provides a
guide on what to measure. The input-output matrix captures all three
elements: inputs or resources, outputs or results, and transformation (or
how inputs are utilized to yield outputs). An activity model of the economy,
the firm and organization contains an input-output matrix ofthe technolog-
ical relations. The input-output matrix is illustrated below in Figure 7.2,
where the inputs are listed down the left side: employees by skill level,
resources by kind, and other required inputs. The outputs are listed along
the top of the matrix: automobiles, trucks, accounting services, and
consulting services. The outputs can be products or services for any
218 Chapter 7. Technology

organization. Consider the automobile column in the matrix; it is a list of

the input resources required to produce one automobile: 30 hours of
autoworker time and .3 hours of the assembly time per automobile and no
other inputs. Similarly, a consulting assignment may require 150 hours of
a skilled consultant's time, 10 hours of computer time, and 400 hours of
technical analysis by specialists. Each column is a list of the inputs
required to yield one unit of the output. It is a list of ingredients. Although
not explicit in the matrix, there is a complementary statement that gives
the process to convert inputs into output; this is also part of the statement
of the technology.

OUTPUTS

Automobile Truck Accounting Consulting

I Employees 30 hours of 150 hours of a

by skill auto work- consultant
N level er time

P Resources .3 hours 10 hours on

by kind assembly- the computer
U line time

T Other 0 400 hours of

required technical
S inputs analysis

Figure 7.2. Technology: An Input-Output-Matrix

The input-output matrix can have various properties. Consider two extreme
cases. First, assume each and every output utilizes every input-that is,
there is an entry in each and every cell in the matrix, which is a very dense
matrix. This dense technology is very complex and interdependent.
Referring to Scott's (1992) concepts of complexity it includes diversity and
interdependency. Diversity is the number of different items requiring
simultaneous consideration and interdependency refers to a change in one
item that necessitates a change in another.
Let us assume that the inputs are in limited supply-for example, there
are 600 autoworkers. Then a change in the number of automobiles
produced decreases the feasible number of trucks. In addition, the
assignment of assembly-line time between automobiles and trucks must be
adjusted. If the matrix has no zeroes, then all inputs affect all outputs
requiring a simultaneous consideration. The interdependency is very high;
any change in an output necessitates a change in another. Second, let us
consider the opposite extreme situation. Assume that only one input is used
for each output. The input-output row has all zero entries, except that each
7.2 A Literature Review 219

row and each column have one and only one nonzero entry. This is a very
sparse matrix. The technology is very diverse and very independent. There
are no simultaneous considerations, and any item can be changed without
affecting another item.
Most technologies fall between these two extreme situations. The first
very dense technology cannot be broken into pieces or be decomposed; the
second can be. The decomposability of a technology is the degree to which
it can be broken into relatively separate parts. The very dense matrix is not
decomposable; the sparse matrix is more decomposable. We can relate the
decomposability of the input-output matrix to Scott's (1992) complexity-
diversity and interdependency measures in Table 7.3.

Table 7.3. Technology Correspondence

Input-output Scott's Measures Decomposability

Matrix
Sparse - Diverse, independent Perfectly decomposable
many zeros
Dense - Complex, interdependent Nondecomposable
few zeros

The technology decomposability is a major contingency for organizational

design. Burton and Obel (1984) investigated the effect of technological
decomposability on the efficiency of two organizational configurations-the
divisional and the functional. Divisional configuration is relatively more
efficient than functional configuration for a more decomposable technology.
Thus, as technological diversity and independence increase, the divisional
configuration is more likely to be recommended. As diversity increases and
independence increases, the divisional configuration becomes relatively
more efficient; when complexity increases and interdependence increases,
the functional configuration is more likely to be recommended.
Uncertainty is also a major technology characteristic. It is Scott's third
technology dimension. Uncertainty is degree of predictability. In the input-
output description oftechnology, technology uncertainty is reflected by the
uncertainty of the input-output elements in the matrix. Consider the time
required for an autoworker to make one automobile. If it is known and
certain, the uncertainty is low. If this time has high variance, then the
uncertainty is high. Technological uncertainty is high when the input-
output elements have high variance, and uncertainty is low for low
variances. The equivocality relates to Perrow's measure of"ill-definedness"
(Daft and Lengel, 1986). It may not be known which resources are needed
to accomplish the task.
220 Chapter 7. Technology

Technology has been discussed by many researchers. Thompson (1967,

pp. 14-16) considers technology in terms oflong linked or serial interde-
pendence, mediating or intensive. Interdependency of technology (pp.
54-55) is described as pooled, sequential or reciprocal. Coordination is
achieved, respectively, by standardization, plan, and mutual adjustment.
Technology has been, then, operationalized in many ways. We have
reviewed the schemes devised by Woodward and Perrow and commented
briefly on Scott's and Thompson's discussion and its relation to the
decomposability scheme. We found that Woodward's unit, mass, and
continuous categories and Perrow's routineness concept are appropriate
operationalizations and measures. Scott's notion of diversity is also utilized.
Most of the research has directly or indirectly been related to production
technology, although Perrow's (1967) measure applies both to production
and service technology. From an information processing point of view, it
does not matter whether it is production or service. The concepts can be
applied equally well as we argue in the next section. However, the
information capacity demand and the means to meet that demand may
differ in a production technology and in a service technology. Similar
arguments can be made for retail and wholesale. To distinguish between
the type oftechnology is therefore appropriate. Next, we consider technol-
ogy and its effect on the configuration and other design variables.

7.3 Technology as a Contingency

7.3.1 Measuring Technology
Technology is the information, equipment, techniques, and processes
required to transform inputs into outputs. The above survey of the
literature shows, as was the case with both size and environment, that
there is not one agreed concept and measure for technology. However, there
are common traits in the technology concepts.
We have found that measuring technology along the following four
dimensions seems appropriate:

• Manufacturing, service, retail, and wholesale,

• Unit, mass, and process (automation),
• Routine or nonroutine, and
• High or low divisibility.

These dimensions capture most of the measures that have been discussed
in the literature and have proven useful in Organizational Consultant. The
7.8 Technology as a Contingency 221

combination of these four dimensions gives 4x3x2x2 = 48 different

technology categories. In Organizational Consultant, we use a finer scale
for the routine and nonroutine and the divisibility dimensions. Our main
dimensions are routineness and divisibility. The two other dimensions are
used to further strengthen our hypotheses. From the description below it
can also be seen that there is some correlation between the measures. This
is particularly true for routineness and unit, mass, and process production
as discussed above. The measurement oftechnology is relatively simple and
yet allows for enough variation to be practical.
The first dimension-manufacturing, service, retail, and wholesale-gives
a broad type of business indicator. It puts the organization in a technology
set. The second dimension follows Woodward's clarification: unit, mass, and
process (automation). Automation is like a process technology, as we argue
later. Technological routineness or nonroutineness is the third dimension.
Scott's uncertainty measure is aligned, but routineness is empirically the
dominant dimension in Perrow's framework related both to uncertainty and
equivocality. Finally, divisibility is similar to decomposability and the
opposite of Scott's interdependence measure. Thus, our technology
descriptors and measurements are similar to existing concepts.
The type of business in which the organization operates seems to make
a difference. The type of business-manufacturing, service, retail, whole-
sale-is important in determining technology. There is evidence that studies
of manufacturing may not be applicable to service firms (Robbins, 1990, p.
198). Every manufactured product ultimately becomes a service, but not
vice versa-for example, an automobile is a manufactured product that
renders a service. General Motors is a manufacturer. Avis is a service
corporation. Some services may, however, involve a few products.
The following definitions have been used:

• A manufacturing firm. A firm whose primary activity is the conversion

of raw materials, parts and subassemblies into finished products for use
by a customer. The manufacturing transformation requires equipment
and personnel. It may be either capital or labor intensive. Firms that
make automobiles, toys, and computers are examples.
• A service firm. A firm whose primary activity is the delivery of a service
directly to a customer or client. In the final stages, individuals usually
deliver the service. Banking is a service which usually involves direct
face to face delivery, but automatic teller machines (ATMs) are an
exception. Service requires both capital and labor. Many services are
labor intensive, but others are not: telecommunications services, for
example, are very capital intensive. Service firms include restaurants,
telecommunications, banking, accounting, health care, and dry cleaning.
222 Chapter 7. Technology

The distinction between a manufactured product and a service is not

always obvious. If you buy an automobile, it is a manufactured product. If
you lease the same automobile, it is a service. Consider the distinction
between cooking a frozen pizza at home and eating a pizza in a restaurant.
The later is a service incorporating a product. Most manufactured products
ultimately provide a service, and some services involve a product. Although
the distinction between a product and a service is sometimes unclear, the
distinction between a manufacturer and a service firm is more apparent.
General Motors manufactures automobiles; Avis rents them. The first is
clearly a manufacturer; the second is clearly a service firm. Additionally,
we distinguish between retail and wholesale firms.

• A retail firm. A firm whose primary activity is the sale of a product or

service directly to a customer who is the final user. Usually, retailers do
not engage in the manufacture of products. Kroger, Bilka and Wal-Mart
are examples.
• A wholesale firm. A firm whose primary activity is the sale of a product
or service to another firm but that does not sell directly to the final user.
Wholesalers do not engage in the manufacture of products. Wholesalers
are less well known. Nonetheless, Coca-Cola sells mostly wholesale as
do many famous brand names.

Working only with manufacturing firms, Woodward (1965) based her

empirical analysis on this type of categorization and found structural
differences between the high and low performers. Her categorization has
solely been associated with manufacturing. Mintzberg (1979 p. 258), in his
discussion of Woodward's categorization, found that the characteristic
process, per se, is not important but that the process industry is highly
automated. The unit production is specialized customer-oriented produc-
tion; the mass production is labor-intensive production of standardized
products, while the process production is highly automated with less labor-
intensive production. Woodward's categories describe the information
demands from various technologies. With this interpretation the categori-
zation can also be used for service, retail and wholesale organizations. Joan
Woodward's categories were

• Mass production: Large batch or mass-produced technology.

• Process production: Highly controlled, standardized, and continuous-
processing technology.
• Unit production: Technology in which units are custom made and
work is nonroutine.
7.3 Technology as a Contingency 223

In the service industry mass production corresponds to standard high-

volume services, such as food chains, airlines, and hotel chains. The unit
production for the service industry is a service that is tailor made for the
customer. Hairdressing, medical services, and specialized portfolio manage-
ment are examples. The process corresponds to highly automated service
industries. These include automated bank tellers, automated telephone
service, cable television, and services using advanced electronics and
computers without employing many people.
For retail, the standard high-volume retail with high labor intensity
would correspond to mass production. An off-the-rack suit is an example of
a standard product sold in this manner, a tailor-made suit is an example of
unit production in the retail industry. A specialty audiostore that combines
the set to match the particular needs of the customer is another example.
The retail process corresponds to stores where the customer does not get
advice and where operations are highly automated. Special supermarkets,
such as Wal-Mart, are examples of such stores. Similar examples can be
given for the wholesale industry. It is clear that Woodward's categories go
beyond manufacturing and include these services. This is particularly true
with respect to the skill level required by employees, as was shown in Table
7.1.
Routineness is a central concept in technology. Analyzability and
uncertainty also have been used. However, a majority of researchers have
used the concept routineness (Robbins, 1990; Miller, Glick, Wang, and
Huber, 1991). Similar to Perrow's (1967) original concept, we define
routineness as follows:

• Routine technology: Contains easy-to-analyze problems and few

exceptions.
• Nonroutine technology: Contains difficult-to-resolve problems and
many exceptions.

A number of more precise definitions of routineness were discussed by

Miller, Glick, Wang, and Huber (1991), and they found that, in general,
relationships between routineness and structure do not depend on the
particular definition of routineness used.
As we discussed earlier, routineness and uncertainty are similar. There
is little uncertainty about a routine technology but a nonroutine technology
presents greater uncertainty. Med Electronics plans to move from
nonroutine unit production to routine mass production, changing both
complexity and uncertainty related to technology.
Technological divisibility is also a central concept. Divisibility is related
to interdependency and decomposability. The less interdependent the
224 Chapter 7. Technology

technology is, the greater the divisibility; the greater the decomposability,
the greater the divisibility
Divisibility then is the degree to which tasks can be divided into smaller,
relatively independent tasks. Returning to our earlier discussion, a highly
divisible technology has a sparse input-output matrix, is decomposable and
independent. There are a number of dimensions to consider: technology can
be divisible across the functions of research and development, production,
marketing, and advertising. Technology can also be divisible across
products. For international organizations, the degree of divisibility across
various countries is important. The technological divisibility is a measure
of these various dimensions. In more technical terms, increasing returns to
scale in R&D, production, or marketing leads to a low degree oftechnologi-
cal divisibility. Satellite television that broadcasts to many countries may
decrease the divisibility of the marketing technology when TV commercials
are used. A Euro Sport commercial will reach most European countries and
require that items advertised are ready for sales in the various countries
at the same time. When LEGO introduces a new design in Europe, the
design is available in approximately 40,000 stores on the same day. This
demands high coordination within the complete logistic operations of the
company. Economies of scope can have a similar effect. The R&D of
telecommunications switches, for example, is not divisible due to the high
cost of development; there are economies of scale and scope. Some
organizations may have high divisibility of technology today, but future
technologies may not.

7.3.2 Technology as a Contingency Factor

The various dimensions of technology have an effect on the organizational
design. We consider technology's effect on formalization, centralization,
complexity, configuration, coordination and control, and incentives.

Table 7.4. Technology Effects on Formalization, Propositions 7.1-7.6

7.1. If technology routineness is low, then formalization should be low.

7.2. If technology routineness is high, then formalization should be high.

7.3. If the organization employs many professionals, then Proposition 7.2 is

not so strong.
7.8 Technology as a Contingency 225

7.4. If the organization is in the service industry, then the strength of

Propositions 7.1 and 7.2 is greater than ifit is in the manufacturing
industry. Retail and wholesale organizations can be expected to fall in
between.

7.5. If the technology type is process (i.e., high automation), then formaliza-
tion should be higher than it would be otherwise.

7.6. If the organization uses modern information technology, then formal-

ization should be high.

7.3.2.1 Technology Effects on Formalization

In Table 7.4, Propositions 7.1 and 7.2 relate technology routineness to
formalization. The more routine the technology, the more the activities are
predictable. Exceptions are few and easy to resolve. Less information needs
to be processed. With a high routineness, it is advantageous and efficient
to establish rules and a program to regulate and coordinate the work
(Perrow, 1967, pp. 199-200). When routineness is low, such rules and
programs are likely to be incorrect much of the time. Thus, a good deal of
information will need to be processed to schedule and coordinate processes.
The relationship has been questioned (Robbins, 1990) but it has obtained
empirical support by many including the metaanalysis by Miller, Glick,
Wang, and Huber (1991).
Miller, Glick, Wang, and Huber (1991) also found that the strength of
the relationship was modified by two factors. These are incorporated into
Propositions 7.3 and 7.4. First, an organization with many professionals
has a mitigating effect on routineness and formalization. The argument is
that professionalization and formalization are alternative forms for
coordination and control so that when one is in place, the effect of the other
vanishes. The important factor is standardized behavior for the organiza-
tion. There are more means to obtain such behavior.
Second, Miller, Glick, Wang, and Huber (1991) found that the type of
industry influenced the strength of Propositions 7.1 and 7.2. They argued
that one would expect that routineness of the technology was more
positively related to centralization, formalization, and specialization in
manufacturing organizations than in service organizations. However, their
metaanalysis showed exactly the opposite. Jackson and Morgan (1978, p.
196) argue that the reason for such findings may be that in manufacturing
the production process is linked to machines and their performance, and
these machines indirectly introduce standardization. In service organiza-
226 Chapter 7. Technology

tions more rules and procedures are needed to obtain the same level of
standardized behavior. Additionally, in manufacturing the quality control
related to the process may secure high quality of the products while rules
are needed to obtain high quality in service organizations. This implies a
higher reliance on formalization and centralization. This also fits the view
expressed in Mills and Moberg (1982) on the differences between manufac-
turing and service technologies.
This kind of analysis leads to Proposition 7.5, which indicates that
highly automated technology should have a greater formalization than
otherwise suggested. This proposition is also supported by Child (1973, p.
183), who states that automation leads to higher formalization. However,
it is directly opposed to Woodward's original results (see Table 7.1). She
found that process production had low formalization. The difference may be
attributed to the difference in size of mass production firms and process
production firms (Hickson, Pugh, and Pheysey, 1979). The mass production
firms were generally larger than the process production firms; therefore,
from a size argument, formalization was higher in the mass production
firms than in the process production firms.
Automation is related to the use of computers and information
technology. Zeffane (1989) found that such use would increase formaliza-
tion. We posit that if the introduction of modern information technology is
not followed by standard rules on how to use it, the likelihood of inefficient
operations is high. We therefore suggest Proposition 7.6. Such standardiza-
tion also may have negative side effects. We all have received letters from
companies urging us to pay their bills when we have done so some days
before. Ifthe company's computer has a standard procedure writing letters
to all those who have not paid within a given deadline, then it may not be
possible to alter procedures.

7.3.2.2 Technology Effects on Centralization

In Table 7.5 the relationship between technology routineness and central-
ization is given in Proposition 7.7 and 7.8. These incorporate issues of size
as well as technology. Ifthe organization is small and has a technology that
is very routine, then the manager can more easily assess the operations
than if the routineness was low. A manager can handle the required
information. Therefore, the argument that for small organizations the
centralization should be high is further strengthened.
When the organization is large and has a technology that is routine,
then it is very likely both from the size paradigm and from Proposition 7.2
that formalization should be high. A control and coordination mechanism,
therefore, is in place (Zeffane, 1989). The size argument presented in
7.3 Technology as a Contingency 227

Chapter 5 that for large organizations centralization should be low is

therefore further strengthened. Propositions 7.7 and 7.8 are supported by
the metaanalysis by Miller, Glick, Wang, and Huber, (1991).

Table 7.5. Technology Effects on Centralization, Propositions 7.7-7.8

7.7. If technology routineness is high and the size of the organization is

small, then centralization should be high.

7.B. If the organization is large and technology routineness is high, then

centralization should be medium.

7.3.2.3 Technology Effects on Organizational

Complexity
The relationship between technology and complexity is not simple either,
as given in Propositions 7.9 and 7.10 (see Table 7.6). Size is a moderator
again. Generally, the larger organization with a routine technology is more
complex.
The argument is that large organizations can better specialize and,
therefore, use the routine technology to create experts for each specialty;
horizontal differentiation increases. Propositions 7.9 and 7.10 are partly
supported by Miller, Glick, Wang, and Huber, (1991). However, a reverse
argument also can be made. If the technology is nonroutine, then the work
is very complex, and it is likely that the appropriate span of control is low.
Therefore complexity increases vertical differentiation particularly. We
therefore state in Proposition 7.11 that for a large organization with a
nonroutine technology, complexity should be high-particularly vertical
differentiation.

Table 7.6. Technology Effects on Organizational Complexity, Propositions

7.9-7.14

7.9. If the size of the organization is large and the organization has a
technology that is routine, then complexity should be high-particular-
ly horizontal differentiation.

7.10. If the size of the organization is small and the organization has a
technology that is routine, then complexity should be medium.
228 Chapter 7. Technology

7.11. If the size of the organization is large and has a nonroutine technol-
ogy, then complexity should be high-particularly vertical differentia-
tion.

7.12. If the organization has a nonroutine technology, then the span of

control should be narrow.

7.13. If the organization has a routine technology, then the span of control
should be wide.

7.14. If the technology type is process (high automation), then complexity is

high.

Both Propositions 7.10 and 7.11 were based on a span-of-control argument;

because supervision has limited information-processing capacity and can
deal with a limited number of issues or exceptions, a nonroutine technology
will yield. These arguments are formalized in Propositions 7.12 and 7.13.
The argument is that the more complex the work, the less people a
manager can supervise and control. This is a bounded-rationality argument
and is widely supported (Robbins, 1990). That both nonroutine and routine
technology may lead to high complexity, but for different reasons, may
explain why Miller, Glick, Wang, and Huber, (1991), in their metaanalysis
using averages, did not find a significant relationship between routineness
and specialization. There may be other technology-based reasons than
routineness that may lead to structural conclusions.
Since process organizations use more automation and more skilled
personnel, the span-of-control argument suggests that process organiza-
tions are more complex than other types. This is Proposition 7.14. However,
Woodward (1965) (see Table 7.1) found the opposite in her research on
manufacturing firms. This contrast also may be related to the discussion
about the effect of size. Woodward's results have been criticized because it
was argued that when controlled for size her results disappeared (Hickson,
Pugh, and Pheysey, 1979). Because organizations that use a process
technology are less labor intensive, they tend to be of smaller size. In most
cases Proposition 5.4 in Chapter 5 will therefore be activated and recom-
mend a low complexity for a highly automated organization. Proposition
7.14 does not take the size of the organization into consideration. The
balancing of the two propositions will result in the correct recommendation
for the particular organization. The balancing issue is treated in more
depth in Chapter 9.
7.3 Technology as a Contingency 229

7.3.2.4 Technology Effects on Configuration

Technology also affects the configuration in many ways (see Table 7.7).
Proposition 7.15 indicates that a unit technology is more likely to require
a matrix organization. One reason for this is that it may be needed to
assign experienced and skilled personnel from one production unit to the
next. This sharing of valued and limited resources requires on-line
coordination, which can be realized in a matrix structure.
We now turn to a number of mismatches between technology and some
configurations. Proposition 7.16 indicates that a functional configuration
for a nonroutine technology is not likely to be efficient because a functional
structural requires high horizontal differentiation, which may be unlikely
for a nonroutine technology. And it certainly will require a lot of cross-
function coordination, which the functional configuration will not do in a
timely fashion.
A differentlimitation is given in Proposition 7.17; a divisional configura-
tion and a nondivisible technology is a mismatch. Divisional organizations
require that the task be divided and placed in each division. Since these
di~sions are relatively autonomous, a high degree of interrelationship
between them is costly to coordinate. On the other hand, a matrix structure
is not needed for a divisible technology, as given in Proposition 7.18. The
argument is as follows. If the technology is divisible, then the work can be
separated into units that are not dependent. A high level of coordination is,
therefore, not required due to technological reasons. A matrix structure
with its lateral relations for coordination is too costly, and there is little to
coordinate.

Table 7.7. Effects on Configuration, Propositions 7.15-7.20

7.15. If the technology type is unit, then it is more likely that the organiza-
tion has a matrix configuration.

7.16. If the organization has a nonroutine technology, then the functional

configuration is not likely to be an efficient configuration.

7.17. If the technology is not divisible, then the configuration cannot be divi-
sional.

7. lB. If the technology is divisible, then it is not very likely that the
configuration should be a matrix configuration.
230 Cha.pter 7. Technology

7.19. [fthe organization has a nonroutine technology, then it is not likely

that a machine or professional bureaucracy is an efficient configura-
tion.

7.20. [fthe technology is not nonroutine, then the configuration cannot be

an ad hoc configuration.

A bureaucracy requires standard behavior either through the use of rules

or the use of professionals. Rules are very likely to obstruct needed
adjustments for a nonroutine technology. This is expressed in Proposition
7.19.
Finally, adhocracies are costly to coordinate and can operate only where
the uncertainty related to the tasks is relatively high. Therefore, an ad hoc
configuration cannot operate if the technology is very routine and will not
operate efficiently.

Information Technology1
Information technology includes computers, e-mail, voice mail, video-conferencing, databases,
expert systems, and other electronic means to store, analyze, move or communicate
information in an organization. Information technology is then a means for an organization to
process information. The organization itself is an information processing entity and thus,
information technology is a means for the organization to accomplish its fundamental work. Of
course, there are many other non electronic means for the organization to process information:
pencil and paper calculations, face to face conversations, paper memoranda, etc. Here, we
want to explore the implications that the organizational design has for the choice of the
information technology.
The connection between the organizational configuration and properties and the organiza-
tion's information technology has a long tradition and vast literature. As Hunter (1998)
suggests, most of the research and studies have focused on the influence of information
technology, computers, email, etc., on the structure, properties, behavior and performance of
the organization, i.e., the information technology is the independent variable and its effect is the
dependent variable (Huber, 1990, Malone and Rockart, 1990). More recent research has taken
a new approach of advanced structuration theory (DeSanctis & Poole, 1994) which focus on
the complex interactions between the organizational actors and the information technology.
The emergent behavior is uncertain and difficult to predict; the research goal is to describe
and understand the interplay between the organization and the information technology without
resorting to an independent variable, dependent variable approach. Here, we want to focus on
the information technology design or choice question, i.e., what information technology should
be adopted by the organization to be compatible with the organizational configuration and the
organizational properties. This switches the independent variable and dependent variable so
that the organizational design is the independent variable and the information system is the
dependent variable.
The organizational design question more directly addresses the managerial question of
"What kind of information system do I need to fit with my organization?" Consequently, we want
to consider what the information technology should be. This seems reasonable, but the most
research questions have not been posited in this manner and thus the empirical evidence is
wanting. There is a good deal of research on the effect on decentralization when an advanced
e-mail system is introduced; however, it is not conclusive.
7.8 Technology as a Contingency 231

The main theme of this book is that the organization is an information processing entity, i.e.,
the organizational task is accomplished by processing information: gathering data, analyzing
information, deciding what to do, communicating information, implementing and controlling
events, measuring events and results. Individuals talk to each other in the hallway, they write
e-mails, they go to meetings, etc. Some organizations have lots of meetings, but discourage
hallway conversations. The list of possibilities is long. The organizational design then helps
rationalize and organize how the information will be processed. Eg, a decentralized organization
processes information differently than a centralized organization to accomplish the sarne
organizational task. We suggest that a decentralized organization will use e-mail differently than
a centralized organization. "Who makes what decision when" is different and we suggest that
the content and frequency of the e-mail would be different. But the e-mail is only a small part
of the organization, and indeed, the information, i.e., the electronic information system is only
a small part of the organizational information processing. Yet, the electronic information system
is an important part of the total organization and it should fit with the rest of the organization.
In Chapter 6 on the Environment, we introduced the concept of media richness: a high media
richness requires that information comes from many sources, many formats, and probably in
large amounts. A low media richness requires much less information from the environment. For
high media richness, the information system is most likely to have many elements: e-mail, list
servers, the web, news services, trade services, telephones with voice mails and ready access
to the outside, multiple formats to receive information and then internally, some information
processing capability to make sense of the diverse and large amount of information. In brief,
the information system should support the gathering of vast and diverse information and also
support its interpretation and meaning for the organization.
A low media richness need can be met with a focused single information system, eg, in a low
complexity environment, low uncertainty and low equivocality, it may be sufficient to look only
at last year's sales. However, this would be very inadequate for a high media richness need.
As we suggested above, the centralization of the organization also affects the choice of the
information technology. Most of the studies consider the opposite question of what is the effect
of a given information element, eg, e-mail on the centralization. Huber (1990, p. 57) suggests
that the greater use of information technology will be mediating: highly centralized organizations
will become more decentralized, and highly decentralized organizations will become more
centralized. This is insightful and summarizes what we know; yet, we can not say definitively
what the effect of the information technology will be on centralization of the organization.
But let us consider the design question, if we want to obtain a decentralized organization,
what kind of information technology do we need? A decentralized organization means that the
decision making is at a low level in the organization. If the relevant information is local and
nearby, then a decentralized organization would not require an advanced information system;
the individuals can look at the situation and talk among themselves to make the decisions. Here
a centralized organization would require the gathering and transmission of the same information
up the hierarchy to make the same decision and hence an advanced information system could
be quite helpful. On the other hand, if the relevant information for decision making is extemal
and widely dispersed, then an advanced information system would support decentralized
decision making. So, the decentralization, advanced information system question requires
information about the source of the relevant information as well.
The organizational design question can be stated: for the organizational configuration and
organizational properties, what is an appropriate information technology? Here we want to
explore some plausible responses to this design question. We will state a few propositions and
then argue for their reasonableness. The supporting arguments do have some research
support, but their validity rests primarily upon their reasonableness.
A highly formalized organization has a large number of rules and standardized routines. What
kind of information technology is appropriate ? Without an electronic information technology,
the formalization was contained in written documents and the rules were realized through the
expertise of the employees who followed the rules. With an electronic information technology,
it is possible to incorporate the rules directly into the operations of the organization through
quick reference or actual control. Many airline reservation systems incorporate detailed rules
and many airline personnel are permitted less discretion than in older paper systems. A highly
232 Chapter 7. Technology

formalized organization can use information technology to operationalize the rules and control
implementation.
A low formalization has few rules. What is an appropriate information technology here? With
few rules, information about the environment, customers, competitors, technology, scheduling,
etc. becomes information for decision making and coordination. Would electronic technology
such as e-mail, voice mail, bulletin boards, shared databases, etc. be helpful? Yes, it seems
reasonable. Here too, an electronic information technology could be helpful.
So we conclude that an electronic information technology can be appropriate for both high
or low formalization. However, there is a difference in detail in what the information technology
does: for a high formalization, the rules are incorporated; for a low formalized, information is
widely made available throughout the organization. Electronic information technology is a
mediating technology; it supports the information processing demands of the organization. In
the above discussion, the discriminating variable is the content of the information technology
and what we want it to do. So, we must conclude that the information technology is not the
primary concern, but what it does is the main issue: how does it help (or, hinder) the
organizational demand for information processing capability.
We can now state design propositions:
If the organizational formalization is high, then the information technology should incorporate
the rules and routines.
If the organizational formalization is low, then the information technology should augment the
availability of information through e-mail, voice mail and shared databases.
Whether a given organizational design increases or decreases the need for information
technology is not the issue. The question is: what kinds of information technology support the
organizational design? Let us explore some other propositions.
If the organizational configuration is matrix, then the information technology should be e-mail,
voice mail, video-conferencing, and shared databases.
A matrix organization requires a good deal of give-and-take and adaptation to circumstances.
The information processing demands are high, both in quantity and type. In addition to face to
face meetings, telephone calls, etc., these information technologies can augment the
availability and communication of information to support the matrix design.
Equally, an information technology which is locked in rules and restricted format may well
hinder the matrix design.
If the organizational complexity is low and particularly the vertical differentiation is low, then
the information technology should facilitate quick hierarchical flow and the aggregation of
information.
A low vertical differentiation means that there are few levels in the hierarchy-top to bottom.
There is no large middle management or the organization is "delayered'" In information terms,
middle management aggregated information as they passed it up, and disaggregated
information as they passed it down. And frequently, they simply passed the information on
without modification. With low vertical differentiation, the information technology should respond
to this information requirement. The specific information technology would also depend upon
the organizational degree of centralization as discussed above.
We could add a number of other propositions on other organizational design configurations
and properties. The supporting arguments must emerge from the information demands of the
organization and how the information technology will aid the particular information needs. It is
too easy to say that we need more information technology. It begs the question: to do what.
Different information technologies are required for different organizational designs.

7.3.2.5 Technology Effects on Coordination and Control

Mechanisms
Propositions 7.21 and 7.22 relate the technology routineness to the recom-
mendations on coordination, media richness, and incentives (Table 7.8).
7.3 Technology as a Contingency 233

Generally, the propositions state that with more routine technology, more
rule-oriented coordination, and less rich media, the incentives can be
procedure based. In contrast, less routine technology calls for coordination
by integrators and group meetings using richer media and results-based
incentives. The supporting arguments are fundamentally information
processing in nature similar to the ones presented in Chapter 6. A routine
technology does not change much. Activities are largely known and can be
planned. There is little new, detailed, or current need for information. The
information is well defined, known for some time, and likely to be numeri-
cal: production quantities, product dimensions, and so on. Incentives can be
procedure based as procedures are known and well defined and the
challenge is to follow them correctly.
The nonroutine technology, in contrast, calls for a large amount of
information. There are many issues to decide, implement, and control. The
products and the procedures themselves are likely to change often.
Galbraith (1974), in his information processing framework, suggested that
the information processing requirements are large for this situation.
Integrators and frequent group meetings are appropriate organizational
strategies to obtain the required coordination. Daft (1992, p. 290) and Daft
and Lengel (1986) argue that relatively rich media will be required to deal
with an ambiguous situation when much is unknown and is to be discov-
ered during the decision-making phase. A nonroutine technology could also
be described as equivocal and uncertain, terms that we used to describe the
environment in Chapter 6. The need for rich media can be realized in a
number of ways. Face-to-face is the richest, which is the medium of
integrators and group meetings. That is, integrators and group meetings
are rich media provided there is truly discussion,joint problem solving, and
a give-and-take atmosphere. Integrators who simply tell and meetings that
only inform will not work. Finally, incentives must fit the routineness of the
technology and the other organizational design recommendations. A routine
technology indicates that we know what to do; the incentive is to do it-that
is, a procedural-based incentive to follow the rules and implement the
plans. The nonroutine technology creates the opposite requirements. The
goal is to obtain results in the face of the nonroutine technology. The goal
is a working product or a satisfied customer. This result is important, and
the procedure is to be developed. Kerr (1975) argues that the best incentive
is to reward what the organization wants, and here the organization wants
results. Many organizations use many different technologies-some routine
and some nonroutine. The various technologies may push the organization
in different directions. If that is the situation, technology is not a strong
contingency on the overall recommendation relative to other contingencies.
Each technology, of course, will be an important factor for the micro design.
This is expressed in Proposition 7.23.
234 Chapter 7. Technology

Table 7.8. Technology Effects on Coordination and Control Mechanisms,

Propositions 7.21-7.23

7.21. If the size of the organization is not small and if the technology is
routine, then coordination and control should be obtained via rules
and planning, and a media with low richness and a small amount of
information can be used. Incentives should be based on procedures.

7.22. If the technology is non routine, then coordination should be obtained

via group meetings, and media with high richness and a large amount
of information should be used. Incentives should be based on results.

7.23. If the organization does not have a dominant technology, then the
technology-structure recommendation should be discounted relative to
other contingency factors.

7.4 Managing Technology

In this chapter, we have presented the rationale for the influence of
technology on the organization. However, there is a counterargument that
the organization must influence and change the technology. This is
managing technology. The organization must choose its technology. From
either point of view the organizational structure and the technology must
fit as well as technology must fit other contingencies.
Most products and services can be produced using a number of different
technologies. Management often faces a choice between a routine and
non routine technology and between mass production and a high degree of
automation. LEGO Systems used to produce its standard plastic bricks in
a kind of mass production setting with many workers operating the various
machines. The same bricks are now produced in a highly automated factory
with numerous robots and basically no workers.
The organization should manage its technology in a comprehensive
manner. Total quality management (TQM) is such a comprehensive
approach to manage technology. TQM focuses on process and product or
service innovation to improve quality continually. Process innovation in
TQM includes changing the organizational structure and the organization's
procedures to achieve quality improvement and innovation.
Earlier, we defined technology (Robbins, 1990, p. 176) as the informa-
tion, equipment, techniques, and processes required to transform inputs
into outputs. There are many ways to achieve the same output, so
7.4 Managing Technology 235

managing technology becomes important. The TQM approach assumes that

all these technology variables can be changed and should be managed to
achieve better quality. Consequently, TQM is a comprehensive approach
with an unending quest to do everything better. It requires innovation on
processes and products and services to serve the organization's customers
and clients better.
Since the early 1950s, quality management has been the mission of
Joseph Juran and W Edwards Deming. One of Deming's fourteen points
(Giffi., Roth, and Seal, 1990, p. 33) is "break down barriers between
departments; encourage different departments to work together on problem
solving." In our model, this recommendation could be a structural change
to "increase lateral communications," "use committees," "form teams," and,
perhaps, "implement a matrix or an ad hoc organization." Whatever the
particular recommendation, the goal is to empower individuals to increase
communications, define new issues, and devise ways to implement these
changes in ways not currently available to the organization. Quality circles
at the worker level and cross-functional teams throughout the organization
are frequently used organizational changes. Deming's advise may seem to
be too general and lacking specificity. It is a noncontingent maxim. It
provides direction for innovation. In a given situation, the particular con-
tingencies of the situation will yield more detailed recommendation and
advice.
TQM is a comprehensive approach to quality. In terms of technology and
structure, both are to be managed: both are variable and can be changed.
New technical processes and products may require a new organization.
Even more important, continual changes in technical processes and
products requires not only changed organizational structures but organiza-
tional structures that foster and drive technology innovation. A fundamen-
tal tenet of TQM is that quality must be improved continually; conse-
quently, technology must be ever innovated, and organizational structures
must drive this innovation. Here, change and innovation are the norm, and
the organization should drive innovation.
The need for TQM has grown out of a need to be competitive in
international and global markets. In the 1950s, the quality of Japanese
products was poor, and they were not competitive. But they changed, and
now Japanese products set the world standards for many products.
Americans, Europeans, and others have adapted TQM as a way to innovate
technology and to answer the competition challenge. The future promises
even greater global competition, which demands continual innovation.
TQM is one approach to managing technology but not the only one. Yet
its comprehensiveness suggests that most other approaches could be viewed
under the TQM umbrella, adding specificity to a general total view. Med
Electronics was considering changing its technology from a nonroutine unit
236 Chapter 7. Technology

production to a routine mass production. Because the environment had

changed, forcing the company to reconsider its strategy, Med Electronics
was trying to obtain a situation fit. A change in technology requires a
change in the overall organization. More formalization and maybe
specialization may be needed. The quality focus will also change. In the
current nonroutine situation, quality is related to creating an apparatus
that fits the needs of a particular physician. In the routine technology
situation, Med Electronics has to assume that each apparatus is very
similar and meets standard specifications with low production cost. More
focus has to be put on the procedure.
Generally the view of quality affects the organizational structure. If
defects are allowed, they have to be handled. A test center has to be created
such that only the allowed amounts of defects leave the organization. A
department that can deal with complaints may be needed, and possibly a
process for replacement or repair has to be established. If defects in
principle are not allowed, more structure and processes have to be allocated
to prevent defects and less resources can be allocated to deal with defects.

7.5 The Technology of SAS

When SAS was established in 1951, managers were engineers, and in the
following years engineers still occupied a majority ofthe company's leading
positions. Managers were recruited internally because they knew the
company from the inside and were in a good position for making decisions.
But new ideas were scarce.
From its beginning until the 1970s, SAS's goal had been to maximize the
number of passenger kilometers, which meant that the main focus was on
the transaction of moving a person from one place to another. Service, as
we know it today, was not a major concern, and therefore the passengers'
wishes were only secondary. But market demands changed. The passengers
wanted more direct departures, and it appeared that passengers were not
as concerned about plane types as employees were.
Airplane purchases had been based on the belief that the market would
be growing throughout the 1980s. But due to the oil crises in the 1970s,
market growth vanished, and overcapacity was an increasingly noticeable
phenomenon.
As a result of market overcapacity, utilization of capacity decreased.
With Carlzon's appointment in 1982, something had to be done. Carlzon
wanted to make SAS more market oriented, which meant that the
production apparatus should be adapted to customers and not the other
way around. SAS had to use smaller planes on routes to increase the
utilization of capacity.
7.5 The Technology of SAS 237

Carlzon's new strategy of making SAS the businessperson's best

alternative (chapter 8) resulted in the introduction of Euroclass-a special
product for the business market segment. Euroclass customers had
separate check-in counters, separate lounges, and hotels on the ground. In
the air, they had roomier seating with more space between the seats in the
front of the cabin.
Carlzon's philosophy was to give individuals individual treatment (SAS,
1982-1983, p. 38) and problems were to be solved as they arose (SAS
1983-1984, p. 5). Therefore, personnel had to be able to make a quick
decision to satisfy the customers. A new control system helped the front-
line personnel with various solutions to various problems. The control
system could be developed because service consisted of certain fixed
components, and consequently, customers' problems were mostly well
defined within fairly narrow categories. Within the framework of this
control system, each employee was responsible for finding a solution that
would satisfy the customer in the given situation (SAS, 1984-1985, p. 12).
The service SAS provided was no longer just transportation but also a good
experience for passengers. From a technology point of view the major
change was the change from a production orientation to a service orienta-
tion. The product was now satisfied customers.
Formalization in airlines has always been high, and that also was the
case at SAS after the change. The individual treatment resulted in a
formalization based on many fine rules, particularly for front-line person-
nel. The employee made decisions within the framework of the many fine
rules-a greater formalization with decentralization. Additionally, social
formalization increased by a massive training and education of all
employees in SAS. The basic technology of SAS did not change with the
changes in 1975, but more emphasis on the front-line personnel and service
issues did have an effect.
Overall, the technology at SAS is routine: there are few exceptions, and
the operating problems are well defined, if not easy to analyze. There are
few surprises from the technology. The technology is nondivisible: the flight
schedules themselves require integrated operations across continents with
narrow time frames. And of course, SAS is a service organization, utilizing
skilled personnel and high-technology equipment-both airplanes and
information systems.
The organizational implications for a routine, nondivisible technology,
service organization follow from Propositions 7.2, 7.4, 7.6, 7.8, 7.9, 7.13,
7.17,7.20, and 7.21. In Proposition 7.2, a routine technology calls for high
formalization. The benefits of standardized behavior leads to predictability
as well as efficiency. In part, the technology is devised to be routine to
eliminate variation and permit high levels of reliability and particularly
safety. For the organization, a high degree of formalization follows. In
238 Chapter 7. Technology

Propositions 7.4 and 7.6, we find the high formalization and an organiza-
tion with modern information technology. Airlines have very sophisticated
information systems.
The implication for centralization is given in Proposition 7.8: a large
organization with a high routine technology should have a medium
centralization. A large organization should not be centralized, and a routine
technology requires high formalization but not high centralization. There
is a tradeoffbetween formalization and centralization. Propositions 7.9 and
7.13 suggest a high complexity with high horizontal differentiation and a
wide span of control. There should be a large number of well-defined highly
specialized jobs-jet mechanics, customer check-in personnel, pilots, and so
on. SAS has a large number of specialized jobs.
Proposition 7.17 suggests that a nondivisible technology is not compati-
ble with a divisional configuration. SAS has nondivisible technology within
its airline company. Its attempt to create a geographical or a customer
divisional configuration within the airline company has been difficult. The
interdivisional externalities are real and difficult to manage. Proposition
7.20 indicates an ad hoc configuration is inappropriate, which is most
reasonable for SAS.
Finally, Proposition 7.21 indicates that a organization that is not small
with a routine technology should use rules, plans, and low media richness
for coordination and control. The incentives should be process based-that
is, how well an employee does his or her job. This recommendation is
consistent with a high formalization. SAS, despite its customer-based
strategy, has a highly routine technology; in fact, it has utilized its routine
technology to make it customer based by giving front-line personnel
decision-making responsibility to meet customer needs but within well-
defined parameters. SAS has routine technology and is customer oriented.

7.6 Summary
Technology is an important determinant for an organization's design.
Technology has been described in numerous ways:

• Production, service, retail, and wholesale,

• Unit, mass, and process,
• Routine or nonroutine, and
• Divisible or nondivisible.

Using these technology descriptions, we relate technology to organizational

design. Generally, mass, routine, divisible technologies require a low
amount of information processing; process, nonroutine and nondivisible
References 239

technologies require a greater amount of information processing; process,

nonroutine, and nondivisible technologies require a greater amount of
information for coordination and control. Consistent with an information
processing view, the recommendations on formalization, complexity,
centralization, and coordination and control schemes are developed.
Basically, mass, routine, divisible technologies require high formalization,
high complexity, and medium centralization. Process, nonroutine, and
nondivisible technologies have lesser formalization, complexity, and
centralization: but a nondivisible technology must be coordinated by some
means.

Notes
1. Starling Hunter contributed greatly to this discussion.

References
Burton, Richard M., and B0rge Obel. 1984. Designing Efficient Organizations:
Modelling and Experimentation. Amsterdam: North-Holland.
Child, John. 1973. "Predicting and Understanding Organization Structure."
Administrative Science Quarterly, 18, 168-185.
Daft, Richard L. 1992. Organization Theory and Design (4th ed.). St. Paul, MN:
West. (2nd 1986, 3rd ed.1989).
Daft, Richard, and Robert H. Lengel. 1986. "Organizational Information Require-
ments, Media Richness and Structural Design." Management Science, (32)5,
554-571.
DeSanctis, Geradine and Marshall Scott Poole. 1994. "Capturing the Complexity in
Advanced Technology Use: Adaptive Structuration Theory," Organization
Science, 5(2), 121-146.
Galbraith, Jay. R. 1974. "Organization Design: An Information Processing View."
Interfaces, 4(3), 28-36.
Giffi, Craig, Aleda V. Roth, and Gregory M. Seal. 1990. Competing in World-Class
Manufacturing: America's Twentyfirst Century Challenge. Homewood, IL: Irwin,
Inc.
http://www.sas.selsas/aboutsas/inbrie£html
Hickson, David J., D.S. Pugh and Diana C. Pheysey. 1979. "Operations Technology
and Organization Structure: An Empirical Reappraisal." Administrative Science
Quarterly, 24, 375-397.
Huber, George P. 1990. "A Theory of the Effects of Advanced Information Technolo-
gies on Organizational Design, Intelligence, and Decision Making." Academy of
Management Review, 15(1) 47 -71.
240 Chapter 7. Technology

Hunter, Starling. 1998. Information Technology & New Organization Forms.

Unpublished Doctorel Dissertation, Duke University.
Jackson, John H., and Cyril P. Morgan. 1978. Organization Theory: A Macro
Perspective for Management, Englewood Cliffs, NJ: Prentice-Hall.
Kerr, Steve. 1975. "On the Folly of Rewarding A While Hoping for B." Academy of
Management Journal, 18, 769-783.
Malone, Thomas and J. Rockart. 1990. "Computers, Networks, and the Corpora-
tion," Scientific American, 128-136.
Miller, Chet, William H. Glick, Yau-de Wang, and George P. Huber. 1991.
"Understanding Technology-Structure Relationships: Theory Development and
Meta-Analytic Theory Testing." Academy of Management Journal, 34(2),
370-399.
Mills, Peter K., and Dennis J. Moberg. 1982. "Perspectives on the Technology of
Service Operations." Academy of Management Review, 7(3), 467-478.
Mintzberg, Henry. 1979. The Structuring of Organizations. Englewood Cliffs, NJ:
Prentice-Hall.
Perrow, Charles. 1967. "A Framework for the Comparative Analysis ofOrganiza-
tion." American Sociological Review, 32(2), 144-208.
Robbins, Stephen P. 1990. Organization Theory: Structure, Design and Application.
Englewood Cliffs, NJ: Prentice-Hall.
SAS. 1982-1990. SAS Annual Reports. Stockholm: SAS.
SAS. 1994. SAS Annual Reports. Stockholm: SAS.
SAS. 1995. Annual Report. Stockholm: SAS.
SAS. 1996. Miljf;1regnskab. Stockholm: SAS
SAS. 1996. Annual Report. Stockholm: SAS
SAS. 1996. De f(Jrste 50 dr. Stockholm: SAS
SAS. 1996. ScandinatianAirlines 1946-1996. Stockholm: SAS
SAS. 1997. Deldrsrapport 1: 1997. Stockholm: SAS
Scott, W. Richard. 1998. Organizations: Rational, Natural and Open Systems.
Englewood Cliffs, NJ: Prentice-Hall.
Thompson, J.D. 1967. Organizations in Action. New York: McGraw-Hill.
Woodward, Joan. 1965. Industrial Organization Theory and Practice. Oxford: Oxford
University Press.
Zeffane, Rachid. 1989. "Computer Use and Structural Control: A Study of
Australian Enterprises." Journal of Management Studies, 26(6), 621-648.
CHAPTERS

Strategy

8.1 Introduction
When Bon Gout faced its problem with Samsonite, described in Chapter 6,
it had a number of ways to deal with the situation. It could fight the
decision made by Samsonite, or it could decide to stay in its normal
business and find a new supplier to replace Samsonite. One of its major
competitive strengths-its ability to deal with manufacturers and retail-
ers-could help it find new products to import and market. Since its
problems arose when the single European market was established, Bon
Gout could decide to play the European game and import products from
East Asia to be sold in the European market. It needed a strategy. Strategy
is both means and ends; it includes the definition of the overall end goals,
and the means of action needed to obtain these stated goals. Bon Gout must
decide what business it wants to be in, and then it has to develop a process
to realize what it wants. Environment is important for this choice. The
process to obtain certain goals may require a particular organizational
structure. For example, if Bon Gout decides that it wants to play the
European game in the high-end fashion game, it will need an organization
that can operate in more countries. This will relate to both distribution and
service to stores as well as to the ability to read fashion trends in various
countries. Additionally, it needs an organizational structure that can deal
with suppliers.
The fit between the strategy and the organizational structure has crucial
implications for the performance of the organization (Miller, 1987b). In the
next section we briefly review some of the theoretical developments that
relate strategy and structure.

8.2 A Literature Review

8.2.1 Structure Follows Strategy
Chandler (1962) demonstrated that "structure follows strategy" in his
historical study ofAmerican business. "Structure follows strategy" remains
242 Chapter 8. Strategy

a dominant proposition. Quite simply, it says that the internal structure of

the corporation must fit with the adopted corporate strategy. Once a
strategy has been chosen, then the appropriate structure follows. Chandler
did not argue that the proposition is normative. Nonetheless, the corpora-
tions he studied were quite successful, which lends support for using the
proposition as a recommendation as well as descriptive. Chandler's
proposition has been challenged, supported, and refined; yet it and its
derivative forms remain a dominant theme.
Chandler investigated about 100 of the largest American firms. He
found that when companies engaged in new strategies, a new organiza-
tional structure was required. His basic thesis was that when an organiza-
tion stayed within a single dominant business, then a functional structure
would fit. If the firm diversified into a number of different businesses, then
a divisional structure would be required. The argument is basically that
when a firm diversifies, coordination and control within a functionally
organized organization will increase up to the point where it is no longer
manageable. A divisional structure is required to internalize some of the
coordination and control, an adjustment to obtain a fit between informa-
tion-processing demand and capacity.
Concurrently, Ansotrs (1965) interest in corporate strategy focused more
on the choice of an appropriate strategy. He offered new tools of analysis
and rationalized the diversification strategy, which was a widely adapted
strategy in the 1960s, 1970s and early 1980s. He noted the importance of
structure and linked it to synergy-a major element in the diversification
strategy.
Another classical study is the one by Miles and Snow (1978). They
developed a typology of prospectors, analyzers, and defenders and argued
that certain organizational structures would best fit each ofthe categories.
Prospectors require a flexible organization, while defenders need a more
mechanistic organizational structure.
Using the industrial organization framework, Porter's (1980) analyzed
the environment in terms of: buyers, suppliers, substitutes, potential
entrants and rivalry among existing firms. (See the capsule in Chapter 6
on the environment.) He then developed three generic strategies: differenti-
ation, cost leadership and focus or niche strategy in terms of strategic
target and strategic advantage. For a particular market segment target, a
focus or niche strategy is called for. A low cost leadership strategy follows
for a wide market strategy and a low cost position. And a differentiation
strategy follows for a wide market and a uniqueness of product or service
as perceived by the customer. Bon Gout had been a niche player with the
Samsonite label. Albani is a niche player with its Girafbeer and Jolly Cola
in the shadow of Carlsberg and Coke. Emerson and Panasonic are low-cost
leaders in electrical and' electronic products, capitalizing on economics of
8.2 A Literature Review 243

scale. Bang & Olufsen and Mercedes-Benz differentiate their products

through advanced and reliable technology for which they command a
premium in the market place. The appropriate structure for each strategy
is the next step.
A recent major study of the relationship between strategy and structure
is the one by Miller (1987b, 1988). He tries to incorporate ideas from Porter
(1980) into the views by Miles and Snow (1978). Miller (1987b) creates a
typology with the following strategic categories: innovation, market
differentiation, breadth-innovation, breadth-stability, and cost control. He
investigates the relationship between strategy and structure in a study of
Canadian firms. His results are very similar to the results posited by Miles
and Snow (1978). For a recent comprehensive review, see Galunic and
Eisenhardt (1994).

8.2.2 The Counter Proposition: Strategy Follows

Structure
"Structure follows strategy" suggests that the firm chooses a strategy and
then chooses a structure that matches the strategy. However, one could
hypothesize that the structure constrains what the firm will do and how it
will do it. That is, the structure helps determine, or severely constrains, the
choice of strategy: strategy follows structure. The proposition "strategy
follows structure" was argued by Hall and Saias (1980) and is consistent
with more recent research of Frederickson (1984). Burton and Kuhn (1979)
took one of Chandler's companies, General Motors, and presented an
analysis using systems concepts showing that the General Motors
structure, as developed by Sloan, did indeed constrain what General Motors
could do and would do from the late 1920s onward. The theoretical model
is fully presented in Kuhn (1986).
Frederickson (1986) began with the structure variables, complexity,
formalization and centralization and developed hypotheses that each will
affect strategy. His results shown in Table 8.1, indeed confirm that
structure affects strategy.
Frederickson finds that increasing complexity makes strategic actions
more political and the strategic actions more incremental. Similarly,
increasing formalization leads to incremental strategic actions. However,
increased centralization leads to strategies that are major departures from
existing ones. Structure does affect strategy.
244 Chapter 8. Strategy

Table 8.1. Propositions Regarding the Effects of Structure on the Strategic

Decision Process

COMPLEXITY

AI:, the level of complexity increases, so does the probability that

1. Members initially exposed to the decision stimulus will not recognize it
as being strategic or will ignore it because of parochial preferences;
2. A decision must satisfy a large constraint set, which decreases the
likelihood that decisions will be made to achieve organization-level
goals;
3. Strategic action will be the result of an internal process of political
bargaining, and moves will be incremental; and
4. Biases induced by members' parochial perceptions will be the primary
constraint on the comprehensiveness of the strategic decision process.
In general, the integration of decisions will be low.

FORMALIZATION

AI:, the level of formalization increases, so does the probability that

1. The strategic decision process will be initiated only in response to
problems or crises that appear in variables monitored by the formal
system;
2. Decisions will be made to achieve precise, yet remedial, goals, and
means will displace ends;
3. Strategic action will be the result of standardized organizational
processes, and moves will be incremental; and
4. The level of detail achieved in the standardized organizational processes
will be the primary constraint on the comprehensiveness or the strategic
decision process. The integration of decisions will be intermediate.

CENTRALIZATION

As the level of centralization increases, so does the probability that

1. The strategic decision process will be initiated only by the dominant few,
and it will be the result of proactive, opportunity-seeking behavior;
2. The decision process will be oriented toward achieving "positive" goals
(i.e., intended future domains) that will persist in spite of significant
changes in means;
8.2 A Literature Review 245

3. Strategic action will be the result of intendedly rational choices, and

moves will be major departures from the existing strategy; and
4. Top management's cognitive limitations will be the primary constraint
on the comprehensiveness of the strategic process. The integration of
decisions will be relatively high.

Source: Adapted from Frederickson (1986, p. 284).

8.2.3 Fit: Strategy and Structure

There are arguments for both "structure follows strategy" and the counter
proposition "strategy follows structure." Amburgey and Dacin (1994) argue
that strategy is more important in determining structure than structure is
in determining strategy. Either way, there must be fit (Naman and Slevin,
1993). From a normative view, the "structure follows strategy" framework
seems the more promising. The organization first sets its strategy, and then
it must choose a structure to implement it. The structure may constrain
future action; indeed, it should. The organization wants to implement a
particular strategy-not all possible strategies. The lack of a strategic focus
is not likely to be viable, and thus, a structure that implements the chosen
strategy is our design goal. Putting these together, we suggest that strategy
and structure should be symbiotic, and fit is very important. This is in
alignment with the findings by Kretchen et al (1994) in an analysis of 40
empirical studies.

Strategy and New Organizational Forms

New organizational forms illustrate the interrelated nature of strategy and structure. New
organizational forms can be viewed as new structures for organizations; but they can also be
viewed as new strategies for organizations as well. We want to explore the new forms as
strategic choices here.
New organizational forms include: joint ventures, partnerships, strategic alliances, outsourc-
ing, dedicated suppliers, franchising, virtual organizations, to name a few, and organizational
properties, such as flexibility as discussed in the hypercompetition capsule in Chapter 6. Many
of these terms and concepts are well known and had been used for some time. So what is
new? Recently we had utilized these approaches more frequently and more strategically than
in the past. Now, large organizations have significant and major aspects of the organization
involved in these new forms and further, the success of the organization depends upon the
success of the new forms. Earlier, these forms were peripheral and minor appendages to the
organization whose success or failure was often not critical to viability. For many small
organizations, they are solely a new form in that the organization is a franchisee, a dedicated
supplier, a strategic partner or an element in a virtual organization. The "newness" results more
from its strategic importance than novelty of concept.
To examine the new forms, it is instructive to examine how they differ from the "old" forms
of organization. Our traditional notion of organization evolved out of traditional notions of proper-
ty rights and the privileges that property rights give the owner. The owner of a piece of land, a
246 Chapter 8. Strategy

building, a factory, an office or a business then had the right to command and control what took
place in that property within very broad limits of the law and society's norms. An owner could
plant or not plant crops, occupy the building or not, choose the products to produce and what
quantities, hire employees for the office or not, and generally could determine the activities
within the organization. Complementarily, the owner could not make these decisions for
properties that were owned by others. The owner could set the course for his/her own
organization, but not for other organizations. The managerial prerogative followed the property
rights for the organization. The organizational boundary was well defined and what is
inside/outside of the organization is well demarcated. The property "line" marked what was
outside and what was inside and therefore could be commanded. The Porter (1980) model of
the firm within an industry is an example where the boundary between the firm and the market
are well defined. Of course, there could be disputes about who owned a given property and
what the boundaries were; the courts could resolve these issues.
Normal relations between organizations with well defined boundaries took place in the
marketplace where owners would engage as buyers and sellers. An owner could participate in
many markets: for supplies, including labor and to sell its products and services. These markets
have many forms: today's spot market to buy and sell and through contracts which include
longer term agreements, future agreements and contingencies for which the future outcomes
depend upon events at a later point in time. Over the last few years, contracts have evolved
from relatively simple to very complex instruments
The traditional boundary of the organization also delineates "who" is inside and who is
outside, and thus sets the boundary. Employees are inside and individuals who are not
employees are outside. In Chapter 5 on size, we defined the size of the organization as the
number of employees, which also sets the boundary of the organization as employees as part
of the organization and everyone else as outside the boundary. More generally, assets and
employees are inside the boundary of the organization. This is the accounting convention as
well as the legal notion. March and Simon (1958, p. 89-90) note the legal boundary as
fundamental, but also acknowledge its limitation for managerial purposes. They (p. 90) continue
that consumers can also be considered "in," but argue there must be a limit; they draw the limit
as the employees of the organization. The point is clear: the boundary of the organization is not
given for all time and purpose, but can be set to meet a particular need or purpose.
The distinction between units in a production-distribution process that are "in" the
organization and those that are "ouf of the organization typically follows the legal definition of
the boundaries of a particular firm. We find it fruitful to use a more functional criterion that
includes both the suppliers and the distributors of the manufacturing core of the organization
(or its analogue where the core of the organization is not manufacturing). Thus, in the
automobile industry it is useful to consider the automobile dealers as component parts of the
an automobile manufacturing organization.
New forms of organization all share one common aspect: they break down the traditional
boundaries and the simple notions of property rights and the associated management
prerogatives and responsibilities. In short, the new forms have complex boundaries of
ownership, but more importantly of managerial prerogative and complexity. Older notions of
authority, responsibility, command and control break down and call for new attitudes and
concepts.
Where is the boundary of the organization? Baligh and Burton (1982) argue that the
managerial boundary of the organization and the legal boundary are normally not the same.
The managerial boundary is a choice for management and should be thought of in information
processing terms. The managerial boundary is only limited by the extent to which the
management chooses to manage the information and take action. Many managerial issues fall
within both boundaries, eg, the management of an owned building or factory, But other
information issues fall outside the property rights of the firm, eg, advertising is a direct
intervention by management beyond the legal boundaries of the firm. (Automobile companies
gather information and take action for automobile dealers, whether they own the dealership or
not.) The managerial information boundary for the modern organization goes "outside" in many
ways: gather environmental information, advertising, influence buyers, influence government
and policy, monitor new product and process technologies, monitor and direct supplier
8.2 A Literature Review 247

activities, etc. It is now a small step to call the extension of these activities: outsourcing,
strategic alliances, joint ventures, franchising, and virtual organizations. From the organization's
point of view, the first step is an extension of the activities into the environment, and then the
second step is the intervention into the activities of another organization. This relation may be
unidirectional as in outsourcing, but it is likely to be reciprocal as in a joint venture or a
strategic partnership. In short, both organizations are managing in part the activities of other
organizations. This is a very major departure from the traditional concept of the organization
with fixed boundaries operating in markets to complex boundary relations where each
organization is managing in part the activities of the other organization. The "in parr'
management does require new notions of authority, responsibility, command, control and how
managerial actions will be realized; it is a new form.
Williamson (1975) posed the boundary issue in different terms. Using transactions cost logic,
he posited that the organizational size will be determined by the optimal mix of activities inside
the organization (hierarchy) and activities outside the firm (market). In this approach, the size
of the firm also sets the boundary of the firm, ie, what is inside and what is outside. Powell
(1990) extended the markets and hierarchies analysis to include hybrids or networks:
organizations that are neither purely market or hierarchy, but a complex combination. Here, the
boundary becomes more complex, involving relations that go beyond either market or
hierarchy. A new form is then an extension of the markets and hierarchies; and, hybrids can
be thought of as a generic category for new forms.
The concept of new forms has been explored further by Daft and Lewin (1993). Among a
large number of characteristics, they include "permeable and internal and external boundaries"
(p. ii) and continue to discuss "Organizational collaborations ... organizations are teaming up with
others to create strategic alliances and other forms of interdependencies that permit them to
find a niche in a turbulent world. Boundaries among organizations are blurring as they explore
hundreds of interorganizational connections and joint ventures. Network forms that emphasize
interdependence rather than independence may emerge as a distinguishing characteristic of
the new organizational paradigm" (p. iv). Besides the blurring of boundaries, they emphasize
the "interdependencies" which transcend these boundaries to create connections, or networks.
Yet, the question remains: are new forms new combinations of traditional forms, or are they
entirely new. If new forms are extensions and new combinations of traditional forms, then the
theory about traditional forms should be relevant for new forms, albeit, modified and extended.
If new forms are truly a departure, then traditional notions should be discarded and we should
begin de novo to think about managing new forms. (Daft and Lewin (1993) entertain this
possibility in their call for new research and new research approaches.) No doubt, there are
arguments for both views, but we suggest that new forms can be fruitfully considered as
extensions of traditional forms as we have developed above. In short, the central theoretical
issue is the boundary of the organization. New forms have new and different boundaries. These
boundaries are managerial and not limited to legal boundaries.
As such, new forms can be thought of as old forms with new boundaries. For a joint venture,
we can think of it as a new and different organization from either of its parents and it has its
own distinct boundary for which we can now analyze its environment, strategy, technology, etc.
to determine an appropriate organizational design. Outsourcing extends the boundary of the
organization to include elements of its vendor. Strategic alliances are organizations for which
each partner will consider elements of other organization as elements of its own. Virtual
organizations are perhaps the purest of information processing organizations where the
property rights are essentially nil and the organization itself only processes information.
So, new organizational forms are more questions of strategy about where managers choose
to place the boundary of the organization than a question of organizational form itself. If the
organization chooses to engage in joint ventures, partners, outsourcing, etc for strategic
reasons, then management has shifted the boundary of the organization or created a new
organization and a new organizational design is required; and, the approach developed in this
book is relevant.
248 Chapter 8. Strategy

8.3 Definition of Strategy

Strategy and strategic planning are important issues for managers and
researchers. There have been numerous attempts to classify the companies
according to their strategic behavior. Our approach is to categorize the
strategy so that a given type of strategy fits with organizational structure.
In the research, this relationship has received less attention than has the
choice of strategy.
To describe strategic choice, Miles and Snow (1978, p. 29) developed a
four-category typology: defenders, prospectors, analyzers, and reactors. In
their typology, the organization is analyzed as an integral and dynamic
whole, taking into account the interrelationship among the strategy, the
process, and the structure. Miles and Snow's categories are characterized
as follows:

• Defenders: organizations that have narrow product-market domains. Top

managers in this type of organization are highly expert in their
organization's limited area of operation but do not tend to search outside
of their domains for new opportunities. As a result of this narrow focus,
these organizations seldom need to make major adjustments in their
technology, structure, or methods of operation. Instead, they devote
primary attention to improving the efficiency of their existing opera-
tions.
• Prospectors: organizations that almost continually search for market
opportunities and regularly experiment with potential responses to
emerging environmental trends. Thus, these organizations often are the
creators of change and uncertainty to which their competitors must
respond. However, because of their strong concern for product and
market innovation, these organizations usually are not completely
efficient but they are effective.
• Analyzers: Organizations that operate in two types of product-market
domains-one relatively stable, the other changing. In their stable areas,
these organizations operate routinely and efficiently through use of
formalized structures and processes. In their more turbulent areas, top
managers watch their competitors closely for new ideas, and then they
rapidly adopt those that appear to be the most promising.
• Reactors: Organizations in which top managers frequently perceive
change and uncertainty occurring in their organizational environments
but are unable to respond effectively. Because this type of organization
lacks a consistent strategy-structure relationship, it seldom makes
adjustment of any sort until forced to do so by environmental pressures.
8.8 Definition of Strategy 249

Miles and Snow (1978) viewed their categories as being points on a scale
going from defenders to prospectors with the analyzers in between. They
asserted (p. 30) "we believe that our formulation specifies relationships
among strategy, structure, and process to the point where entire organiza-
tions can be portrayed as integrated wholes in dynamic interaction with
their environments." Indeed, the empirical evidence supports their claim.
To validate the strategy typology, researchers have performed empirical
research where some form of cluster analysis has been used. The clusters
are generally few in number and match the Miles and Snow typology
remarkably well (Smith, Guthrie, and Chen, 1989; Roth and Miller, 1990).
It may be useful, however, to develop the typology further. Nicholson
Rees, and Brooks-Rooney, (1990) present a typology with the following five
categories:

• Defenders: Organizations whose strategy is to produce efficiently a

limited set of products directed at a narrow segment of the total poten-
tial market.
• Prospectors: Organizations whose strategy is to find and exploit new
products and market opportunities.
• Analyzers: Organizations whose strategy is to move into new products
or new markets only after their viability has been shown, yet they
maintain an emphasis on their ongoing products. Analyzers have limited
innovation and the innovation is related to the production process and
generally not to the product.
• Hybrids: Organizations that combine the strategy ofthe defender and
the prospector. They move into production of a new product or enter a
new market only after viability has been shown. But unlike analyzers,
they do have innovations that run concurrently with their regular
production. They have a dual technology core.
• Reactors: A residual strategy that describes organizations that follow
inconsistent and unstable patterns.

Nicholson, Rees, and Brooks-Rooney, (1990) tested their categories on only

a very limited set of cases, but their categorization fits well with the
categories developed by Miller (1987a, 1987b, 1986). Miller combined the
categories developed by Porter (1980) and Miles and Snow (1978). His
results are summarized in Table 8.2.
Miller's and Nicholson, Rees, and Brooks-Rooney's categories can be
compared as follows:
250 Chapter 8. Strategy

Miller Nicholson, Rees, and Brooks-Rooney

Complex product innovation Prospectors
Breadth-innovation Hybrids
Breadth-stability Analyzers
Cost control Defenders

Table 8.2. Miller's Integrative Framework

Strategic Challenge Predicted Structural

Dimension Characteristics

Innovation To understand and manage Scanning of markets to discern

more products, customer customer requirements, low for-
types, technologies, and malization, decentralization, ex-
markets tensive use of coordinative com-
mittees and task forces

Market To understand and cater Moderate to high complexity,

differentia- to consumer preferences extensive scanning and analy-
tion sis of customer's reactions and
competitor strategies, moder-
ate to high formalization,
moderate decentralization

Breadth To select the right range of

products, services, custom-
ers, and territory
High complexity, low formaliza-
- innovation, tion, decentralization

High complexity, high formali-

-stability zation, high centralization

Cost control To produce standardized High formalization, high centra-

products efficiently lization

Source: Adapted from Miller (1987a, pp. 55-76).

Smith, Gutherie, and Chen (1989) did a cluster analysis investigating the
strategic behavior in forty-seven electronic firms. Their purpose was to test
the validity of Miles and Snow's (1978) typology. They identified four
clusters and three ofthe clusters could easily be identified with prospectors,
analyzers, and reactors. However, for the fourth cluster, it looked like the
defender but with more product diversity. Small "defenders" outperformed
Definition of Strategy 251

large "defenders" in their study. Comparing the empirical results by Smith,

Gutherie, and Chen (1989) with the typologies developed by Nicholson,
Rees, and Brooks-Rooney (1990), one might argue that there are two types
of defenders-those with a narrow market diversity, low cost, and little
environmental monitoring and those that have a more broad product
diversity, which comes from copying successful markets and products from
other companies within the same industry. These later companies have no
innovation.
Segev and Gray (1990) developed eleven metavariables to describe what
they called strategic content. They used nine different typologies to describe
the organization's strategy and strategic process. For each typology they
assigned values for each metavariable that would best fit a category in a
given typology. For the Miles and Snow (1978) typology, they had the scores
shown in Table 8.3. In Table 8.3 eleven metavariables describe the strategic
category of the organization. Scores range from 0 to 100 with 50 the mean.

Table 8.3. Strategy Content Scores

Metavariable: Content D A P R
5. Technologicalpro~ess 64 64 77 20
6. ProducVDlarketbreadth 10 70 78 32
7. Product innovation 11 61 99 17
8. Quality 81 72 75 20
9. Price level 5 57 94 55
10. Active Dlarketing 36 74 92 18
II. Control systeDllevel 100 58 24 18
12. Resources level 55 70 64 28
13. InvestDlentinproduction 86 60 32 43
14. NUDlber of technologies 1 46 96 37
15. ProfessionaHzation 22 62 95 20

Note: D =defender, A =analyzer, P =prospector, and R =reactor.

Source: Adapted from Segev and Gray (1990, pp. 255-256).

Hambrick. (1983) analyzed functional attributes similar to those shown in

Table 8.3. In general, his results support the scores in Table 8.3, except for
252 Chapter 8. Strategy

the price-level metavariable. Hambrick did not find a significant difference

in the price level between defenders and prospectors.

Table 8.4. The modified Miles and Snow Strategy Categories

Prospector: an organization that almost continually searches for market

opportunities and regularly experiments with potential responses to
emerging environmental trends. Thus, the organization often is the
creator of change and uncertainty to which its competitors must
respond. However, because of its strong concern for product and market
innovation, it usually is not completely efficient.

Analyzer with innovation: an organization that combines the strategy of

the defender and the prospector. It moves into the production of a new
product or enters a new market after viability has been shown. But
contrary to an analyzer without innovation, it does have innovations
that run concurrently with the regular production. It has a dual
technology core.

Analyzer without innovation: an organization whose goal is to move into

new products or new markets only after their viability has been shown
yet maintains an emphasis on its ongoing products. It has limited
innovation related to the production process and generally not the
product.

Defender: an organization that has a narrow product market domain.

Top managers in this type of organization are highly expert in their
organization's limited area of operation but do not tend to search outside
their domains for new opportunities. As a result of this narrow focus,
these organizations seldom need to make major adjustments in their
technology, structure, or methods of operation. Instead, they devote
primary attention to improving the efficiency of their existing opera-
tions.

Reactor: an organization in which top management frequently perceives

change and uncertainty occurring in their organizational environments
but are unable to respond effectively. Because this type of organization
lacks a consistent strategy or structure relationship, it seldom makes
adjustment of any sort until forced to do so by environmental pressures.

We have used a five-category typology to describe the strategic behavior of

the organization. Prospector, defender, and reactor which follow the usual
definitions by Miles and Snow (1978). The analyzer strategy is divided into
Definition of Strategy 253

two categories: (1) analyzers with innovation, which is defined as the hybrid
category by Nicholson, Rees, and Brooks-Rooney (1990) and (2) analyzers
without innovation, which is defined as the analyzer category by Nicholson,
Rees, and Brooks-Rooney (1990). The analyzer without innovation category
can be seen as an extension of the defender strategy. The category also fits
the breadth-stability category, which is one of Miller's (1986) typology
categories. Before we move on, we want to compare this framework with
other well known schemes.
Chandler (1962) implicitly used a product based strategy typology. He
argued that a functional organization perfonned well for a firm with few
products; a divisional organization perfonned better for a firm with a large
number of products. The strategy scheme is then the number of products:
a small number and a large number. He demonstrated clearly through the
historical studies that the firm got into management trouble ofcoordination
and control when the number ofproducts grew large and the firm continued
with a functional organization. The Miles and Snow typology does not
consider the number of products as a primary dimension of strategy, but
focuses more on the innovation and the attitude toward the market.
Ansoff(1965) focused on synergy and the implications for diversification
in strategy. Diversification is related to the number ofproducts-the greater
the diversification, the greater the number of products. The Miles and Snow
typology is not explicit on the level of diversification. We might argue that
the hybrid strategy of blending is a form of diversification. But this is a
different notion than Ansofl's diversification of products and markets.
Porter's (1980) strategies of differentiation, cost leadership and niche
are more closely related to the Miles and Snow typology. The differentiation
strategy focuses on the product or service characteristics. Innovation is one
approach to obtain and maintain products which are different. The Miles
and Snow categories of prospector and analyzer with innovation also focus
on uniqueness of product, which permits the firm to obtain higher profit
margins, which is a return for higher risk. Cost leadership is a defender
strategy, where low cost permits a profitable low price for the product and
fends off imitators and competition. The niche strategy focuses on a small
segment of the market with unique customer needs. The Miles and Snow
prospector may be a niche player, but from a different point of view. The
prospector is more driven by technology and what is possible than a direct
orientation to the customer needs. Yet in both, we can see a need for
innovation and uniqueness.
In Table 8.4, the first four categories do follow an ordinal scale.
Beginning with the prospector and moving to a defender, the prospector is
an environment-oriented, risk-taking organization, frequently an inefficient
organization with a zealous spirit for the "new"; the defender is an internal-
oriented, risk-avoiding, cost-conscious organization that evades the "new."
254 Chapter 8. Strategy

The two analyzers fall in between. The analyzer with innovation is more
like the prospector in its external orientation, but it accepts the "new" only
after careful consideration. The analyzer without innovation only copies
what is known to work; it is not as dogged in its protection of its turf as the
defender. Examples abound.
3M Corporation is a prospector. It exists to develop and market new
products. It thrives on the new, using technology and developing commer-
cial application.
Traditionally, IBM has been an analyzer with innovation. IBM
simultaneously defended its position in the marketplace, carefully
developed new products as demanded, but was infrequently first to the
market. IBM has a large R&D effort as a base for its innovation.
Matsushita can be categorized as an analyzer without innovation.
Matsushita is a Japanese-owned and -managed global manufacturer and
distributor ofconsumer electronics. Panasonic and National are well-known
brands. Matsushita enters markets that are established but not exhausted,
with high-quality low-cost products. It will undertake process development
to reduce costs.
Coca-Cola, in recent years, can be called a defender. It defends its
position as the top worldwide purveyor of cola drinks. Frequently, it is
pitted against Pepsi. The strategy is nonetheless the same; it uses
advertising and promotion to attract customers and maintain market
share.
These five categories are used in the next section when we develop the
propositions that set the relationships between strategy and structure.

8.4 The Two-Stage Model: Description,

Strategy, Structure
We recommend an organizational structure that adopts "structure follows
strategy." Describing and categorizing the strategy is the first stage of a
two-stage process: (1) describing the strategy in detail and the categorizing
of the strategy into one of the modified Miles and Snow (1978) five
categories, and (2) recommending a structure for the strategy. The first
stage is descriptive; we want to state the strategy. The second stage is
normative; we recommend an organizational structure for the given
strategy.
The two-stage model is given in Figure 8.1. Stage 1 describes the
strategy in more basic terms and categorizes that strategy. The second
stage takes the strategy and then recommends an organizational structure.
8.5 The Propositions (or Strategy Description 255

Description Strategy Structure

Product innovation Prospector Configuration
Process innovation Analyzer with innovation Complexity
Product and market Analyzer without innovation Centralization
breadth Defender Formalization
Concern for quality Reactor
Price level
Control level
Technology
Capital requirement

Figure 8.1. Two-Stage Model: Description, Strategy, Structure

Miles and Snow (1978) argued for a fit among environment, strategy, and
organization. They found that a defender would fit best with a stable and
simple environment and an organization that was rather mechanistic while
a prospector at the other extreme would fit a dynamic and complex
environment and would require a more organic organization. Miller (1987 a,
1989) in his empirical investigations basically came to the same results. We
refer to his results in more detail in the next sections. Doty et al (1993)
support the Miles and Snow typology relating it to organizational effective-
ness.
The typology seems to be rather robust and a good way to categorize
strategic behavior in many industries such as: banking (James and Hatten,
1994), transportation (Murphy and Daley, 1996), health (Byles and Labig,
1996), service (Rajaratnam and Chonko, 1995), and biotechnology
(Weisenfeld-Schenk, 1994) among others.

8.5 The Propositions for Strategy

Description
The modified Miles and Snow typology provides the basic strategy concepts
and framework for structural choices. First, the strategy is categorized from
a description of the organization's innovation, capital requirement, price
level, product and market breadth, technology, concern for quality, and
control level. Second, strategy helps prescribe the appropriate structure for
the organization.
Each of the strategy descriptors must be assigned a value, or category.
For example, production innovation is high, or it is low. An organization
which develops and introduces new products on a regular basis would be
high, where an organization that seldom has a new product would be low.
256 Chapter 8. Strategy

Process innovation is perhaps more difficult to observe than product

innovation, but the same concept appears. An organization which changes
technical and administrative procedures has high process innovation. An
organization which seldom changes has low process innovation. In the car
industry with the introduction of flexible manufacturing systems, process
innovation is high. Product/market breadth is defined according to number
of products and number of markets.
Coca-Cola has few products-it sells carbonated soft drinks of various
kinds in bottles and cans. 3M has a broad product/market; a local company
with few products does not have a broad product/market.
The concern for quality is an attitude about quality. Does the organiza-
tional value focus on delivering high quality products and services. This
concern is not only a management issue, but one that must permeate the
entire organization if it is to be realized. 3M, IBM, Matsushita and Coca-
Cola-all have a high concern for quality. Today, General Motors and SAS
have a much higher concern for quality than they did in the 1970s. In
general, concern for quality has gone up in the last decade.
Price level is relative; price level is high if it is higher than the average
price in the industry, a low price level is lower than the average. 3M, IBM,
and Coca-Cola have consistently commanded high prices for their products
in competitive markets. But Matsushita is a low price competitor: good
quality at a low price.
Control level is a measure of preferred level of control. A focus on short-
time microlevel feedback is a high level of control. A focus on long-term
trends is a low level of control. The technology dimension that is used to
define strategy is the routineness dimension defined in Chapter 7. And
finally, high capital requirement is a high capital-to-Iabor ratio. A steel mill
has high capital requirement. A consulting firm does not. 3M also has low
capital requirement, whereas Matsushita scores high on the dimension. The
capital requirement is also related to a "barrier-to-entry" dimension.
There are alternative ways to determine an organization's strategy and
categorize it into prospector, analyzer, defender or reactor. Perhaps the
most straightforward approach is self-typing; simply ask managers, "what
is your strategy type?" Gr~nhaug and Falkenberg (1989) indicate that
managers frequently have a bias in judging their own strategy. The
manager may report his own strategy as analyzer, where a competitor sees
it as a defender. Second, an outsider-competitor or expert-can also be used
to indicate an organization's strategy. And finally, strategy descriptors as
discussed above can be used.
In most empirical studies self-categorization or expert categorization has
been used. But as managers have difficulties consistently categorizing their
own company strategy, we have introduced a choice. The strategy can be
8.5 The Propositions for Strategy Description 257

categorized using the strategy descriptors or the strategy can be categorized

by an ad hoc procedure.
We now turn to describing the strategy using the strategy descriptors.

8.5.1 Describing a Prospector

Table 8.5. Describing a Prospector, Propositions 8.1-8.7

B.l. Ifproduct innovation is high, then the strategy is likely to be prospector.

B.2. If the organization has many products, then the strategy is likely to be
prospector.

B.3. If capital requirement is high, then the strategy is likely to be prospec-

tor.

B.4 If technology is nonroutine, then the strategy is likely to be prospector.

B.5. Ifprice level is high, then the strategy is likely to be prospector.

B.6. If concern for quality is high, then the strategy is likely to be prospector.

B.7. If the preferred level of control is low, then the strategy is likely to be
prospector.

We begin with a description of a prospector. A prospector is likely to have

a number of characteristics. At least one of these characteristics must be
present, and the greater the number of operative characteristics, the more
likely the strategy will be a prospector. Proposition 8.1 in Table 8.5 is, in
part, definitional; clearly, a prospector is likely to have high product
innovation. Miller (1987a) found significant correlation, and the Segev and
Gray (1990) score also indicates that a prospector has a high production
innovation. The hypothesis is generally supported by the literature. 3M is
a product innovator and a prospector. It cannot sustain its new product
strategy without product innovation. However, an organization can have
high product innovation and not be a prospector. Xerox is an example of a
product innovator that was not a prospector. Xerox developed much of the
PC basic technology. Yet it failed to exploit the unknown market. Xerox
focused more narrowly on the known but changing copier market and
missed the personal computer revolution. Product innovation is an activity
258 Chapter 8. Strategy

for prospectors and is a primary characteristic. The high requirement for

innovation requires either a prospector or an analyzer with innovation
strategy.
Proposition 8.2 indicates that an organization with many products is a
prospector. A prospector is constantly seeking new product opportunities
to serve existing and potentially new customers and usually has an array
of products at anyone time.
Proposition 8.3 suggests that a prospector strategy that is aggressive in
product development or market opportunities exploitation requires high
capital and also skilled individuals. A nonroutine technology is also
consistent with a prospector, as given in Proposition 8.4. A routine
technology will not generate the required new products. A nonroutine
technology is likely to be costly. If the organization engages in product
development where margins are likely to be high, nonroutine technology is
very reasonable.
Proposition 8.5 states a high price is a prospector characteristic. Segev
and Gray (1990) put a high score on the price level; there is a high
association. A prospector is not likely to survive with a low price as it is
unlikely to generate sufficient revenue; other strategies may also require
or allow high prices. Hambrick (1983) did not find that price level is a
significant descriptor of strategy.
Proposition 8.6 associates a concern for high quality with a prospector.
This fits the results by Hambrick (1983). However, Segev and Gray (1990)
did not have significant scores on the quality dimension. This relation is not
as strong as the price relation above. It is possible to innovate with new
products that are not necessarily of the higher quality nor produced in a
high-quality process.
The argument to support the low-control prospector relation in
Proposition 8.7 follows. A prospector strategy requires that risks are taken
and new ideas can be generated freely. This can be obtained only if the
management prefers a relatively low level of control. This relation is
supported by Segev and Gray (1990).
Thomas, Litschert, and Ramaswamy (1991) found that firms with a
greater alignment between strategy and the profiles of top managers
generally realized superior performance outcomes. Their definition of
manager profiles is not the one we have used, but they are related. As we
do, they emphasize that the level of risk aversion is one critical dimension
of the manger profile. Of these relations Proposition 8.1 that related
product innovation with a prospector is the most important one. It is nearly
definitional where the other propositions are more indicators of a prospec-
tor.
8.5 The Propositions for Strategy Description 259

8.5.2 Prospector Effects on Structure

We now recommend some organizational design properties that are
required for the prospector to be effective (Table 8.6). We begin in Proposi-
tion 8.8: the prospector should have low formalization. Prospector strategy
requires a low formalization so that the organization can react quickly to
the new situation and not be constrained by rigid rules. Low formalization
is required because of the need for innovations (Frederickson, 1986; Miles
and Snow, 1978; Miller, 1987b). Complex information has to be processed,
and task uncertainty is high, which from the information-processing
argument leads to low formalization.
Complexity should be either low or high when the organization has a
prospector strategy. It should either have generalists or specialists to deal
with the high requirements for innovation (Frederickson, 1986; Miles and
Snow, 1978; Miller, 1987b). Again, demand and capacity of information
processing have to be balanced.
In Proposition 8.10, the recommendation is that for an organization that
has a prospector strategy, a low centralization is required so that the
organization can react quickly. Low centralization is also required because
of the need for innovation (Frederickson, 1986; Miles and Snow, 1978;
Miller, 1987a).
Finally, Propositions 8.11 and 8.12 indicate that a prospector strategy
should be configured as a simple, ad hoc, or matrix organization but not a
bureaucracy-machine or professional. Miles and Snow (1978), Miller
(1987a, Table 24.2) and Daft (1992, Exhibit 14.4) all support these
propositions. The basic idea behind Propositions 8.11 and 8.12 is that if the
organization has a prospector strategy, then a configuration that will
enable it to react fast and engage is innovation as necessary. The descrip-
tion of a prospector and its recommended effect for the organizational
structure are summarized in Table 8.7.

Table 8.6. Prospector Effect on Structure, Propositions 8.8-8.12

B.B. If the strategy is prospector, then formalization should be low.

B.9. If the strategy is prospector, then complexity should be low or high.

8.10. If the strategy is prospector, then centralization should be low.

B.ll. If the strategy is prospector, then configuration should be either a

simple, an ad hoc, or a matrix configuration.
260 Chapter 8. Strategy

8.12. If the strategy is prospector, then configuration should be neither a

professional bureaucracy nor a machine bureaucracy.

Table 8.7. Prospector Strategy: Description and Effect

Description Strategy Structure

High product innovation Prospector Low formalization
Many products High or low complexity
High capital requirement Low centralization
Nonroutine technology Simple, matrix or ad hoc
High price level configuration
High quality concern Not a bureaucracy
Low control level

8.5.3 Describing an Analyzer Without

Innovation
An analyzer without innovation is likely to have a number of characteristics
that describe it. At least one of these characteristics must be present, and
the greater the number of operative characteristics, the more likely the
strategy will be an analyzer without innovation. Proposition 8.13 in Table
8.8 is almost definitional. Product innovation is an important element in
innovation. The firm can copy effectively proven ideas and not incur the
cost of product innovation. It does require selectivity in the choice of what
to copy.
Proposition 8.14 indicates that an analyzer without innovation will
likely have many products. An analyzer searches the environment for new
opportunities to maintain a reasonable level of product diversity and
appropriate match with the market.
Proposition 8.15 suggests that a low or medium capital requirement is
consistent with an analyzer without innovation. An organization with a
medium capital investment is likely to have some fixed capabilities but can
adjust also. The analyzer without innovation that seeks new opportunities
but also maintains its profitable position is appropriate.
Proposition 8.16 states that a routine technology is characteristic of an
analyzer without innovation. It is important to handle the old products or
services in a routine way to maintain quality. New products will be
included in the routine operations quickly.
Process innovation, however, is quite different, as given in Proposition
8.17. An organization with medium or high process innovation devises new
8.5 The Propositions for Strategy Description 261

ways to make a product or deliver a service. The focus is the process and
making it more efficient. Usually, the product or service is known. It is not
likely that an organization can accomplish simultaneously the process
development and matching new products. Rarely, a process innovation will
change the product. A more likely approach is that the organization sees a
product with market potential; it copies the product and attempts to
become efficient and competitive through process innovation.

Table 8.8. Describing an Analyzer Without Innovation, Propositions

8.13-8.21

8.13. Ifproduct innovation is low, then the strategy is likely to be analyzer

without innovation.

8.14. If the organization has many products, then the strategy is likely to be
analyzer without innovation.

8.15. If capital requirement is medium or low, then the strategy is likely to

be analyzer without innovation.

8.16. If technology is routine, then the strategy is likely to be analyzer

without innovation.

8.17. If process innovation is medium or high, then the strategy is likely to

be analyzer without innovation.

8.18. If technology is mass production, then the strategy is likely to be

analyzer without innovation.

8.19. Ifprice level is low, then the strategy is likely to be analyzer without
innovation.

8.20. If concern for quality is high, then the strategy is likely to be analyzer
without innovation.

8.21. If preferred level of control is high, then the strategy is likely to be

analyzer without innovation.

For an analyzer without innovation, cost efficiency is important. A routine

technology that fits the mass production is appropriate and may be
necessary. This is stated in Proposition 8.18.
262 Chapter 8. Strategy

Proposition 8.19 considers price level. When prices are low, the
organization must be cost effective to survive. For a copy strategy, the firm
can maintain low costs and be profitable with low prices.
Propositions 8.20 and 8.21 relate to quality and control, respectively.
This strategy focuses more on cost and quality, and it is, thus, appropriate
to require a high level of control.

8.5.4 Analyzer Without Innovation Effects on

Structure
An analyzer without innovation looks very much like a defender (Table 8.9).
An analyzer does not invent new things but copies when it is safe. The
Danish charter-tour-operator Spies is a very good example of an analyzer
without innovation. The company never develops a new destination. It
waits, and when some destinations seemed to be ready for mass-marketing,
it moves in a very formalized way. It markets low prices and high quality.
Compared to a prospector, the demand for information processing is lower
and less complex.

Table 8.9. Analyzer Without Innovation Effect on Structure, Propositions

8.22-8.26

8.22. If the strategy is analyzer without innovation, then formalization

should be medium or high.

8.23. If the strategy is analyzer without innovation, then centralization

should be medium or high.

8.24. If the strategy is analyzer without innovation, then complexity should

be high.

8.25. If the strategy is analyzer without innovation then the configuration

should be functional or divisional.

8.26. If the strategy is analyzer without innovation, and the organization is

small, then the configuration should be simple.

Proposition 8.22 recommends that an analyzer without innovation should

have a medium or high formalization. This organization has a need for
limited change and in a reasonable time. Internal efficiency is a goal, and
8.5 The Propositions for Strategy Description 263

formalization is one means to achieve the economics of regularity in

process.
Proposition 8.23 recommends a medium or high centralization. The top
management can process the required information and can make decisions
that are coordinated and consistent in the organization's goals. High
specialization is required to be effective as an analyzer without innovation.
Additionally, the organization should operate both current and new
products and services. This generally leads to a high organizational
complexity, Proposition 8.24.

Table 8.10. Analyzer Without Innovation: Description and Effect

Description Strategy Structure

Low product innovation Analyzer Medium to high formali-

Medium to high process without zation
innovation innovation High complexity
Many products Medium to high centraliza-
Medium to low capital re- tion
quirement Functional or divisional
High quality concern configuration: may be sim-
Routine technology pIe configuration if small
High control level
Low price level

Propositions 8.25 recommends a functional configuration for an analyzer

without innovation. The functional configuration captures the efficiencies
of specialization when the need for adjustment is slow and reasoned. For
an analyzer without innovation, a functional organization is normally best.
If the technological diversity is low due to no economies of scale, then a
divisional configuration may be appropriate. However, if the organization
is small, a simple structure may do, as is suggested in Proposition 8.26.
These propositions are summarized in Table 8.10.

8.5.5 Describing Analyzer with Innovation

We begin with a description of the characteristics of an analyzer with
innovation (Table 8.11). At least one of these characteristics must be
present, and the greater the number of operative characteristics, the more
likely the strategy will be an analyzer with innovation. Proposition 8.27 is
definitional, stating that a high product innovation firm is likely to be an
analyzer with innovation, but to be efficient it also has to have some
process innovation. With many products, the firm can maintain its diversity
264 Chapter 8. Strategy

by developing new products, which is also consistent with an analyzer with

innovation, as suggested in Proposition 8.28.

Table 8.11. Describing an Analyzer with Innovation, Propositions 8.27-8.33

8.27. Ifproduct innovation is high and process innovation medium, then the
strategy is likely to be analyzer with innovation.

8.28. If the organization has many products, then the strategy is likely to be
analyzer with innovation.

8.29. If the capital requirement is medium, then the strategy is likely to be

analyzer with innovation.

8.30. If technology is medium routine, then the strategy is likely to be

analyzer with innovation.

8.31. Ifprice level is medium, then the strategy is likely to be analyzer with
innovation.

8.32. If concern for quality is high, then the strategy is likely to be analyzer
with innovation.

8.33. If the preferred level of control is medium, then the strategy is likely
to be analyzer with innovation.

An organization with a medium capital investment is likely to have some

capabilities rather fixed but can adjust also. This is consistent with an
analyzer with innovation, as suggested in Proposition 8.29.
For a medium routine technology, the organization has some flexibility.
It is consistent with an analyzer with innovation strategy as in Proposition
8.30.
The price level in Proposition 8.31 is medium for an analyzer with
innovation. Moderate prices create some flexibility but also require
attention to efficiency. An analyzer with innovation strategy, where the
organization both attends to existing profitable markets and seeks selective
new opportunities, seems an appropriate description.
Proposition 8.32 indicates that a concern for high quality is also
consistent with an analyzer with innovation. With a medium preference for
control the management wants some influence. This can be obtained via
8.5 The Propositions for Strategy Description 265

control over current operations. Product innovation should then be less

controlled. This is Proposition 8.33.

8.5.6 Analyzer with Innovation Effects on

Structure
An analyzer with innovation strategy requires that formalization should be
medium. There should be high levels of standardization in current activities
and high levels of flexibility in new undertakings. This is the essence of
Proposition 8.34.
In an analyzer with innovation strategy, the organization is set up to
deal with both new and current activities, which leads toward high
complexity. An analyzer with an innovation strategy is actually a dual one.
The current operations are different from new undertakings. Proposition
8.35 captures this idea.
When the organization has an analyzer with innovation strategy, the
centralization should be medium, as given in Proposition 8.36. There should
be tight control over current activities and looser control over new ventures.

Table 8.12. Analyzer with Innovation Effect on Structure, Propositions

8.34-8.38

8.34. If the strategy is analyzer with innovation, then formalization should

be medium.

8.35. If the strategy is analyzer with innovation, then complexity should be

high.

8.36. If the strategy is analyzer with innovation, then centralization should

be medium.

8.37. If the strategy is analyzer with innovation, then the configuration

should be a matrix or divisional configuration.

8.38. If the strategy is analyzer with innovation and the organization is

small, then the configuration should be simple.

The analyzer with innovation requires a focus on change and efficiency.

Propositions 8.37 and 8.38 recommend a matrix or divisional configuration,
except for a small organization where a simple configuration is appropriate.
266 Chapter 8. Strategy

These Propositions are developed from Miles and Snow (1978), Miller (1989,
Table 24.2) and Daft (1992, Exhibit 14.4).
The propositions for an analyzer with innovation are summarized in
Table 8.13.

Table 8.13. Analyzer with Innovation: Description and Effects

Description Strategy Structure

High product innovation Analyzer Medium formalization

Many products with High complexity
Medium routine technology innovation Medium centralization
Medium capital requirement Matrix, divisional or sim-
High quality concern pIe configuration, if small
Medium control level
Medium price level

8.5.7 Describing a Defender

The defender is identified by a number of characteristics (Table 8.14). At
least one of these characteristics must be present, and the greater the
number of operative characteristics, the more likely the strategy is to be a
defender.

Table 8.14. Describing a Defender Strategy, Propositions 8.39-8.46

8.39. Ifproduct innovation is low, then the strategy is likely to be defender.

8.40. Ifprocess innovation is high, then the strategy is likely to be defender.

8.41. If the organization has few products, then the strategy is likely to be
defender.

8.42. If capital requirement is high, then the strategy is likely to be defender.

8.43. If technology is routine, then the strategy is likely to be defender.

8.44. If concern for quality is high, then the strategy is likely to be defender.

8.45. If price level is low or high, then the strategy is likely to be a defender.
8.5 The Propositions for Strategy Description 267

8.46. If preferred level of control is high, then the strategy is likely to be a

defender.

We begin with low product innovation in Proposition 8.39. The low-

innovation requirement fits the defender strategy where the organization
can utilize its particular expertise.
As argued above, Proposition 8.40 states that high process innovation
is a focus that tends to drive out other potential competitors. This fits well
with an organization that is protecting its markets and products: the
defender. The realized efficiency and cost reduction can yield a competitive
advantage.
Proposition 8.41 suggests that a defender will have few products. It
needs to defend these products well in the marketplace and with an
appropriate technology. Viability depends on being successful with these
limited activities. For a company with a high capital investment, the ability
to adjust its capital base quickly is not likely. Thus, it needs to protect and
defend its position; a defender strategy and technology protection are
appropriate, as stated in Proposition 8.42.
A defender is characterized by a routine technology in Proposition 8.43.
Consequently, new products for new customers are less likely to be possible.
It needs to defend its position for the technology it has or copy well-known
products or markets.
A defender uses high quality to keep its markets and keep competitors
out. The defender will have a high concern for quality, as suggested in
Proposition 8.44.
In Proposition 8.45, a low price is a sign of a defender. To defend its
position the organization needs generally to be competitive both on quality
and prices. The efficient operation makes it possible to defend on prices,
and this is normally the case when the organization produces a standard
product. If the organization defends a high-quality brand, like LEGO, then
a high price can be charged.
Finally, in Proposition 8.46 a high level of control signals an element in
a defender strategy.

8.5.8 Defender Effects on Structure

The effects of a defender strategy on the organizational structure are
recommendations that follow the general proposition that "structure follows
strategy" (see Table 8.15). Generally, the demand for information-process-
ing capacity is low and the task uncertainty is low. In Proposition 8.47, we
state that a defender strategy generally requires high formalization. The
268 Chapter 8. Strategy

reason is that a defender needs cost efficiency, which can be obtained

through formalization (Miles and Snow, 1978; Miller, 1987a).
Proposition 8.48 holds that a defender strategy generally requires
medium to high complexity. The reason is again that a defender needs cost
efficiency, which can be obtained through specialization (Miles and Snow,
1978). Empirical results have shown that defenders may also have low
complexity for small companies (Miller, 1987a).

Table 8.15. Defender Effect on Structure,- Propositions 8.47-8.50

8.47. If the strategy is defender, then formalization should be high.

8.48. If the strategy is defender, then complexity should be medium or high.

8.49. If the strategy is defender, then centralization should be high.

8.50. If the strategy is defender, then the configuration should be func-

tional.

Proposition 8.49 states that a defender usually requires a high centraliza-

tion. Again, the reason is that a defender needs cost efficiency, and that can
be obtained through centralized coordination (Miles and Snow, 1978; Miller,
1987a).
Finally, in Proposition 8.50, a defender strategy usually requires a
functional configuration. The emphasis on specialization and efficiency is
consistent with a defender's goals and its competitive position.

Table 8.16. Defender: Description and Effects

Description Strategy Structure

Low product innovation Defender High formalization

High process innovation Medium complexity
Few products High centralization
Normally high capital requirement Functional configuration
Routine technology
High quality concern
High control level
High or low price level
8.5 The Propositions (or Strategy Description 269

The defender two stage model is summarized in Table 8.16. The first stage
gives the description of a defender, and the second stage states the
structural recommendations.

8.5.9 Reactor Strategy

In the previous section we described four strategic types. For each of these
types we have assumed that some consistency between descriptive factors.
Some descriptors may lead to more than one strategy categorization. For
example, a high price level is consistent with both the prospector and the
defender strategies. However, a consistent pattern must exist for the
strategy to be viable. If that is not the case, the strategy is termed a reactor
strategy. This is the case when half the descriptor variables would lead to
a defender strategy while the rest would lead to a prospector strategy. In
such a case consistency is absent.
A reactor is also normally characterized by a low concern for quality. In
fact, it may be the only significant factor separating a reactor from the
other categories. However, concern for quality does not mean the same for
a prospector and a defender. High quality for a prospector means that the
organization can react quickly to trends and changing needs. It is flexible
with an effective product innovation. Cost is normally not a major concern.

Table 8.17. Reactor Proposition, Proposition 8.51

8.51. If the descriptor factors in Figure 8.1 are inconsistent and quality
concern is low, then the strategy is reactor.

For a defender, quality concern normally means that the product has to
meet standard specification, cost has to be low, and machinery has to be
effective and efficient. For Med Electronics, the change from nonroutine
unit production to a routine mass production would change the quality
focus dramatically. Therefore, the relationship among TQM, technology,
and strategy is important. Med Electronics now has high product innova-
tion, high price level, low standardization, and low process innovation. It
has relatively few products in a narrow market segment. The quality
concern is high. Med Electronics has many of the characteristics of a
prospector. With its new strategy ofmoving into the consumer market, Med
Electronics will have to increase process innovation, control level, and
standardization. Additionally, the price has to be lowered for consumer
products. The concern for quality should still be high but of a different
270 Chapter 8. Strategy

nature. The strategy now looks more like an analyzer with innovation with
a mixture of prospector and defender activities. Proposition 8.51 summa-
rizes the above discussion (see Table 8.17).

8.6 International Dimensions

The international dimensions of a strategy do not affect the choice of basic
structure per se. Of course, if an international strategy means that the
organization increases its market or product diversity, that affects
structural choice.
To further strengthen the hypotheses on the choice of configuration, we
add four propositions related to international issues (Table 8.18). They are
based on whether the organization has a high or low foreign product or
service diversity and how intensive foreign involvement is. The propositions
are based on Phatak (1992) and Habib and Victor (1991).
Propositions 8.52 through 8.55 follow the general argument that product
and market diversity may lead to divisionalization. Again, an information-
processing perspective has been used. The more diversity, the more demand
for information processing capacity. The higher the involvement in interna-
tional markets or production the more important it is to reduce the
information-processing demand by creating units that can internalize the
coordination requirements. When that is not possible, a matrix configura-
tion is needed. In fact, the choice of an information system with its channels
and capacity is a strategic choice.

Table 8.18. International Configurations, Propositions 8.52-8.55

8.52. If foreign product or service diversity is low and foreign involvement

is low, then configuration should be either an international functional
configuration or an international division configuration.

8.53. If foreign product or service diversity is medium and international

involvement is medium, then configuration should be a global
configuration.

8.54. If foreign product or service diversity is low and international

involvement is high, then configuration should be a global functional
configuration.
8. 7 Choosing the Right Strategy 271

B.55. If foreign product or service diversity is high and international

involvement is high, then the configuration should be a multidi-
mensional, global configuration.

The contingency theory of international configuration has not been

researched to the same degree as our basic configurations. The strengths
of Proposition 8.52 through 8.55 are therefore offered with caution (Habib
and Victor, 1991).

8.7 Choosing the Right Strategy

Bon Gout has to figure out what to do to solve the crisis: it has to select a
strategy. The strategy of Bon Gout could be described by accessing the
values in the descriptive part of Figure 8.1: finding the value for innovation,
product and market breadth, quality, control, technology, and capital
requirement.
From an organizational point of view the" structure follows strategy"
hypothesis, as discussed previously, would be to choose values for the
descriptor variables that could result in a strategy type that fits the
organizational structure.
The organization has its choice of the amount of product and process
innovation to employ. Additionally, it can choose the number of products
and markets. The choice could be the organizational evolution that was
previously discussed going from exploiting the domestic market to diversi-
fying into other markets and, finally, to diversifying into more products.
The choice of technology-including quality, control, and standardiza-
tion-is a part of the strategic choice. Investments in long-term commit-
ments like Rand D equipment and market development are also a part of
choosing the right strategy.
Generally, managing size, environment, and technology to fit the
environment is a part of selecting a strategy. Choosing objectives, goals,
and ways to obtain those goals re_quires choosing and implementing a
strategy. The strategy may be chosen to fit the environment and technolo-
gy, and then a structure is selected, or as discussed above, the strategy is
chosen with the limits given by the organizational structure.
The literature on strategic choice is vast (Porter, 1980; Quinn, Mintz-
berg, and James, 1988), and a discussion of it lies outside the scope of this
book.
272 Chapter 8. Strategy

8.8 The Strategy of SAS

To review briefly, during the 1950s and 1960s the airline market developed
very favorably. It was a stable, rapidly growing, and very profitable market.
The challenge was to deliver a given number of ton miles on every route
(Buraas, 1972, p. 113). During this period, SAS had established a very
bureaucratic and rigid organization. Problems were realized with the first
oil crisis in the beginning of the 1970s. Suddenly, the market was stagnant
because of the general depression that followed the oil crisis. It became of
utmost importance for organizations to be efficient in order to make a
profit. The SAS management recognized the necessity for cost reductions
and therefore set up some internal groups to find ways of cutting costs
without firing employees.
Management initiated cost cutting where the most savings could be
realized. Customer services that were very important for SAS to maintain
its market share were cut, whereas less important costs were left un-
touched. A survey among customers showed that the cuts were noticed and
that customers generally felt that customer service at SAS had deterio-
rated. It was not until 1980 that a major change was initiated.
On March 2,1981, a strategy outline was presented and approved by the
board of directors. The outline consisted ofa three-year plan for the
organization in which major changes were described.
The main strategy was to be the best alternative for the full-fare
passenger who wanted "value for money." The idea was to give special
service to business travelers and thereby create an image of being "the
businessman's airline."
The first action was to introduce a special category for businesspeople
on the planes. In Europe, it was called Euroclass. After 1988, all business-
class passengers traveled Euroclass irrespective of their destination (SAS,
1988, p. 12). Euroclass passengers got roomier seating and better on-board
service. In some larger airports special Euroclass check-in counters were
introduced for quicker check-in. Special lounges were available for the
exclusive use of Euroclass passengers. In this way SAS tried to create an
environment specially designed for business travelers.
It was important for employees to be able to give service in their daily
contact with customers ("Moment of truth"). As a first sign of a changed
attitude toward employees, Carlzon began to inform employees about the
changes as soon as the board of directors approved his strategy outline. All
20,000 employees received a red booklet called "Now We Start Fighting."
This booklet presented the main concepts of the new strategy. Through this
action Carlzon showed that employees should take active part in the
changes necessary for the future. This marked the beginning ofteam spirit
among SAS employees, which was an important part of implementing the
B.B The Strategy ofSAS 273

strategy. Therefore the new strategy focused strongly on the role of front-
line personnel in the customer's perception of service.
As part of the increased service for business travelers, SAS had to
become a more precise airline company. Many departures had earlier been
delayed to await the arrival of other planes so that transit passengers could
catch their planes. From now on SAS would compensate transit passengers
who did not catch their SAS departure by offering alternative ways of
reaching their destination. In that way service for transit passengers
remained almost unchanged, and planes could depart on time. Furthermore
a serious drive was launched aiming at an improvement of the flight
schedules and the punctuality. In that connection, Carlzon even had a
monitor placed at his office so that he could control take offs and landings
of all planes.
Besides becoming "the businessman's airline" a complementary strategy
was devised. The main idea was to utilize available capacity to increase
SAS's total revenue within the frame of the main strategy (Obel, 1986). A
full airplane generated more revenue. The good years of the 1960s resulted
in an overcapacity of planes. Furthermore, the new jet engine airplanes
were much larger than earlier planes. The result was too many seats
compared to the number of passengers, so airline companies reduced their
prices to fill up their planes. In the 1970s, stagnation really set in, but price
agreements and regulation of the airline industry meant that companies
were not hit hard. The oil crises, however, hit the entire world economy
very hard. At the same time, the U.S. government began liberalizing
American airline policy, which meant increased competition on the already
tough market for trans-Atlantic air travel. The oil crisis and liberalization
in the United States made the overcapacity even more pronounced.
As a consequence a substrategy of capacity utilization was developed to
offer the remaining seats on each flight to leisure travelers. These tickets
were cheaper, but passengers received less service. On board the plane, this
distinction between the customers was particularly important. In order to
further distinguish SAS's products from those of other airlines, Carlzon
introduced "the total traveling concept." Surveys had shown that business-
men demanded the service ofnot having to worry about transport and hotel
bookings. This was why in 1985 SAS established destination service, whose
core product was a ticket, transport and hotel package. Passengers were
now able to book airline tickets, ground transportation and confirm hotel
reservations with one call. To pursue this vision, SAS now had to develop
a hotel network. Because it would take SAS at least 10 years to build up an
appropriate amount of hotels, Carlzon decided to buy a 40% share in the
Intercontinental Hotels, giving SAS immediate access to 106 hotels
worldwide. This investment later turned out to be a very unprofitable
274 Chapter 8. Strategy

stake, and when SAS finally gave it up in 1992, many judged the invest-
ment to be the biggest mistake Carlzon had ever made at SAS.
Because the size of SAS's domestic market limited the competitiveness
of the company, SAS would need to co-operate with another airline
company in order to gain access to a secondary domestic market. This was
why in the eighties several alliances were made, with THAI and the
Japanese ANA, and shares were bought in Continental Airlines, LanChile
and British Midland. In 1989 SAS formed the European Quality Alliance
with Swissair, Austrian Airlines and, in the beginning, also Finnair.
A similar, but larger alliance was the goal of the Alcazar negotiations.
Alcazar was projected to include SAS, KLM, Swissair and Austrian
Airlines. Because Carlzon was preoccupied with the Alcazar negotiations,
Jan Reinas was appointed as new CEO of SAS whereas Carlzon was
announced coming leader of the Alcazar group. Reinas' job was to create a
"rescue" plan in case the Alcazar negotiations failed. This plan included a
harsh rationalization program, aiming at severe cost reductions, extensive
firings and a renewed focus on SAS's core business: The airline.
At the end of 1993 the Alcazar negotiations broke down, and Carlzon
was left with no position to fill out at SAS. Jan Reinas, whose appointment
had from the beginning only been temporary, was succeeded by Jan
Stenberg on April 1 1994.
Jan Stenberg followed up with the focus now being on the airline,
several of SAS's non profitable side business such as Diner's club Nordic
and SAS Leisure were sold off. Furthermore extensive firings were made,
and the life long employment concept was abandoned along with the idea
of the total traveling concept.
Stenberg agreed with Carlzon in that the company needed to co-operate
with other airlines to avoid becoming a feeder company. By co-operating
with other airlines, SAS would not only save administrative costs, but the
company would also gain access to more airport hubs. But Stenberg did not
want these co-operations to take form of actual mergings, but rather as
strategic alliances.
As part of its strategy to create a Baltic junction , in 1995 SAS became
co-owner of the Lettish company Air Baltic Corporation. In 1996 Lufthansa
and SAS signed a strategic alliance, the core of which was a code share of
all route networks between Scandinavia and Germany. The SAS-Lufthansa
created a strong network between Denmark and Germany, and because
Lufthansa has traditionally been strong in Africa and the Middle East, the
alliance gave SAS the ability to exploit these areas as well. Further
alliances were made with Thai Airways International and United Airlines,
giving SAS access to the Far East and the American market. Recently, the
so-called Star Alliance was created counting Lufthansa, United Airlines,
Air Canada, Thai Airways and Varig (May 1997).
B.B The Strategy ofSAS 275

Prior to Carlzon's changes, SAS's strategy was technology based because

a growing market and high prices pennitted a continuing investment in the
latest equipment. During this period, SAS could be described as a defender,
and the organization's high fonnalization, medium to high complexity, high
centralization and the functional configuration matched the organizational
implications of a defender strategy (propositions 8.47, 8.48, 8.49 and 8.50)
excellently.
The new strategy was to create SAS as an analyzer with innovation.
There was a definite emphasis on innovation. Propositions 8.27, 8.28, 8.29,
8.30,8.31,8.32 and 8.33 give the definitional characteristics of an analyzer
with innovation. With the exception of proposition 8.30 which indicates
medium routine technology, whereas the aircraft requires technology that
is defined in detail, SAS clearly moved from a defender to an analyzer with
innovation. The new strategy was clear and consistent, and it called for
organizational change. In propositions 8.34, 8.35, 8.36 and 8.37 we find
these implications:

• Medium formalization
• High complexity
• Medium centralization
• Matrix or divisional configuration

Carlzon changed the organization's direction. Formalization remained high,

but many rules were abolished, and new ones gave front-line personnel
much greater freedom of action. Complexity remained high. Centralization
was decreased, and decision making was pushed down in the organization.
The company's previous functional configuration was changed a number of
times in the 1980s-beginning with the divisional organization and
eventually incorporating more complex matrix relations to coordinate the
nondivisible technology.
However, one of the things that caused SAS's problems was that over
time, the company's strategy became similar to that ofthe prospector. Once
Carlzon had brought SAS ''back on its feet" he started focusing on new
market opportunities which were conceptualized in his "total traveling
concept." Many of his visions made SAS operate in areas that were very
new to the company and most of which were based on emerging environ-
mental trends. Carlzon became more of a prospector, but did not change the
organization to match this strategy. The prospector strategy fits with low
formalization (proposition 8.8) low or high complexity (proposition 8.9) and
low centralization (proposition 8.10). This structure was also not compatible
with SAS's nondivisible technology.
In brief, the SAS organization was not geared to co-ordinate the many
new areas and there was a misfit between the organization's strategy, and
276 Chapter 8. Strategy

its structure. Stenberg's new strategy meant that SAS was to focus on its
core business: the airline. The company sold off many of its sidebusinesses
and subsidiaries, but it still seeks selective new opportunities through its
many strategic alliances. SAS' strategy today can therefore be described as
an analyzer with innovation. Following propositions 8.34, 8.35, 8.36 and
8.37, SAS's current structure fits well with these recommendations. SAS
has medium formalization, high complexity, medium centralization and
although its overall structure is functional it has a number of established
divisions as well as cross organizational matrix relations.

8.9 Summary
"Structure follows strategy" is a basic proposition. An organization's
strategy is an important determinant of its organizational design.
A modified Miles and Snow strategy categorization is both compre-
hensive and discriminating. They are prospector, analyzer without
innovation, analyzer with innovation, defender, and reactor. Each strategy
calls for a different organizational design.
The prospector requires an open, flexible organization that adjusts
quickly. The analyzer adjusts more slowly and deliberately. The defender
needs a stable organization. In general, information processing require-
ments decrease moving from prospector to defender; however, the defender
needs particular information to fight off competitors. Prospectors need a lot
of information to deal with an unknown environment and technology;
everything can be in flux. Analyzers fall in between. We do not offer
organizational recommendations for a reactor strategy, but rather suggest
a reactor strategy is not a viable strategy in the long run. No organization
can salvage it.
Strategy and structure should fit, and that fit is realized with a
structure that processes the right amount of the right kind of information.

References
Ansoff, H. Igor. 1965. Corporate Strategy. New York: McGraw-Hill.
Amburgey, Terry L., and Tina Dacin. 1994. "AF, the Left Foot Follows the Right? The
Dynamics of Strategic and Structural Change." Academy of Management
Journal, 37(6), 1427-1452.
Baligh, Helmy H. and Richard M. Burton. 1982. "The Movable Boundaries Between
Organizations and Markets." International Journal of Policy Analysis and
Information Systems, 6(4), 435-449.
Buraas, Anders. 1972. Fly over Fly. OIso, Norway: Gyldendal.
References 277

Burton, Richard M., and Arthur J. Kuhn. 1979. "Strategy Follows Structure! The
Missing Link of Their Intertwined Relation." Fuqua School of Business.
Research Paper.
Byles, Charles M., Labig, Chalmer E. Jr. 1996, "Hospital Strategy and its
relationship to administrative practice and performance: A partial test of the
Miles and Snow Typology." Journal of Managerial Issues, 8(3), 326-342.
Carlzon, Jan. 1985. Riv pyramideme ned! Copenhagen, Denmark: Gyldendal.
Chandler, Alfred D. 1962. Strategy and Structure: Chapters in the History of the
Industrial Enterprises. Cambridge, MA: MIT Press.
Daft, Richard L. 1992. Organization Theory and Design (4th ed.). St. Paul, MN:
West (2nd ed., 1986, 3rd ed., 1989).
Daft, Richard L. and Arie Y. Lewin. 1993. "Where Are the Theories for the 'New'
Organizational Form? An Editorial Essay." Organization Science, 4(2), i-vi.
Doty, D. Harold, W.H. Glick, and G.P. Huber. 1993. "Fit, equifinality, and
organizational effectiveness: A test of two configurational theories." Academy of
Management Journal. 36(6), 1196-1250 (55 pages).
Frederickson, J.W. 1984. "The Effect of Structure on the Strategic Decision Process."
Academy of Management Proceedings, 12-16.
Frederickson, J.W. 1986. "The Strategic Decision Process and Organization
Structure." Academy of Management Review, 11(2),280-297.
Galunic, D. Charles, and Kathleen M. Eisenhardt. 1994. "Reviewing the Strategy-
Structure-Performance Paradigm." In Barry Staw and Larry Cunnings (Eds.),
Research on Organizational Research, 16, 215-255.
Grf/lnhaug, Kjell, and Joyce S. Falkenberg. 1989. "Exploring Strategy Perceptions
In Changing Environments." Journal of Management Studies, 26(4), 249-359.
Habib, Mohammed M., and B. Victor. 1991. "Strategy, Structure, and Performance
of US Manufacturing and Service MNCS: A Comparative Analysis." Strategic
Management Journal, 12, 589-606.
Hall, David, and Maurice Saias. 1980. "Strategy follows Structure!" Strategic
Management Journal, 1, 149-163.
Hambrick, Donald C. 1983. "Some Tests of the Effectiveness and Functional
Attributes of Miles and Show's Strategic Types." Academy of Management
Journal, 25(1), 5-26.
http://www.sas.se/sas/aboutsaS/inbrief.html
James, William L., Hatten, Kenneth J. 1994. "Evaluating the Performance effects
of Miles' and Snow's Strategic Archetypes in Banking, 1983 to 1987: Big or
Small. Journal of Business Research. 31(1,2),145-154.
Kuhn, Arthur J. 1986. Organizational Cybernetics and Business Policy: System
Design for Performance Control. University Park: Pennsylvania State University
Press.
"Lufthansa, SAS to start integrating networks." 1996. Aviation Week & Space
Technology, 2211.
March, James G. and Herbert A. Simon. 1958. Organizations. New York: John
Wiley & Sons, Inc. 262p.
Miles, Raymond E., and Charles C. Snow. 1978. Organizational Strategy, Structure
and Processes. New York: McGraw-Hill.
Miller, Danny. 1987a. "The Structural and Environmental Correlates of Business
Strategy." Strategic Management Journal, 8, 55-76.
278 Chapter 8. Strategy

Miller, Danny. 1987b. "Strategy Making and Structure: Analysis and Implications
for Performance." Academy of Management Journal, 30(1), 7-32.
Miller, Danny. 1988. "Relating Porter's Business Strategies to Environment and
Structure Analysis and Performance Implications." Academy of Management
Journal, 31(2), 280-308.
Miller, Danny. 1989. "Complexurations of Strategy and Structure: Towards a
Synthesis." In D. Asch and C. Bowman, Readings in Strategic Management. New
York: MacMillan.
Murphy, Paul R. and James M. Daley. 1996. "A Preliminary analysis of the
Strategies ofIntemational Freight Forwarders." Transportation Journal, 35(4).
Naman, John L. and Dennis P. Slevin. 1993. "Entrepreneurship and the concept of
fit: A model and empirical Tests." Strategic Management Journal, 14(2),
137-153.
Nicholson, Nigel, Anne Rees, and Anette Brooks-Rooney. 1990. "Strategy,
Innovation and Performance." Journal of Management Studies, 27(5), 511-534.
Obel, Borge. 1986. "SAS: Changes in Competition, Strategy and Organization." In
Richard M. Burton and Borge Obel (Eds.). Innovation and Entrepreneurship in
Organizations: Strategies for Competitiveness, Deregulation, and Privatization.
Amsterdam: Elsevier.
Phatak, Arvid V. 1992. International Dimensions of Management. Boston, MA:
PWS-Kent.
Porter, Michael E. 1980. Competitive Strategy: Techniques for Analyzing Industries
and Competitors. New York: The Free Press.
Powell, Walter W. 1990. "Neither Mrket Nr Hierarchy: Network Forms ofOrganiza-
tion." Research in Organization Behavior, 12, 295-336.
Quinn, James Brian, Henry Mintzberg, and Robert M. James. 1988. The Strategy
Process: Concepts, Contexts, and Cases. Englewood Cliffs, NJ: Prentice Hall.
Rajaratnam, Daniel, Chonko, Lawrence B. 1995. "The Effect of Business Strategy
type on Marketing Organization Design, Product-market growth Strategy,
Relative Marketing Effort, and Organizational Performance." Journal of
Marketing Theory & Practice, 3(3), 60-75.
Roth A.v., and J.G. Miller. 1990. "Manufacturing Strategy, Manufacturing
Strength, Managerial Success, and Economic Outcomes." In J.E. Ettlie, M.C.
Burstein, and A. Fiegenbaum (Eds.), Manufacturing Strategy. Norwell, MA:
Kluwer.
SAS. 1980-1991. SAS Annual Reports. Stockholm: SAS
SAS. 1994. SASAnnual Reports. Stockholm: SAS.
SAS. 1995. Annual Report. Stockholm: SAS.
SAS. 1996. Miljr)regnskab. Stockholm: SAS
SAS. 1996. Annual Report. Stockholm: SAS
SAS. 1996. De fr)rste 50 ar. Stockholm: SAS
SAS. 1996. ScandinatianAirlines 1946-1996. Stockholm: SAS
SAS. 1997. Deldrsrapport 1: 1997. Stockholm: SAS
Segev, Eli, and Paul Gray. 1990. Business Success: Strategic Unit Comprehensive
Computer-Based Expert Support System. Englewood Cliffs, NJ: Prentice-Hall.
Smith, Ken G., James P. Gutherie, and Ming-Jen Chen. 1989. "Strategy, Size and
Performance," Organization Studies, 10(1),63-81.
References 279
Thomas, Anisya, Robert J. Litschert, and Kannan Ramaswamy. 1991. "The
Performance Impact ofStrategy-Manager Coalignment: An Empirical Examina-
tion." Strategic Management Journal, 12,509-522.
Weisenfeld-Schenk, Ursula. 1994. "Technology Strategies and The Miles and Snow
Typology: A Study of the Biotechnology Industry." R&D Management, 24(1),
57-64.
Williamson, Oliver E. 1975. Markets and Hierarchies: Analysis and Antitrust
Implications. New York: The Free Press.
CHAPTER 9

Organizational Design: A Synthesis

9.1 Introduction
"Design is concerned with how things ought to be, with devising artifacts
to attain goals" (Simon, 1981, p. 133). A number of concepts are important
here. "Ought to be" is a normative; it is a statement of the ideal and what
we would like to see. It does not necessarily exist anywhere in reality, but
it could be and it is desired; it is a creation of the mind and imagination,
and perhaps the heart. As we noted in Chapter 1, positive science is a
description of the world as we find it and some statement of what it could
be without any statement of desirability, or "ought to be." An "artifact" is
of man's making and not natural. The Golden Gate bridge is an artifact and
we find no such bridges in nature. Artifacts were part ofthe beginning: an
ax or a folded leaf to drink water from. We made the ax and built the
Golden Gate bridge. We devised them, ie, created the idea in the form of a
blueprint and then took steel and concrete and the efforts of men, women
and machines to realize the bridge. Our "goal" for this bridge is to convey
individuals and motor vehicles across the Golden Gate. Design of a bridge
is readily accepted. In this book, we create a statement of what we mean by
design for an organization.
Organizational Design is concerned with what the organization ought to
be, ie, what should be the configuration and properties of the organization.
(Refer to Figure 1.3.) An organization is an artifact, indeed an old artifact
(China, 5000BC and the Middle East, 2000BC or earlier), but an artifact,
nonetheless. Organizations are devised, created and changed under the
direction of management. We imagine what an organization should be and
then we build the organization for the purpose to do something: make
automobiles, transfer monies around the world, or provide health care for
individuals. Organizations are artifacts in another significant way: it is an
artifact of abstraction which is unlike a bridge. A bridge is real in that it
contains steel and we can walk on it. An organization remains an abstrac-
tion; it exists as a restriction on behavior and guide on behavior, but it also
facilitates purposeful information processing as we have developed in this
book. The organization guides what individuals can do, but it is not
individual behavior per se. Individuals behave and take action; organiza-
tions facilitate behavior, but they do not behave in the same way. Of course,
282 Chapter 9. Organizational Design: A Synthesis

we impute behavior to the organization; it is an analytical convenience that

we adopt in this book. An organization as an information processor does
behave and we can think about the organization as reading, transmitting,
and processing of information as behavior.
In this chapter we further develop the neo information processing model
for organizational diagnosis and design. Using the empirical base as
supporting and augmenting the information processing view, we propose
means and rules to obtain particular organizational configurations and
properties which fit. From the design point of view we examine fit:
contingency misfits, design fit and total fit as part of a dynamic and
continuing process of design.

9.2 Diagnosis and Design

Growth at Applied Computer Science, INC, created a conflict for the design
of the organization. Management's preference for a high level of micro-
involvement suggested a simple and centralized organizational design. But
the company's increased size suggested that an organization with speciali-
zation and decentralization would be more appropriate. Generally, to decide
on a suitable organizational structure Applied Computer Science has to
assess the effect of all contingency factors.
Diagnosis is the description of the organization and how it works, and
further the assessment of whether the organization meets the needs, i.e.,
is it efficient, effective and viable. Design is devising structures to attain
goals. Referring back to the basic model in Figure 1.2, the goals are
effectiveness, efficiency and viability ofthe design. The "devised structures"
are the properties (complexity, formalization, centralization, and so on) and
the structural configuration of the organization. The "ought to be" or
normative aspects of design are the recommendations themselves-that is,
a recommended design is one that will be effective, efficient and viability.
As we have argued throughout this book, choosing a good design recommen-
dation is a difficult task. There are a number of issues. First, most of the
knowledge base was originally developed in the positive science tradition
to explain the world, not directly to recommend good practice and aid
managers; some translation is necessary. Second, the knowledge is diverse
and fragmented as there are numerous models using different variables
and which are only loosely related. Third, design must be a synthesis or
putting together of knowledge to obtain a recommendation.
To illustrate the difficulty we presented Duncan's (1979) and Perrow's
(1967) models in Chapter 1. Stated in normative terms, Duncan's six design
recommendations rely solely on the environment; Perrow's four design
9.2 Diagnosis and Design 283

recommendations rely solely on the technology. Both are part of our

knowledge base and what we know about good design. Yet, it is not clear
how to put the two separate pieces of knowledge together; how do we create
a synthesis of what we know for design purposes. As we argued earlier both
the environment (Chapter 6) and the technology (Chapter 7) are important
determinants of a good design. So the question is now: how does one put the
two together and what is the proper weighing and balance of the environ-
ment and the technology in determining a good design. Further, the design
recommendations are stated in different terms. Duncan's design terms are
functional, decentralized, mixed functional, and mixed decentralized, where
Perrow uses craft, nonroutine, routine and engineering. If we could apply
each set of recommendations separately, then the synthesis would be
relatively simple and straightforward; just apply both recommendations.
Unfortunately, this approach cannot work. For example, let us assume that
the environment suggests a mixed functional design and the technology
suggests an engineering design. What does it mean to recommend a mixed
functional and engineering design? We do not know. Similarly, there are
fifteen (16-1) other designs which would result from a Duncan-Perrow
combination which are very difficult to interpret. We must then develop a
concept of what a design is, so we can devise recommendations in a
common, consistent vocabulary which fit together.
The properties and structural configurations on the right side of Figure
1.2 are the design concepts we utilize. In the previous chapters we have
stated the knowledge within organizational theory in propositions using the
same consistent set ofterms and concepts. Duncan's term functional does
apply, but none of Perrow's terms are used directly. His terms have been
related to technological routineness. We have chosen design concepts that
are broadly used in the literature and for which there are numerous
empirical studies and a vast literature. Additionally, the concepts fit into
the neo-informational processing view of organizational design. Yet, this
literature is not consistent and terms can be defined and operationalized
differently. The syntheses of this knowledge, including Perrow's insight on
technology, must be considered and incorporated into what we know about
good design. For design purposes, concepts and terms must be defined.
A design recommendation is a set of statements about the organization's
properties and structural configuration. For example, the configuration is
functional, centralization is high, and so on. How do we arrive at this
recommendation? Let us examine again the recommended "centralization
is high." Many contingencies could influence this result. The environmental
uncertainty is low, the environmental equivocality is low, the management
has a high preference for microinvolvement, and so on. And, of course, there
can be influences which suggest a high decentralization, such as a large
organization. How does one combine in a reasonable way these numerous
284 Chapter 9. Organizational Design: A Synthesis

influences; some of which are mutually supportive and some opposing?

There are many alternative approaches-for Organizational Consultant we
chose the Mycin rule, explained in Chapter 1. The idea behind the rule is
straightforward; for a given recommendation, more supporting information
strengthens the recommendation, but at a decreasing marginal rate. For
example, a low environmental uncertainty suggests a high centralization
as does management's high preference for microinvolvement. The two
influences are mutually supportive and the two together yield a stronger
recommendation than either one by itself. However, the two supporting
influences do not add linearly, but at a decreasing rate. Contingent
influence can pull in opposite directions. For example, a group climate and
a prospector strategy lead to opposite recommendations about centraliza-
tion, resulting in a relatively weak combined recommendation about
centralization. It is equally important to recognize these situations, which
may suggest that the contingent influences themselves are not compati-
ble-that is, a situational misfit. Management should address the situa-
tional misfits first.
In Chapter 1, we introduced four fit criteria: situation fit, contingency
fit, design parameter fit, and total design fit. Refer to Figure 1.3. Situation-
al misfits occur when the situation in which the organization has to operate
is not internally consistent. As discussed in previous chapters, Bon Gout
experienced a change in environmental uncertainty and environmental
equivocality that might not fit with its strategy and technology. Similarly,
the airline industry became more dynamic and unpredictable, making
SAS's environment more uncertain and not compatible with management's
style. Most situation misfits occur when a balanced situation is changed.
The environment may change, new technologies may be adopted, or new
management or new strategies may create an unbalanced situation. It is
unlikely that a proper organizational design exists in an unbalanced
situation. To obtain a balanced situation strategic choices may be neces-
sary. The organization may temporarily or permanently be designed to
make such strategic decisions.
Contingency fit was the topic of Chapters 3 through 8, and in the next
sections we return to that subject. Contingency misfits arise when design
parameters are not aligned with a particular situation. For example, there
is a contingency misfit when the organizational design is very formalized
and bureaucratic and the environment is uncertain. Environmental,
technological, management, or strategic changes may result in a contin-
gency misfit requiring an organizational design change. If no total design
misfits existed in a situation with a stable environment, then a change to
a uncertain environment creates both a situation misfit and a contingency
misfit. A change in the situation may thus create both a situation misfit
and a contingency misfit. A contingency misfit may also obtain if one
9.2 Diagnosis and Design 285

designs an inappropriate organizational structure-one that violates the

propositions presented in Chapters 3 through 8.
When contingency misfits are observed, a natural reaction is to fix
them-to change the organization so it matches the situation. However, two
problems have to be dealt with. First, with situational misfits present, the
various contingency propositions may lead to conflicting recommendations
for the design parameters, making it difficult or even impossible to obtain
design parameter fit. Second, the organizational design theory that
provides the basis for the contingency fit does not provide a one-to-one
correspondence between situations and design parameters. As is discussed
in more detail in the next sections, there may be other elements of
organizational design that may fit a particular situation-a nonunique
recommendation or equifinality in the terms of the goals. For example, both
a matrix configuration and a functional configuration with liaison devices
could be appropriate for the same situation. Similarly, there may be more
recommendations for other design parameters. However, not all possible
combinations may be appropriate. The situation may recommend both high
and low centralization and high and low formalization. This gives four
possible combinations. But only high centralization with low formalization
and low centralization with high formalization are normally appropriate
designs. High centralization with high formalization is an overkill while low
centralization with low formalization is appropriate only with a proper
incentive structure.
From this discussion it follows that one cannot infer from a design
parameter misfit to either situation or contingency misfit. A situation misfit
may lead both to contingency and design parameter misfits. There may be
no situation misfit and no contingency misfit and still design parameter
misfit. Strategic organizational diagnosis is the analysis of situation,
contingency and design parameter misfits. Strategic organizational design
is creating total design fit.
Fit simply means that things have to fit together. Either side can be
changed. As discussed in Chapters 3 through 8, either situation can be
manipulated or the organizational design can be changed. It also may be
necessary to change both to obtain total design fit. In the next sections we
reexamine the proposition developed in Chapters 3 through 8 from the
point of view of obtaining design parameter fit. Then we discuss the
problem of obtaining total design fit and rules that lead toward design
parameter fit. Similarly, rules that help companies recognize situation
misfits are discussed, providing a basis for the strategic organizational
diagnosis and a statement of design as product.
286 Chapter 9. Organizational Design: A Synthesis

9.3 Contingency Fit

We now examine the design synthesis, the contingencies that support a
given recommendation or design. The idea is to examine when a given
design is appropriate. For example, a simple configuration is appropriate
when the management has a high preference for microinvolvement
(Proposition 3.25). Other contingent propositions lead to the same recom-
mendation, such as a small organization (Proposition 5.15). Neither is
sufficient as both contingencies can be compatible with other configura-
tions. In design, there is always a possible nonuniqueness or equifinality in
the recommendations. Nonetheless, the two supporting contingencies for a
simple configuration strengthen the recommendation and make that
recommendation more likely than alternative recommendations.
The presentation of the many propositions in this book may give the
impression that organizational strategic diagnosis and design is complex
and difficult. It is. We have tried to simplify the presentation by presenting
only rules where the contingency factor normally could take two values:
high and low. We have used two values for the management style, four
values for the climate measure, three values for the size measure, 3x2
values for the environment, 4x2x3x2 values for the technology, and finally
five values for the strategy. In total, this gives 34560 different situations.
Increase the values for environment and technology to three-low, medium,
and high-and the number of situations increases to 524880. Of course, it is
impossible to capture this many situations simply. Additionally, many of
the situational combinations do not fit together.
It is this complexity that makes design difficult. There are thousands of
possible design alternatives from which to choose. The examination of the
designs themselves and when each property and configuration is appropri-
ate is a straightforward way to examine this complexity. We now turn to
the design recommendations themselves. These configurations and proper-
ties were introduced, discussed and illustrated in Chapter 2. In the next
section, in each table we put the rules together that relate to the various
design properties. Then we sort the rules according to the particular value
of a design parameter recommendation.

9.3.1 The Simple Configuration

The simple configuration consists of a top manager and individuals. There
may be little functional specialization and no well-defined departmental
structure with departmental heads. Decision making, coordination, and
control are usually done by the top manager. For the simple configuration
9.8 Contingency Fit 287

the organization relies on the top manager (Table 9.1). An important

contingency for the simple configuration to be viable is that the top
manager is ready to assume that role. Thus, the simple organization
requires a top manager who will assume responsibility and authority to get
things done-Proposition 3.25.

Table 9.1. Recommendations for a Simple Configuration

A simple configuration may be recommended when there is at least one of the

following:

Contingency Factor The Values Proposition

Management High microinvolvement 3.25

Size Small 5.15

Environment Max two of 6.10

Complexity 6.12
Uncertainty 6.14
or Equivocality are high 6.16
High hostility 6.18

Strategy Prospector 8.11

Analyzer and small 8.26
8.38

The strength of the recommended simple configuration is greater as the number

of operative contingencies increases.

Proposition 5.15 suggests a small organization may be a simple configura-

tion. When both of these conditions are met, then the simple configuration
is more likely to be recommended. Taken together they provide strong
support for a simple configuration. If the manager is not ready to take the
responsibility, an ad hoc configuration may be recommended. But that will
depend on other contingency factor values.
Propositions 6.10, 6.12, and 6.16 state that the simple configuration may
be recommended when only one of the environment factors-complexity,
uncertainty, or equivocality-is high. Ifmore than one is high, the informa-
tion processing capacity of the manager may be exceeded. Proposition 6.14
allows both equivocality and uncertainty to be high, but environmental
complexity must be low for a simple configuration to be recommended. A
hostile environment also suggests a simple configuration. A prospector
strategy is compatible with a simple organization. The size of the organiza-
288 Chapter 9. Organizational Design: A Synthesis

tion is important for the choice of the simple configuration in a different

way. If it is small and has an analyzer strategy, it strengthens the
recommendation for a simple configuration.
When all the premises are true, it is very likely that the simple
configuration will be appropriate. For example, when the top manager has
a high preference for microinvolvement and the organization is small, has
an equivocal but simple and less certain environment, and has an analyzer
strategy, then the simple configuration is a very highly recommended
choice. Many different situations may lead to a recommendation that a
simple configuration is appropriate. The more propositions that support
such a recommendation, the stronger the recommendation. The Mycin
certainty factor principle captures this idea; more supportive information
strengthens the recommendation.

9.3.2 The Functional Configuration

The functional configuration has more levels and more horizontal special-
ization than a simple configuration. There is a well-defined departmental
structure. The departments are created based on the functional specializa-
tion in the organization.
Generally, a medium or large organization is compatible with a func-
tional configuration-but with other configurations as welL With medium or
large size the functional specialization is usually desirable (Table 9.2). The
functional organization is limited in its capacity to process large amounts
of information and particularly information which is unusual and non
standard in form. This is particularly true, when the climate is an internal
process climate. High equivocality creates difficulties and excessive in-
formation processing demands. The functional configuration does not deal
well with ambiguity and the unknown. If the equivocality of the environ-
ment is high, then the uncertainty and complexity have to be low for the
functional organization to be appropriate. Propositions 6.10, 6.11, 6.13 and
6.16 relate the environment to the functional form.
A functional organization operates efficiently and will continue within
narrow environmental variations. It can deal with environmental uncer-
tainty and complexity provided these parameters are well defined and
value variation is small. Even so, high complexity and high uncertainty are
likely to require additional information processing capacity in liaison
activities and integrators. The functional configuration deals well with
routine technology that does not have large information processing
demands, Proposition 7.16. Finally, innovation is difficult to obtain with a
functional configuration, as suggested in Propositions 8.25 and 8.50. The
defender and analyzer without innovation are compatible, but innovative
9.3 Contingency Fit 289

strategies should use the functional configuration with caution. Innovation

requires large amounts of information processing to effect the requisite
change. In brief, the functional configuration is oriented to efficiency of
specialization and does not support well large information processing
demands.

Table 9.2. Recommendations for a Functional Configuration

A functional configuration may be recommended when there is at least one ofthe

following:

Contingency Factor The Values Proposition

Climate Internal process 4.38

Size Medium or large 5.16

Environment Low equivocality, 6.10

low uncertainty,
low complexity

Low equivocality, 6.11

low uncertainty,
high complexity

Low equivocality, 6.13

high uncertainty,
high complexity

High equivocality, 6.16

low uncertainty,
low complexity

Technology Not nonroutine 7.16

Strategy Analyzer without innovation 8.25

Defender 8.50

The strength of the recommended functional configuration is greater as the

number of operative contingencies increases.
290 Chapter 9. Organizational Design: A Synthesis

9.3.3 Divisional Configuration

The divisional configuration is characterized by organizational subunits
based on a grouping of products, markets, or customers. The units are
relatively autonomous unlike the units in the functional configuration.

Table 9.3. Recommendations for Divisional Configuration

A divisional configuration may be recommended when there is at least one ofthe

following:

Contingency Factor The Values Proposition

Climate Rational goal 4.53

Size Medium or large 5.16

Environment Low equivocality, 6.11

high complexity,
low uncertainty

Low equivocality, 6.12

low complexity,
high uncertainty

Low equivocality, 6.13

high complexity,
high uncertainty

Technology Divisible 7.17

Strategy Analyzer 8.25

8.37

The strength of the recommended divisional configuration is greater as the

number of operative contingencies increases.

A medium or large organization is compatible with a divisional con-

figuration, Proposition 5.16 (Table 9.3). In this example, there are multiple
design choices for a particular contingency variable value. For low equivo-
cality the uncertainty and complexity can be either high or low, but not at
the same time for a divisional configuration, Propositions 6.11, 6.12, and
6.13. This medium to low information processing situation is compatible
with a divisional organization, provided that the nondivisible technology is
not present, as implied in Proposition 7.17. In brief, the divisional
9.8 Contingency Fit 291

configuration deals with some information processing demands of a

medium to large organization and a somewhat complicated environment by
creating relatively self-contained divisions that employ divisible technolo-
gies, Propositions 8.25 and 8.37. The divisional configuration is compatible
with an analyzer strategy with or without innovation. This also fits a
rational goal climate, Proposition 4.53.
The divisional organization can deal with a number of products and
markets, provided they are relatively independent. A divisional carbon copy
can be repeated many times in different locations.

9.3.4 Matrix Configuration

The matrix configuration introduces a dual hierarchy; it incorporates the
essential functional and divisional configuration of an organization
simultaneously.
There are a large number of matrix configuration propositions to match
its complicated, dual focus and hierarchy (Table 9.4). The high managerial
preference for microinvolvement is not very compatible with a matrix,
Proposition 3.29. To handle the dual hierarchy a sharing or trustful climate
is appropriate, Propositions 4.8 and 4.23. Top management involvement in
details will likely create an information overload at the top and delay
activities as well as potentially create motivational issues for the matrix
managers.
A matrix is compatible with a medium or large organization, but it is
less likely to be appropriate for a small organization, Proposition 5.16. The
environment is a very important contingency for a matrix recommendation,
Propositions 6.15 and 6.17. The matrix is not necessary unless the
environment is not well understood and requires a large information
processing capacity, which a matrix organization can supply. By omission
there is an implication that a matrix configuration may not be needed for
a better defined environment.
A unit technology where there are a number of customers or projects is
well suited for a matrix configuration, Proposition 7.15 and 7.18. However,
that if the technology is divisible, the matrix is not needed-probably a
divisional configuration is more appropriate, as argued above.
Finally for a prospector and an analyzer with innovation, a matrix
configuration is a recommended configuration, Propositions 8.11 and 8.41.
These strategies with their demand for innovation have large information
processing demands along multiple dimensions, and the matrix can be
effective as well as efficient.
292 Chapter 9. Organizational Design: A Synthesis

Table 9.4. Recommendations for a Matrix Configuration

A matrix configuration may be recommended when there is at least one of the

following:

Contingency Factor The Values Proposition

Management style Not high microinvolvement 3.29

Climate Group 4.8

Developmental 4.23

Size Medium or large 5.16

Environment High equivocality, 6.15

high complexity,
high uncertainty

High equivocality, 6.17

high complexity,
low uncertainty

Technology Unit 7.15

Not divisible 7.18

Strategy Prospector 8.11

Analyzer with innovation 8.37

The strength of the recommended matrix configuration is greater as the number

of operative contingencies increases.

The matrix configuration is too costly to utilize in a small or simple

situation. It can handle large information demands along multiple
dimensions where both effectiveness and efficiency are important. The
efficiency can be realized only ifvery scarce resources are utilized without
opportunity losses.

9.3.5 Ad Hoc Configuration

The ad hoc configuration is a very loose configuration that employs Theory
Y people with a high capacity for information processing. The ad hoc
configuration has much in common with the matrix configuration; it is
recommended in similar situations (Table 9.5).
9.3 Contingency Fit 293

Table 9.5. Recommendations for an Ad Hoc Configuration

An ad hoc configuration may be recommended when there is at least one of the

following:
Contingency Factor The Values Proposition

Management style Low microinvolvement 3.28

Climate Group 4.8

Size Small 5.15

Not important 5.17

Environment High equivocality, 6.14

low complexity,
high uncertainty

High equivocality, 6.15

high complexity,
high uncertainty

Technology Nonroutine 7.20

Strategy Prospector 8.11

The strength of the recommended ad hoc configuration is greater as the number

of operative contingencies increases.

The ad hoc configuration is advisable only for very large or complex

information processing situations where the level of the unknown is very
high. Top management cannot be involved in details; there are too many
new decisions to be made too often, Proposition 3.28. People have to be able
to work together in this loose organizational structure, so a group climate
is appropriate, Proposition 4.8. Size does not play an important role, but an
ad hoc configuration functions best when the organization is small,
Propositions 5.15 and 5.17. An ad hoc configuration can cope with the ill-
defined environment of high equivocality and complexity, which require a
large information processing capacity, Propositions 6.14 and 6.15. The ad
hoc configuration is not suited for a low information "routine" technology,
Proposition 7.20. Whatever the source, the ad hoc makes sense only for a
large information processing situation. A prospector strategy fits this
configuration well, Proposition 8.11.
294 Chapter 9. Organizational Design: A Synthesis

9.3.6 The Professional Bureaucracy

A bureaucracy can be defined as an organization with the following
characteristics (Weber, 1946):

• Division oflabor
• Well-defined authority hierarchy
• High formalization
• Impersonal in nature
• Employment based on merits
• Career tracks for employees
• Distinct separation of members' organizational and personal lives.

The above is also often called a machine bureaucracy. If the members in a

"bureaucracy" are highly skilled professionals and if some of the standard-
ization is obtained via the professionalization, giving the professionals some
decision authority, it is called a professional bureaucracy (Table 9.6).

Table 9.6. Recommendations for a Professional Bureaucracy

A professional bureaucracy may be recommended when there is at least one of

the following:
Contingency Factor The Values Proposition

Management style Not a high microinvolvement 3.26

Climate Internal process 4.38

Size Medium or large 5.16

Technology Routine 7.19

Strategy Not a prospector 8.12

The strength of the recommended professional bureaucracy is greater as the

number of operative contingencies increases.

The professional bureaucracy configuration is recommended for medium

and large organizations, Proposition 5.16. If the manager has a high
preference for microinvolvement, then the professional bureaucracy is less
likely, Proposition 3.26. A strong statement is Proposition 7.19, which
indicates that a nonroutine technology is not likely to be compatible with
a bureaucracy. Even the professional bureaucracy relies heavily on routine
9.3 Contingency Fit 295

procedures, albeit at professional levels, and nonroutine procedures require

too much nonstandard information processing for the bureaucracy.
Therefore, a prospector strategy will not be compatible with a professional
bureaucracy, Proposition 8.12. This is very compatible with an internal
process climate, Proposition 4.38.

9.3.7 The Machine Bureaucracy

The machine bureaucracy focuses on the adherence to rules. In a bureau-
cracy, the information processing is very dependent upon the bureaucratic
characteristics above; they define what information is required and how it
is to be processed; the information technology can then be well specified. A
bureaucracy can process large amounts of standardized information.

Table 9.7. Recommendations for a Machine Bureaucracy

A machine bureaucracy may be recommended when there is at least one of the

following:
Contingency Factor The Values Proposition

Management style High microinvolvement 3.27

Climate Internal process 4.38

Size Medium or large 5.16

Technology Routine 7.19

Strategy Not a prospector 8.12

The strength ofthe recommended machine bureaucracy is greater as the number

of operative contingencies increases.

The rationale for a bureaucracy is given in Section 9.2.6. A manager who

wants to be involved in detail is compatible with a machine bureaucracy
(Table 9.7). Medium and large organizations also can support a machine
bureaucracy. As with the professional bureaucracy, the machine bureau-
cracy cannot support a nonroutine technology, Proposition 7.19. At the
heart of the machine bureaucracy are programmed procedures that are not
compatible with the variations demanded with a non routine technology.
This also precludes a prospector strategy, Proposition 8.12.
296 Chapter 9. Organizational Design: A Synthesis

9.3.8 Organizational Complexity

The configuration is a general description of the organizational structure.
Other characteristics are important to give a more complete design
specification, which can be stated in numerous ways.

Table 9.8. Organizational Complexity Recommendations

A HIGH organizational complexity may be recommended when there is at least

one of the following:

Contingency Factor The Values Proposition

Management style Low microinvolvement 3.15

Climate Internal process 4.39

Rational goal 4.54

Size Large and not pubic 5.1

Environment Low equivocality, 6.3

low complexity,
high uncertainty

Low equivocality, 6.4

high complexity,
high uncertainty

Technology Routine 7.9

7.11
Process 7.14

Strategy prospector 8.9

Analyzer 8.24
8.35

The strength of the recommended high organizational complexity is greater as

the number of operative contingencies increases.

Organizational complexity is defined in Chapter 2; it is the degree of

horizontal, vertical, and spatial differentiation. Horizontal differentiation
is greater when there are several small tasks and specialization by
experience, education, and training. Vertical differentiation is the number
of hierarchical levels between top management and the bottom of the
hierarchy. Spatial differentiation is greater when there are many locations
9.3 Contingency Fit 297

of facilities and personnel. As the degree of organizational complexity

increases, then the difficulty of coordination issues and the requirements
for information processing increase as well.

Table 9.8. Continued.

A MEDIUM organizational complexity may be recommended when there is at

least one of the following:
Contingency Factor The Values Proposition

Climate Group 4.9

Developmental 4.24

Size Large 4.2

4.3

Environment Low equivocality, 6.1

low complexity,
low uncertainty

Low equivocality, 6.2

high complexity,
low uncertainty

High equivocality, 6.7

low complexity,
low uncertainty

High equivocality, 6.8

high complexity,
low uncertainty

Technology Routine and 7.10

small organization

The strength of the recommended medium organizational complexity is greater

as the number of operative conting~ncies increases.

Organizational complexity depends on management's preference for

microinvolvement, size, environment, technology, and strategy (Table 9.8).
Proposition 3.14 suggests that a high preference for microinvolvement calls
for a low organizational complexity, and vice versa in Proposition 3.15: low
microinvolvement calls for high complexity. A high management involve-
ment in details is likely to create a management overload for high
complexity; the information processing demands are too large. However, the
298 Chapter 9. Organizational Design: A Synthesis

information overload may be decreased by an information processing

system. Climate is related to the information processing capacity of the
individuals. Thus, an internal process and rational goal climates with their
requirements for a more structured workplace and less emphasis on the
group can allow for a high complexity with specialization. In the group and
developmental climates the complexity may be lower.
Size is related to organizational complexity; increased size calls for
increased organizational complexity-both vertical differentiation and
horizontal differentiation, Propositions 5.1 through 5.4. These propositions
are well supported and widely accepted.
The environmental Propositions 6.1 through 6.8 are quite involved. Two
arguments underlie these propositions. First, greater environmental
equivocality calls for less organizational complexity. With high equivocality,
it is not clear what will need to be done to survive, and thus, general skills
will be more appropriate than specialized ones. Second, greater uncertainty
calls for greater complexity. Here, small adjustments will need to be made
to understand environmental forces. Specialists can react quickly and
effectively in the small. However, if the environment is very hostile, the
organizational complexity should be low to be able to react quickly.
Propositions 7.9 through 7.14 consider the effect of technology on
organizational complexity. A more routine technology supports greater
horizontal differentiation and a greater span of control. A less routine
technology suggests a greater vertical differentiation. The routine
technology permits greater specialization without greater information
processing demands. The nonroutine technology requires greater informa-
tion processing demands that can be supported with greater vertical
differentiation. However, modern information systems mitigate this effect.
Propositions 8.9, 8.24, 8.35, and 8.48 consider the strategy and organiza-
tional complexity. A prospector needs low or high complexity and either
generalists or specialists but not hybrids. The defender needs the high
internal efficiency ofhigh organizational complexity with its specialization,
Proposition 8.48. Finally, the analyzer with or without innovation relies on
high complexity to focus on many areas in an efficient manner.
In summary, a low organizational complexity recommendation has the
greatest support when the manager has a high preference for microin-
volvement, the organization is small, the environment is somewhat
unsettled or hostile, and the strategy is prospector (Table 9.8). Notice here
that the situation fit is realized only for some of the possible situations. A
prospector strategy is appropriate in an unsettled environment and with a
management that participates and makes quick decisions. Referring back
to the configuration discussions, such a situation will also call for a simple
configuration-for a design parameter fit. However, the simple structure is
not recommended when all three dimensions ofthe environment score high.
9.3 Contingency Fit 299

Table 9.B. Continued.

A LOW organizational complexity may be recommended when there is at least

one of the following:

Contingency Factor The Values Proposition

Management style High microinvolvement 3.14

Climate Group 4.9

Size Small 5.4

Environment High equivocality, 6.5

low complexity,
high uncertainty

High equivocality, 6.6

high complexity,
high uncertainty,
hostile 6.9

Strategy Prospector 8.24

The strength ofthe recommended low organizational complexity is greater as the

number of operative contingencies increases.

A high organizational complexity is recommended when the manager has

a low preference for microinvolvement, the organization is large and
private, the environment has high uncertainty but scores low on the other
environmental dimensions and the technology is a routine process
technology. All strategy types can fit a high organizational complexity.
Again, multiple situations can lead to high organizational complexity.

9.3.9 Formalization
For many organizations it is efficient to obtain a standardized behavior of
the members of the organization. This standardization can lead to low cost
high product quality, and generally efficient operations.
Formalization is one way to obtain such standardized behavior and, as
such, is a means to obtain coordination and control. Formalization
represents the rules in an organization.
300 Chapter 9. Organizational Design: A Synthesis

Table 9.9. Formalization Recommendations.

A HIGH formalization may be recommended when there is at least one of the

following:

Contingency Factor The Values Proposition

Management style High microinvolvement 3.17

Climate Internal Process 4.40

Size Large 5.12

Environment Low equivocality, 6.1

low complexity,
low uncertainty

Low equivocality, 6.2

high complexity,
low uncertainty

Technology Routine 7.2

Process 7.5

With modem information 7.6

technology

Strategy Analyzer without innovation 8.22

Defender 8.48
The strength of the recommended high formalization is greater as the number
of operative contingencies increases.

Formalization is related to management preference, size, environment,

technology, and strategy (Table 9.9). A management preference for a high
microinvolvement calls for high formalization and vice versa, Propositions
3.17 and 3.18. Generally, formalization gives the detailed control desired
by the management. When trust is low, and the level of conflict is high, the
rules and regulations must be in place. Formalization should increase with
size to capture the economics of specialization, Propositions 5.10,5.11, and
5.12. There is a caveat and clarification for professionals: namely, that
professionals should have formalization between their professional units.
The professionalization itself serves the same purpose.
9.3 Contingency Fit 301

Table 9.9. Continued.

A MEDIUM formalization may be recommended when there is at least one of the

following:

Contingency Factor The Values Proposition

Climate Rational goal 4.55

Size Medium 5.11

Environment Low equivocality, 6.3

low complexity,
high uncertainty

Low equivocality, 6.4

high complexity,
high uncertainty

High equivocality, 6.7

low complexity,
low uncertainty

High equivocality, 6.8

high complexity,
low uncertainty

Strategy Analyzer 8.22

8.34

The strength of the recommended medium formalization is greater as the

number of operative contingencies increases.

Propositions 6.1 through 6.8 relate the environment to formalization.

Generally, as the environment becomes less well defined, formalization
should decrease. Low equivocal and low uncertainty environments call for
high formalization. Rules and procedures can be effective for this relatively
certain situation, and efficiencies can be realized. For the other extreme of
high equivocality and high uncertainty, formalization should be low.
Appropriate actions are not known, adaptation is mandatory, and rules will
only impede the needed flexibility. The intermediate environments call for
medium complexity. As before, if environmental hostility is high, then the
organization has to react fast and low formalization is needed.
302 Chapter 9. Organizational Design: A Synthesis

Table 9.9. Continued.

A LOW formalization may be recommended when there is at least one of the

following:
Contingency Factor The Values Proposition

Management style Low microinvolvement 3.18

Climate Group 4.10

Developmental 4.25

Size
Small 5.10
Environment
High equivocality, 6.5
low complexity,
high uncertainty

High equivocality, 6.6

high complexity,
high uncertainty

Hostile 6.9
Technology
Nonroutine 7.1
Strategy
Prospector 8.8

The strength of the recommended low formalization is greater as the number of

~erative contingencies increases.

A routine technology calls for high formalization; nonroutine technology,

low formalization, Propositions 7.1 and 7.2: moderated for a professional
organization and strengthened for a service organization. Service organiza-
tions do not have machines to regulate work, so formalized rules take on
increased importance and relevance. Technology affects the formalization:
increased formalization for process technology, as the technology itself does
not drive the workplace, Proposition 7.5. And Proposition 6.6 suggests that
information technology itself leads to greater formalization. The last four
Propositions-8.8, 8.22, 8.34, and 8A8-relate the organization's strategy,
particularly innovation, to formalization. Generally, increased innovation
calls for less formalization. A defender needs the economies offormalization
to survive; an innovator needs the flexibility of less formalization.
9.8 Contingency Fit 303

Again, high and low fonnalization can fit with a number of situations.
Low formalization is strongly recommended when the manager has a low
preference for microinvolvement, the environment is unsettled, and the size
of the organization is small with a nonroutine technology. A prospector
strategy would be appropriate.
A manager with a preference for high microinvolvement in a large
organization in a stable environment using a routine technology following
a defender strategy will have a perfect contingency fit with high formaliza-
tion.

9.3.10 Centralization
Centralization is the degree to which fonnal authority to make discre-
tionary choices is concentrated in an individual, unit, or level (high in the
organization). Decentralization is low centralization. We measure centrali-
zation by how much direct involvement top management has in gathering
and interpreting the information it uses in decision making and the degree
to which top management directly controls the execution of a decision. The
above issues are important to determine who has authority to influence a
decision aside from actually making the decision.
The organization's centralization is influenced by a number of contin-
gency factors (Table 9.10). Centralization is quite dependent on the
preference of the management. The management's high preference for
micro involvement requires a centralized organization; its low preference,
a decentralized organization, Propositions 3.19 and 3.20. This is at variance
with the maxim of universal greater decentralization. Rather, these
propositions indicate that the organization should match and follow the
style of the management. With a high level of trust, high work morale, and
low level of conflict, the centralization can be lower leading to a higher
capacity for information processing than otherwise. When the opposite is
the case more centralization may be needed.
Large organizations should be decentralized; small organizations need
not be so, Propositions 5.5 through 5.9. These propositions are strongly
supported in the literature and managerial experience and widely
recommended. Large organizations simply cannot be run only by those at
the top; infonnation demands are too great, and motivational incentives for
lower management are wanting. Smaller organizations can be centrally
managed.
304 Chapter 9. Organizational Design: A Synthesis

Table 9.10. Centralization Recommendations

A HIGH centralization may be recommended when there is at least one ofthe

following:

Contingency Factor The Values Proposition

Management style High microinvolvement 3.19

Climate Internal process 4.41

Size Small 5.9

Environment Low equivocality, 6.1

low complexity,
low uncertainty

High equivocality, 6.5

low complexity,
high uncertainty

High equivocality, 6.7

low complexity,
low uncertainty

Hostile 6.9

Technology Routine 7.7

Strategy Analyzer without innovation 8.23

Defender 8.49

The strength of the recommended high centralization is greater as the number

of operative contingencies increases.

The environment affects the desired centralization. The well-defined

environment in calls for high centralization; the information processing
demands are low, permitting a high degree of centralization, Proposition
6.1. High centralization is warranted in a very equivocal and uncertain
environment but low complexity, Proposition 6.5. Here centralization
permits a coordination unity and quick reaction when the environment does
not have many factors in the face ofthe unknown. In contrast, for the high-
complexity environment with high equivocality and high uncertainty calls
for low centralization, Proposition 6.6 Here the environment has many
9.3 Contingency Fit 305

factors that are unknown, and decentralization is the only hope as a

centralized organization cannot handle the information demands in a timely
fashion. Environmental complexity is quite important, as it affects the
amount of the information processing demands. Medium centralization is
proposed for intermediate situations of moderate information demands,
Propositions 6.2 and 6.3. High hostility requires a centralized decision
making in order to react quickly to the threats. Technology routineness
suggests a centralized organization: there are few decisions to make, and
top management can make them.

Table 9.10. Continued

A MEDIUM centralization may be recommended when there is at least one of

the following:

Contingency The Values Proposition

Factor

Climate Group 4.11

Developmental 4.26
Internal process 4.41
Rational goal 5.56

Size Large and private 5.6

5.7
Medium 5.8

Environment Low equivocality, 6.2

high complexity,
low uncertainty

Low equivocality, 6.3

low complexity,
high uncertainty

Technology Routine and large size 7.8

Strategy Analyzer 8.23

8.36
The strength of the recommended medium centralization is greater as the
number of operative contingencies increases.
306 Chapter 9. Organizational Design: A Synthesis

Table 9.10. Continued

A LOW centralization may be recommended when there is at least one of the

following:

Contingency Factor The Values Proposition

Management style Low microinvolvement 3.20

Climate Developmental 4.26

Size Large and not private
5.5
Environment Low equivocality,
high complexity, 6.4
high uncertainty

High equivocality,
high complexity, 6.6
high uncertainty

High equivocality,
high complexity, 6.8
low uncertainty

Strategy Prospector and large size

8.10

The strength of the recommended low centralization is greater as the number of

operative contingencies increases.

Finally, strategy and centralization are related. A prospector should be

decentralized, Proposition 8.10. Both information demands and motiva-
tional issues support this proposition. In contrast, the defender in Proposi-
tion 8.49 should be centralized to obtain a coordinated response to the
environment. The analyzers in Propositions 8.23 and 8.36 suggest a
medium centralization. The analyzer without innovation also suggests a
high centralization, as does the defender. The best level of centralization
can be assessed by the organization's need for timely activity but, more
important by the information processing demands. High information
processing demands indicate decentralization; low information processing
dern.ands perrn.it centralization.
A perfect fit with a recommendation for a high centralization is a
situation with a manager with a high preference for microinvolvement in a
small organization. An environment that is stable or at least not complex,
a routine technology, and a defender strategy fit as well. A decentralized
9.3 Contingency Fit 307

organization provides a perfect contingency fit when the organization is

large, management has a low preference for microinvolvement, the
environment is unsettled, and the technology is nonroutine. A prospector
strategy is appropriate as well.

9.3.11 Coordination and Control

Coordination and control can be obtained in many ways. As discussed
earlier, when both centralization and formalization are low, then coordina-
tion and control have to be obtained using other means. This is also one of
many rules that address the situation fit problem. Additionally, there may
be interaction effects: the effect of technology, for example, depends on the
size of the organization. In this section propositions on coordination and
control are summarized.
Size, by itself, does not tell us much about coordination and control
requirements. It must be coupled with another contingency. The environ-
ment is a major influence on coordination and control requirements, where
the less well defined the environment, the greater the demand for coordina-
tion mechanisms. Finally, technology routineness is important.
Consider management's preference for microinvolvement on the
requirement for coordination and control. The situation of a small organiza-
tion that is not centralized or formalized, with a management with a high
preference for microinvolvement, suggests that control will be required and
liaison mechanisms will be appropriate, Proposition 4.22. Quite simply,
management, despite its preferences, cannot be involved in everything and
will require some means of coordination. Other mechanisms have been
eliminated, and thus, liaison activities are to be utilized. A functional
configuration with a management that has a high preference for micro-
involvement will control information and make decisions, Proposition 4.23.
In contrast, Proposition 4.24 considers the functional configuration, low
formalization, and management's low preference for microinvolvement and
suggests that other coordination approaches such as meetings and liaison
activities will be required to obtain the needed coordination among the
functional units. Otherwise, the functional units will drift off into non-
related activities.
The environment affects coordination and control requirements,
Propositions 6.10 through 6.17. A low equivocality, low complexity, and low
uncertainty environment can be coordinated by rules, procedures, planning,
and direct supervision. It is a low information processing situation. As the
need for information processing increases due to a less simple environment,
planning becomes important and liaison activities, integrators and group
meetings will be required. The distinction between environmental
308 Chapter 9. Organizational Design: A Synthesis

equivocality, complexity, and uncertainty yields more precise recommenda-

tions for particular situations. Finally, consider technology. A more routine
technology can be coordinated with direct supervision and planning; as it
becomes nonroutine, the information needs increase and group meetings
will be required. Finally, if the technology has no dominant technology,
other factors must be relied on, Propositions 7.21, 7.22, and 7.23.

9.3.12 Media Richness and Incentives

The appropriate media richness and incentives depend upon the environ-
ment and the technology. Generally, the well-defined environment in
Propositions 6.10 and 6.11 calls for procedure-based incentives; the
increasingly ill-defined environment in Propositions 6.12 through 6.17 calls
for incentives based on results. As the environment becomes ill defined, it
is less clear what to do a priori, and thus, the incentives should reward
adaptability and quick response by rewarding results. To follow incorrect
procedures precisely is the height of folly. Similarly, a well-defined
environment requires a less media richness (Propositions 6.11 and 6.11)
than does a less defined environment (Propositions 6.13 through 6.17).
Finally, consider the effect of technology routineness on incentives. For
a routine technology, procedures are well defined, and it is appropriate to
perform them well. With a nonroutine technology, the goal is to obtain the
derived outcome or results; it is less important which procedures are
utilized. A routine technology (Proposition 7.21) requires a low media
richness where a nonroutine technology (Proposition 7.22) requires a high
media richness.
In general, incentives should be procedure based when the procedures
are well known and will yield appropriate results; if not, the incentives
should focus on the results and encourage adaptation to obtain the desired
results.

9.4 Situation Fit and Misfits

In Chapter 1, the idea of a situation fit was introduced. Basically, it is
impossible to design an organization for optimal performance if it has put
itself into a bad situation. The organization may go through a number of
deliberate stages where misfits are created. This was discussed in Chapter
1, and will be further elaborated on in section 9.7.3.
In this section a number of potential misfit situations are presented. The
concept of situational misfits incorporates theories from strategy, human
resource management, management and leadership. It is beyond the scope
9.4 Situation Fit and Misfits 309

of this book to go into the specifics of all these theories, but we do provide
a link to these theories. Table 9: 11 provides a list of important situational
misfits which are discussed then discussed together with possible ap-
proaches to eliminate the misfits. This also summarizes the relationships
between the variables comprising the organizational context as they were
described in chapters 3-8.

Table 9.11. Propositions on Misfits

9.1. If professionalization is low and technology-routineness is low then

there is a potential situational misfit.

9.2. If the technology routineness is high and the environmental uncer-

tainty is not low then there is a potential situational misfit.

9.3. If the technology routineness is high and environmental equivocality

is high then there is a potential situational misfit.

9.4. If the product innovation is low and the environmental uncertainty is

not low then there is a potential situational misfit.

9.5. If the technology routineness is high and the product innovation is

high then there is a potential situational misfit.

9.6. If the market diversity is high and the product innovation is low then
there is a potential situational misfit.

9.7. If the product diversity is high and the product innovation is low then
there is a potential situational misfit.

9.B. If the strategy is defender and market diversity or product is high then
there is a potential situational misfit.

9.9. If the environmental equivocality is high and the strategy is defender

then there is a potential situational misfit.

9.10. If the environmental uncertainty is high and the strategy is defender

then there is a potential situational misfit.

9.11. If the environmental complexity is complex and the strategy is

defender then there is a potential situational misfit.
310 Chapter 9. Organizational Design: A Synthesis

9.12. If the strategy is prospector and product innovation is low then there
is a potential situational misfit.

9.13. If the strategy is prospector and the environmental uncertainty is not

high then there is a potential situational misfit.

9.14 If the environmental equivocality is low and strategy is prospector

then there is a potential situational misfit.

9.15. If the top management is risk averse and the strategy is prospector
then there is a potential situational misfit.

9.16. If the strategy is prospector and market and product diversity is low
then there is a potential situational misfit.

9.17. If the strategy is prospector and technology routineness is high then

there is a potential situational misfit.

9.18. If the top management has a short term view and the strategy is
analyzer then there is a potential situational misfit.

9.19. If the strategy is analyzer and market or product diversity is low then
there is a potential situational misfit.

9.20. If the environmental equivocality is high and the strategy is analyzer

without innovation then there is a potential situational misfit.

9.21. If the environmental uncertainty is high and the strategy is analyzer

without innovation then there is a potential situational misfit.

9.22. If the environmental complexity is high or environmental complexity

is low and the strategy is analyzer without innovation then there is a
potential situational misfit.

9.23. If the environmental complexity is simple and the strategy is analyzer

with innovation then there is a potential situational misfit.

9.24. If the environmental equivocality is low and the strategy is analyzer

with innovation and then there is a potential situational misfit.

9.25. If the strategy is reactor then there is a potential situational misfit.

9.26. If the capital requirement is high and size is small then there is a
potential situational misfit.
9.4 Situation Fit and Misfits 311

9.27. If the capital requirement is low and size is large then there is a
potential situational misfit.

9.28. If the environmental equivocality is high and the technology type is

mass production then there is a potential situational misfit.

9.29. If the technology type is mass production and the strategy is prospector
then there is a potential situational misfit.

9.30. If the management's preference for microinvolvement is high and the

environmental complexity is high and environmental uncertainty is
high and size is large then there is a potential situational misfit.

9.31. If the climate is an internal process climate and the environmental

equivocality is high or the environmental equivocality is medium then
there is a potential situational misfit.

9.32. If the climate is an internal process climate and the strategy is

prospector then there is a potential situational misfit.

9.33. If the climate is an internal process climate and the technology

routine ness is low then there is a potential situational misfit.

9.34. If the climate is an internal process climate and the managements

preference for microinvolvement is low then there is a potential
situational misfit.

9.35. If the climate is an internal process climate and the strategy is

analyzer with innovation then there is a potential situational misfit.

9.36. If the climate is a rational goal climate and the technology routineness
is low then there is a potential situational misfit.

9.37. If the climate is a rational goal climate and managements preference

for microinvolvement is low then there is a potential situational misfit.

9.38. If the climate is a rational goal climate and the environmental

equivocality is high then there is a potential situational misfit.

9.39. [(the climate is a rational goal climate and the strategy is prospector
then there is a potential situational misfit.

9.40. If the climate is a developmental climate and top management is risk

averse then there is a potential situational misfit.
312 Chapter 9. Organizational Design: A Synthesis

9.41. If the climate is a developmental climate and the technology routine-

ness is high then there is a potential situational misfit.

9.42. If the climate is a developmental climate and managements preference

for microinvolvement is high then there is a potential situational
misfit.

9.43. If the climate is a developmental climate and the environmental

equivocality is low then there is a potential situational misfit.

9.44. If the climate is a developmental climate and the strategy is defender

then there is a potential situational misfit.

9.45. If the climate is a group climate and the management's preference for
microinvolvement is high then there is a potential situational misfit.

9.46. If the climate is a group climate and the environmental equivocality

is high then there is a potential situational misfit.

9.47. If the climate is a group climate and the strategy is prospector then
there is a potential situational misfit.

9.48. If the climate is a group climate and technology routineness is high

then there is a potential situational misfit.

When the organization has a non routine technology, and professionaliza-

tion is low and thus the workforce has a low level of education and training
the situation can create production and service difficulties which usually
require an investment in the education and training. A non routine
technology usually requires that individuals adapt work methods to the
particular task at hand. The individuals must have a sufficiently high level
of skill to make these adaptations. Low levels of education and training do
better at routine tasks and technologies. With a non routine technology and
low level of education and training, new training will be required for the
workforce. This training should emphasize individual responsibility and
decision making for the quality ofthe service. It should provide new skills
which permit the individual to take the initiative for actions which meets
the customers' requirements, Proposition 9.1. A routine technology produces
goods and services efficiently which are standard and without variation.
In an uncertain environment, it is very likely that the customers will
prefer variation in products and services. Competitors are likely to vary
their strategies in products, prices, advertising, etc. New innovative
strategies may be called for. A more non routine technology will likely be
9.4 Situation Fit and Misfits 313

required to adapt to an uncertain environment, Proposition 9.2. In a highly

equivocal environment, it is likely that customers will demand variation in
the product and service characteristics. Further, in the equivocal environ-
ment, large changes can come from unforeseen actions by competitors,
government, and breakthrough innovations. A more non routine technology
will be required to adapt to the unknowns and changes of an equivocal
environment, Proposition 9.3.
With a low product innovation and an uncertain environment, the
situation calls for a review and suggests that the organization should
consider greater product innovation. Low product innovation means the
same products are available for an extended period. In a certain environ-
ment with little change in customer demands and preferences, there is little
need for new products. But, with increasing uncertainty in customer
demand, new competitor strategies, possible governmental actions, shifting
customer tastes, etc., current products are likely to be mismatched with this
changed environment. New products and innovation will likely be required
to adapt and meet the emerging needs and opportunities of the new
environment, Proposition 9.4. Similarly, a routine technology will not
support high product innovation. A routine technology yields standard
products with low variation. The need for product innovation creates a
mismatch. Product innovation will be difficult to manage, expensive and
inefficient. For product innovation, a more non routine and adaptable
technology is required. Of course, the organization may also shift to markets
and products where less product innovation is required and a routine
technology is suitable, Proposition 9.5.
With many markets and a low product innovation, the organization may
want to consider whether the product innovation is high enough. Unless the
many markets are very stable, it is likely that new products should be
introduced; thus, some product innovation is needed to meet these
variations in market demand, Proposition 9.6. Additionally, with many
products and a low product innovation, the organization may want to
consider increasing the product innovation. Low product innovation will
likely create a mismatch with the product demands of the marketplace.
New products and new product innovation will be required to meet the
changing product needs, Proposition 9.7.
A defender strategy is easier to manage for a few products or markets as
management can focus attention on a few issues welL Defender strategies
are difficult to sustain for a large number of products; there is a high
probability that some products and markets will require innovation and
new developments. The recognition of when to give up a defensive strategy
for a given product or market requires a good deal of management attention
and the need for change is frequently missed or realized late; needed
adaptation is then even more difficult. For many products or markets, a
314 Chapter 9. Organizational Design: A Synthesis

uniform defender strategy is likely to be a mismatch for some products or

markets. A complex and mixed strategy is required, Proposition 9.8.
When the environment is not well defined - that is, when the organiza-
tion cannot anticipate the important environmental factors - the defender
strategy is not appropriate, Proposition 9.9. A defender strategy is not
innovative or adaptive. In an ill-defined environment, adaptation will be
required to survive as new situations and issues will emerge. Here, the
organization should change its strategy to an analyzer or prospector to
adapt to the evolving and changing environment. With a highly uncertain
environment, the environmental variable values change in large amounts
and can vary in the extreme. The defender is trying to maintain the status
quo and defend its market share, product lines, production processes, etc.
This is risky as the market or other environmental values change outside
the range of the defender's position. Then there is a serious mismatch
between what the defender is doing and what the market will support.
Here, the organization must prepare to shift quantities and perhaps
products to meet the uncertainty in the environment. An analyzer strategy
is probably more appropriate, Proposition 9.10.
Additionally, when many factors in the environment affect the organiza-
tion, it may make it difficult for a defender to protect what it does and also
difficult to protect its established market position, Proposition 9.11. An
analyzer strategy is more appropriate for this complex environment. Here
the analyzer should seek out opportunities in the complex environment.
A prospector strategy and low product innovation are a serious mismatch
of strategy and capability, Proposition 9.12. Without product innovation, the
organization can develop nothing new. With nothing new to offer, the
prospector will incur high costs and not likely be able to obtain a greater
sales and/or market share. The prospector is committed to offering new and
innovative products and services. With low environmental uncertainty,
there may be little need for new ideas and products. Here, the prospector
may incur high costs with little return. An analyzer, even a defender
strategy would be more appropriate, Proposition 9.13.·
Similarly when the organization both has a prospector strategy and an
environment with low or medium equivocality. With low equivocality, there
is little need for such innovation. The prospector will incur high costs with
little return. An analyzer, even a defender would be more appropriate here,
Proposition 9.14.
A prospector strategy demands a projection into the unknown with new
and innovative products and services, where the returns are uncertain. A
risk adverse management will be very uncomfortable with this high level
of risk. Risk adverse managers prefer situations with less uncertainty. It is
possible to either change the prospector strategy or hire more risk assuming
managers. Usually a risk adverse management will control expenditures to
9.4 Situation Fit and Misfits 315

reduce or eliminate the prospector projects. If the environment and markets

call for a prospector strategy, a new management would be preferable. Some
risk adverse managers can adapt, but it is very difficult, Proposition 9.15.
With a routine technology, developing new products or services will also be
very difficult. With limited product and market opportunity, the range of
prospector possibilities may exceed the environmental possibilities. The
prospector needs to seek new markets as well as new products. If the
markets do not exist or cannot be created, the prospector will incur high
costs ofinnovation without return, Proposition 9.16. The routine technology
is limited in its capacity to vary products or processes. For a prospector, a
more non routine technology is required, Proposition 9.17.
Conflict and confusion are likely results when the organization both has
an analyzer strategy and a management with a short time horizon. An
analyzer is searching for opportunities which may not be within the current
activities of the organization. Frequently, investment and startup costs will
be incurred which will decrease short term returns. Management should
then develop a longer term outlook for the organization, Proposition 9.18.
Generally, more products are required for an analyzer.A few products may
be reasonable in the short run, but an analyzer should be in constant
consideration of new possibilities. When a few, unchanging products become
the norm, the analyzer should broaden its scope of new opportunities,
Proposition 9.19. Likewise when the equivocality the environment is low,
the analyzer with innovation strategy may not be a suitable one. With low
equivocality, the environment is well known and understood. An innovative
strategy works best when the environment offers new opportunities for
products and services, Proposition 9.20. However, even an uncertainty in
the environment may deem an analyzer without innovation problematic,
Proposition 9.21. The analyzer without innovation works best with a
medium environmental complexity-neither too much nor too little,
Proposition 9.22.
When only few factors in the environment affect the organization, the
analyzer with innovation strategy may not be a suitable one, Proposition
9.23. There may be limited need for innovation and adaptation and probably
limited opportunities to which to adapt. An analyzer without innovation, or
a defender strategy which focuses directly on the few environmental factors
and meets market needs efficiently will usually yield better results. A high
equivocality in the environment calls for a capability to vary products and
services as the environment becomes clear. Without an innovative capa-
bility, it may be very difficult to adjust. Copying what others have done may
be possible, but it is not likely to be viable for the long run, Proposition 9.24.
Again without innovation, the organization is limited to copy what others
have done. The organization needs to develop some innovative capabilities
316 Chapter 9. Organizational Design: A Synthesis

to adjust and adapt to the uncertainties in the environment, Proposition

9.23.
A reactor strategy is inherently risky and will ultimately lead to failure.
It is time to reconsider the strategy, Proposition 9.25
With a high capital requirement and not a large organization, the
organization can be vulnerable. An organization with a high capital
requirement and a few employees usually makes a few standardized
products. Further, the technology is likely to be very limited in adaptitive-
ness. The organization is then vulnerable to changes in the environment,
market and products changes. Smaller organizations with small capital
requirements are frequently more adaptive. To reduce this vulnerability,
the organization should consider creating a greater capability for adapta-
tion, which will usually require more employees of higher skill, education
and training, Proposition 9.26.
Low capital requirements are usually associated with low barriers to
entry. Small competitors can enter and frequently, with more advanced
technology and lower costs. Large organizations are frequently slow to
adjust and adapt. The smaller competitor will have the advantage and thus,
be a threat, Proposition 9.27. One alternative is to break up the large
organization into a number of smaller ones.
Most mass production operations are very limited in capacity to adapt
and make different products. Mass production optimizes on the economies
of specialization and standardization. A highly equivocal environment
requires adjustment to the unknown as that environment becomes clearer.
The possibility for mismatch of what the existing mass production can do
and what will be required in the new environment is very high and further
the economic consequences are likely to be great with low return. A highly
equivocal environment calls for a more non routine production capability
than most mass production operations have, Proposition 9.28. Normally a
prospector strategy does not fit well with a mass production technology. A
prospector explores with new products and services with innovative
technology. A mass production will not be able to adjust quickly and support
the prospector. The prospector requires a more non routine technology that
most mass production operations can support, Proposition 9.29.
A large organization with a complex and dynamic environment may not
fit with the managements preference for a high level of microinvolvement,
Proposition 9.30. With a complex and dynamic environment, there are a
very large number of changing situations to which to adjust. Management
cannot access all the situations, analyze what needs to be done and oversee
the implementation. There is simply too much to do; there is too much
information to deal with. A high level of microinvolvement will usually lead
to an information overload at the top and a delay in action when it is most
needed. Despite a tendency for management to become even more involved
9.4 Situation Fit and Misfits 317

in details, the situation requires less microinvolvement and alternative

approaches, such as greater decentralization.
An internal process climate may cause problems in a high or moderately
high equivocal environment. An internal process climate focuses more on
the inside of the organization than on the outside. In an equivocal environ-
ment which is likely to require change and adaptation, the internal process
climate may not either see the shift, understand the need for change and
does not have a capacity which supports adaptation to such needed change.
There is high resistance to change. An equivocal environment requires an
external orientation which is found in the rational goal and development
climates, Proposition 9.31. Similarly an internal process climate does not fit
a prospector strategy. An internal process climate is focused on its own
processes of managing with non equitable rewards, high resistance to
change high conflict and low leader credibility, among other characteristics
which lead to an emphasis on process control according to the procedures
and rules. A prospector strategy requires new products on a continuing
basis with sense of experimentation and exploration. In most situations, the
internal process will compromise the prospector strategy. To realize the
prospector strategy, the climate must change to one that will have a more
external orientation, Proposition 9.32.
An internal process climate is also a mismatch with a non routine
technology, Proposition 9.33. An internal process climate is internally and
control oriented. A routine technology with a similar focus on implementing
standard ways of doing things is a better fit for an internal process climate.
For a non routine technology, a climate with more flexibility would be
better. An internal process climate is also a mismatch with the leader
having a preference for a low microinvolvement!An internal process climate
is internally oriented with a focus on controL Low morale, high conflict, low
trust and high resistance to change, among others require more hands on
management who drive the activities of the organization. This is more
compatible with a higher level of microinvolvement by the leader, Proposi-
tion 9.34. Similarly an internal process climate is a mismatch with an
analyzer with innovation strategy! More flexibility and a more external
orientation are desirable for innovation. An internal process climate
supports better an analyzer without innovation and defender strategy,
Proposition, 9.35
A rational goal climate is a mismatch with a nonroutine technology. A
rational goal climate has an external orientation with a focus on control.
Usually, there is high conflict, low trust and low leader credibility, among
others. A routine technology is a better match. A non routine technology is
difficult to support in this climate. A non routine technology is better
supported with more flexibility and where there is less conflict, higher trust
and greater leader credibility, Proposition 9.36. A rational goal climate is
318 Chapter 9. Organizational Design: A Synthesis

also a mismatch with the leader having a low preference for microinvolve-
ment, Proposition 9.37. The low leader credibility requires that the
management must drive the activities to assure implementation. A leader
with a low microinvolvement is not concerned with sufficient detail to
assure that the activities are in fact realized. A leader who gives attention
to detail and implementation would be preferred for a rational goal climate.
A rational goal climate may not fit an environment that is highly equivocal,
Proposition 9.38. It may be difficult to adapt and adjust to an equivocal
environment where many important variables are not understood. A
rational goal climate is better suited to a less equivocal environment where
there may be some complexity and uncertainty as well. A highly equivocal
environment requires more flexibility A rational goal climate is a mismatch
with a prospector strategy as well, Proposition 9.43. This climate is more
compatible with a defender or analyzer without innovation. A prospector
thrives on innovation and the creation of new things. An organization with
more flexibility, openness and support would work better for a prospector.
A development climate is relatively flexible and externally oriented,
characterized by low conflict, low resistance to change, high leader
credibility, among others. Risk adverse leaders generally avoid uncertainty
and generally also minimize change. The risk adverse leader will not be
comfortable in a developmental climate and may introduce more control to
reduce the uncertainty, Proposition 9.40.
A developmental climate is also a mismatch with a highly routine
technology. A routine technology is more compatible with a climate of
stability, Proposition 9.41. A developmental climate can support a more non
routine technology where adaptation and variation are the norm. Thus, a
developmental climate is also a mismatch with the leader having a
preference for a high microinvolvement, Proposition 9.42. A leader with a
preference for high microinvolvement may challenge some of the develop-
mental characteristics. A leader with a lower need for microinvolvement is
a better fit.
In a low equivocal environment, the environment is known and
understood and the need for adaptability is not great. A developmental
climate can adapt to a more equivocal environment. A developmental
climate and environment with a low equivocality do not fit, Proposition 9.43.
As such a developmental climate is also a mismatch with a defender
strategy, Proposition 9.44. A defender strategy focuses on the status quo
and its continuation. A developmental climate is more compatible with an
analyzer or prospector strategy.
A group climate has a high degree oftrust, low conflict, medium to high
morale, among others. It is internally oriented and has high resistance to
change. A high degree of microinvolvement may threaten the trust and
morale of the organization and actually increase the level of conflict. A
9.5 Design parameter Fit 319

leader with a preference for high microinvolvement may threaten the

climate with hislher involvement and perhaps interference in the organiza-
tion. To maintain the climate, less microinvolvement is required, Proposi-
tion 9.45. A group climate does fit well with predictable environment,
Proposition 9.46. Similarly, a group climate is a mismatch with a prospector
strategy, Proposition 9.47.
A group climate is also a mismatch with routine technology, Proposition
9.48. A group climate can support a more non routine technology with its
capacity to handle and process more complex information.

9.5 Design Parameter Fit

Khandwalla (1973), in a study of seventy-nine manufacturing firms, did not
find a significant difference on average scores on organizational variables
between two groups of high and low performers. However, he did find a
significant difference on the correlation of organizational variables. He
investigated differentiation and integration along the lines of Lawrence and
Lorsch (1967). Those organizations that had both a high integration and
high differentiation or that had a low score on both did better than those
that had a high-low combination. This is consistent with Lawrence and
Lorsch's (1967) argument, but it also shows that design parameter fit is
important. Only those organizations that have the right combination of
design parameters will do well. This is also related to the discussion about
configurations. Should we consider only a limited set of prototype configura-
tions' as Mintzberg (1983) did? It may be easier to consider only such a
limited set of configurations, but the applicability may be equally low.
To obtain a design parameter fit is difficult. Khandwalla's research
showed that it is the interaction effects that are important. Interaction
effects can come from either the contingency factors or from the design
parameter factors. One approach to solving this problem would be for each
possible situation to find the proper organization form. This was the
approach used by Duncan (1979) using decision trees. Even in our simple
setup, however, with 34560 possible situations it is impossible to do so.
Some compound rules are appropriate as we have discussed previously,
but too many will make any model too complex. Another approach is to have
simple rules (as presented in Chapters 3 through 8), use the certainty factor
principle presented in Chapter 1 and then explicitly introduce design
parameter propositions as specific rules.
320 Chapter 9. Organizational Design: A Synthesis

Table 9.12. Design parameter fit, Propositions 9.49-9.53

9.49. If centralization is low and formalization is low, then incentives could

be based on results.

9.50. If centralization is low and formalization is high, then incentives

could be based on procedures.

9.51. If organizational complexity is high, then it is less likely that central-

ization should be high.

9.52. Ifformalization is high, then it is less likely that centralization should

be high.

9.53. If organizational complexity is low, then it is less likely that central-

ization should be low.

The propositions in Table 9.12 are based on the information processing

argument that the capacity should be neither too low nor too high.
Propositions 9.49 and 9.50 suggest that a low centralization coupled with
low formalization requires results-based incentives, but low centralization
with high formalization calls for procedures-based incentives-that is, high
formalization requires procedures-based incentives. Proposition 9.51 states
if organizational complexity is high, it is less likely that a centralized
organization can process the required information. There needs to be a very
strong argument here if high centralization should be recommended. This
view is in the knowledge base represented with a negative certainty factor,
as described in Chapter 1. Proposition 9.52 indicates a tradeoff between
formalization and centralization. Proposition 9.53 gives the tradeoff
between complexity and centralization.
Similarly, Proposition 9.52 says that both a high formalization and a
high centralization may be too costly and even produce conflicts. A
centralized management may not want to obey the rules given and then ask
the employees to break them too.

9.6 Total Design Fit

The objective of strategic organizational design is to obtain total design fit:
situational fit, contingency fit, and design parameter fit. These criteria were
developed in Chapter 1. Briefly, situational fit is fit among the input factors:
9.6 Total Design Fit 321

management style, size, environment, technology, and strategy. Contin-

gency fit ensures that the contingency relations-if-then statements-have
been followed and are compatible with each other. Design parameter fit
considers the compatibility ofthe design recommendations on configuration,
formalization, centralization, and so on. Total design fit is a consideration
of all three fit criteria simultaneously. To illustrate we examine a new case,
Charles Jones, and return once again to SAS.
Charles Jones operates The Jones Company, a small computer-based
design company that provides individualized custom orders. Each order is
an individual job, and procedures require considerable variation and
adaptation to meet customer specifications. The environment is not
complex: there are only a few customers, who focus on price and quality. Yet
the environment is uncertain, as the price competition makes prices difficult
to predict. The nature of the quality is continually evolving, and it, too, is
difficult to predict. The equivocality is high because Mr. Jones is convinced
that he does not understand what will be important to customers next year
or even who his customers might be next year: this year his customers are
advertising boutiques, last year they were architects and next year who
knows.
He thinks that his employees respect him, and there is a great deal of
trust with little conflict in The Jones Company. The resistance to change is
moderate, so the climate could be categorized as a developmental climate.
Mr. Jones runs his business and is involved in every detail-at least, he
would like to be. Despite a predilection to be involved in the designs
themselves, he knows that he must search widely for new business in order
to survive. He experiments with new customers and new design procedures
and spends a good deal of his time developing technical proposals for new
customers. He is a prospector. Can we recommend an organizational design
for the Jones Company?
For Jones, there are no situational misfits. We can therefore without
hesitation concentrate on contingency fit propositions.
The issue is to balance the effect of the various contingency factors. We
introduce the certainty factor approach (presented in Chapter 1) to balance
and weigh the various factors. The effect of management style has been
argued to be strong, and we have assigned a certainty factor (cf) of 40.
Individually, the other contingencies are less important and have been a
assigned a certainty factor of 20. Although we have assigned numerical
values for the certainty factors, the literature gives only qualitative
guidance on the relative importance. The certainty factors are qualitatively
similar to those utilized in Organizational Consultant. A certainty factor of
10 is a weak factor. The factors in our example are summarized in Table
9.13. The summary in the row labeled Total design has been calculated on
the basis of the Mycin principle discussed in Chapter 1.
Table 9.13. Total Design Fit: 'The Jones Compay. Note: cf= certainty factor

Organizational Centralization Formalization Coordination

Configuration Complexity and Control Incentives
Management Style with a high Simple cf40 Low cf40 High cf40 High cf40
_preference for microinvolvement.

Developmental Climate Matrix cf20 Medium cf20 Low cf20 Low cf20 Planning, integrators, Results cf 20
meetings, high media
richness

Size small Ad hoc cf20 Low cf20 High/medium cf 20 Low cf20

Simple cf20

Environmental equivocality, high, Simple cf20 Low cf20 High cf20 Low cf20 Direct supervision, Results cf 50
uncertainty high, complexity low Ad hoc cf20 group meetings, high
media richness

Service organization, unit Ad hoc cflO - Low cf20 Group meetings Results cf 50
production divisible, nonroutine high media richness
technology
Prospector strategy Simple cf20 High or low cf 20 Low cf20 Low cf20
Ad hoc cf20

Total design Simple cf69 Low cf69 High cf62 Low cf67 Planning and inte- Results cf 75
Ad hoc cf54 High cf20 Low cf36 High cf40 grators, direct super-
Matrix cf20 Medium cf20 vision, group
meetings, high media
--- - - - ------ - - --------
richness --- -

-
9.6 Total Design Fit 323

With respect to configuration, the situation fits both a simple, matrix and
an ad hoc configuration, except Mr. Jones's management style favors a
simple configuration. His management style has a strong effect on the
choice, and thus a simple configuration is recommended with a high
certainty factor. -J

The organizational complexity is clearly low. The uncertain environment,

unit production, and nonroutine technology do not favor specialization or
high vertical differentiation. There are contingency conflicts on the centrali-
zation issue. Most contingency parameters lead toward a high centraliza-
tion. Only the prospector strategy recommends a low centralization.
Therefore, centralization is recommended to be high but with a slightly
lower certainty factor than for either configuration or complexity.
Formalization presents an interesting conflictual situation. All con-
tingency factors, except management style, favor low formalization. Despite
the fact that management style individually has a higher weight than the
other contingency factors, the case for low formalization is very strong.
Additionally, the design parameter fit rule suggests that centralization and
formalization should not both be high. The recommendation is that formali-
zation should be low. There is a unanimous recommendation, that direct
supervision, group meetings, and a result-oriented incentive system should
be the coordinating devices. For coordination and control, there is a general
agreement that direct supervision and group meetings are appropriate.
High media richness is required for the environment. And incentives should
be results-based.
This analysis is confirmed by Organizational Consultant. The proposi-
tions presented in Chapters 3 through 8 have been the basis for the Organi-
zational Consultant expert system. In this system, each rule has been
assigned a certainty factor according to the validation procedure described
in Chapter 1. Thus, Organizational Consultant represents a particular
weighing of the various rules. The input is stored in the file "Jones.ocd."
To obtain a design parameter fit the totality must make sense. In this
case it does. All the design parameters fit nicely together, and a total design
parameter has been obtained.
Now assume that Mr. Jones leaves and the new person has a manage-
ment style with a low preference for microinvolvement. This means that
management favors an ad hoc configuration, high organizational complexi-
ty, low centralization, and low formalization. The total design recom-
mendation then changes to an ad hoc configuration. Organizational
complexity will probably still be low, primarily due to the size. Centraliza-
tion will now change to a decentralized structure, primarily due to
management's preference. Formalization remains low, and the incentive
and coordination devices remain the same. However, this design still
represents a total design fit but for a very different organization. The simple
324 Chapter 9. Organizational Design: A Synthesis

structure was centered around a strong management. The ad hoc structure

is much more centered around a strong incentive scheme. This example
shows the process of diagnosing and designing. In this particular organiza-
tion there were no situation misfits. When there are, they must be
addressed first. This issue will be dealt with in a later section on SAS.
Additionally, the example shows that changes in the situation may
change the organizational design dramatically. One has generally to pay
very close attention to minor misfits. As in chaos theory, small changes or
small misfits may have large consequences. The change in management
style rippled throughout the organization. Also small design parameter
misfits involving the incentive structure may have a dramatic effect on the
performance of the organization.

New Forms and for Flexible Organizational Design

In earlier capsules new organizational forms were discussed. In chapter 2, we argued that the
traditional vocabulary and concepts can be utilized for new forms. In Chapter 6, we examined
the relation between hypercompetition and flexibility. In Chapter 8, we examined strategy and
new forms. Here, we want to extend that analysis to suggest the new organizational designs for
flexibility can fit into the contingency framework.

Contingency factors: Flexible Flexible

management style high microinvolvement low

climate rational goal rational goal

size small large

environment high uncertainty, complexity, high uncertainty, complexity,

equivocality equivocality
technology nondivisible, nonroutine divisible, nonroutine

strategy prospector prospector, analyzer

Design:

configuration ad hoc divisional (5.16) (7.17)

organizational complexity low (3.14) (5.4) high (5.1) (8.9)

formalization low (5.10) (6.6) (7.1) low (3.18) (6.6) (8.8)

centralization high (3.19)(5.9) low (3.20) (6.6) (8.10)

coordination meetings, integrators (6.15) meetings, integrators (6.15)

media richness high (6.15) high (6.15)

incentives results based (6.15) results based (5.15)

Figure 9.1 New Organizational Forms

9.7 Designing the Organization 325

Flexibility usually means to be able to change quickly. First, we begin with two organizational
designs which meet the notion of being "flexible." Second, we look at the situation or
contingencies for which flexibility fits. In Figure 9.1 the two designs are given and the associated
contingencies. The contingencies are not unique, i.e., for any resulting design there are a
number of contingency sets which could yield that design. We have only included one
contingency set for each design and included some supporting propositions for the fit.
The first flexible organization is an ad hoc configuration with low organizational complexity,
low formalization, high centralization, meetings for coordination, high media richness and results
based incentives. Can this organization adjust quickly? We argue it can. There are no rules to
restrict change and tasks are general. Centralization is high, yet decisions can be made by
many individuals. Success is measured directly by achieving the organizational goals and the
incentives are matched to that goal. There is little organizational inertia here and it can change
quickly. The next issue is the contingencies which give rise to this organizational design. These
contingencies are not the only ones which could yield this particular design, but these particular
contingencies fit and are possible. The managerial style is one of high microinvolvement where
the climate is rational goal for a small organization. The environment is highly uncertain,
complex, and equivocal. The technology is nondivisible and nonroutine. A prospector strategy
then fits. In brief, this flexible organization is a small, goal driven organization which has a
strategy of exploring the possible. It seems to meet the common notion of flexibility.
The second flexible organization is a divisional configuration with high complexity, low
formalization, low centralization, coordination by planning and meetings, high media richness
and results based incentives. Can it adjust quickly? Again, we suggest that it can. The divisions
are only loosely coupled and can adjust independently. Here again, the low formalization and
low centralization facilitate change. The results based incentives support the change as needed.
The contingencies here are somewhat different. The environment is highly uncertain, complex
and equivocal. The organization can be large and the technology needs to be divisible across
the divisions. Indeed, the technology can be nonroutine and the strategy can also be a
prospector, but could also be an analyzer. It is a large organization which can adjust as needed.
What is a flexible organization? From these two examples, we find low formalization, low
organizational complexity, high media richness and results based incentives for both. For the
contingencies, the environment is highly uncertain, complex and equivocal. The strategies are
change oriented: prospector or analyzer with a nonroutine technology. Do they meet the fit
criteria. The supporting propositions are given in parenthesis in Figure 9.1.
We conclude that new organizational forms can be described and designed in sensible ways
using an information processing framework and also utilizing the knowledge base from
contingency theory. Yet, there is much to be learned about these new forms of organization to
augment and supplement our understanding of organization.

9.7 Designing the Organization

Design is the fit between strategy and structure. Earlier in Chapter 8,
Strategy, we reviewed Chandler's (1962) famous dictum and its counter
proposition that "strategy follows structure." As design, the choice of the
strategy fits the other contingencies: management style, climate, size,
environment, and technology. This is situation fit. Structure follows
strategy is then realized in our design approach by applying the contingency
fit criteria and the design fit criteria. These fit criteria comprise the "if-
then" rules of the previous chapters 3-8. The knowledge base in the OrgCon
embeds the three kinds offit rules.
326 Chapter 9. Organizational Design: A Synthesis

Organizational design is both process and product. The product is the

statement ofthe organizational design in tenns ofthe configuration and the
properties. The process is the set of organizational activities we undertake
to realize the design product. To have a design process, we must have an
operating organization. That is, an organization exists which undertakes an
organizational process to create an organizational design as product, and
on and on. The process is the "thinking" about the desired product.
Organizational design as process and product is sequential and dynamic
and occurs over time. Organizational design is re-design. An organization
exists which undertakes a design process to realize a new design product.
In this sense, organizational design as process is one behavior of the
organization, what happens, and organizational design as product is an
artifact or ideal statement of what we want. Design is both process and
product. Process without product is wandering and aimless. Product
without process is idealism without means of realization.
Misfits enable the process of diagnosis to proceed to design as product.
Earlier in Section 9.4 we stated 48 misfit propositions which involve the
organization's situation. Anyone of these misfits should trigger a diagnosis
to assess the gravity of the misfit and what should be done. Eg, if the
strategy is defender and the environment is uncertain, the organization is
likely to have a serious problem which should be addressed (9.10). This
diagnosis indicates that unless the strategy and the environment are
brought into alignment, there is no organizational design which will
perform well.
The OrgCon incorporates both diagnosis and design. It aids the design
process by asking the designer questions about the current organization,
the contingency factors and then offers recommendations on the design, the
configuration and properties in Figure 1.3.
Design as process involves organizational change; it is time dependent
and dynamic. The process of design is embedded in the reality ofthe current
organization; we seldom if ever begin with a blank piece of paper and
consider the organization at the beginning. The lifecycle ofthe organization
has only one new beginning, but many changes in its evolution. The
management of the organization's lifecycle requires a number of different
organizational designs over time and for different circumstances. The
understanding for the need to change and the realization of the change are
part of the process of organizational design.
Design as process is the way we proceed to create and realize the design
as product. Without design as product, we do not know where we are
heading and thus, the design process is undirected. Design as product
without process is a statement of the ideal without a means to get there. In
this book, we have focused on design as product, but also embedded process
considerations as well.
9.7 Designing the Organization 327

Design is engineering; it is the creation and realization of useful things:

a bridge, a software program or a paperclip (Petroski, 1993). It is a
pragmatic art of error and learning, which is driven by "need" to create a
useful and purposeful tool. The purpose is to create a tool or aid to help
select a design for an organization, also to help the individual learn about
organizational design through practice on cases, projects and real organiza-
tions. Engineering is a normative science; out of all the possible choices, we
want to recommend one or more that is better than others. The recom-
mended designs meet the needs better and further they are feasible or
possible to realize. We want to recommend an organizational design,
configuration and properties, which is efficient, effectiveness and more
viable than those not recommended. It should be possible to realize and
implement; it should work.
Simon (1981, p. 129) summarizes: "it has been the task of the science
disciplines to teach about natural things: how they are and how they work.
It has been the task of engineering schools to teach about artificial things:
how to make artifacts that have desired properties and how to design."
Engineering is also to "error;" it is not an exact science and errors and
failure are part of the process (Petroski, 1985). Bridges fall down, and ifwe
are fortunate, we learn why and correct the error. "To engineer is to learn."
To fail alone is not enough. Failure must be accompanied by new under-
standing and correction. Learning can be the positive side of error and
failure. "To engineer" must also include: to experiment, to fail, to forgive,
indeed, to reward, and to move to a higher level of knowledge. The design
of an organization is a piece of engineering, an ongoing process of learning
how to make the organization work within the limits ofwhat we know about
how organizations can process information.
We explore and examine: design is lifecycle management, design is
exploration and exploitation, and design is learning. Each view incorporates
both process and product as design. These design perspectives provide
insight into its normative basis, its analytical complexity and the process
of search and selection for an organizational design.

9.7.1 Design is Lifecycle Management: Evolu-

tion and Revolution
Design is managing the lifecycle.The lifecycle includes an evolutionary
phase of moderate stability, which is interrupted by short periods of
revolutionary change between the evolutionary changes. The main idea in
lifecycle management is that over time contingencies and circumstances
change, and thus, the organizational configurations and properties should
328 Chapter 9. Organizational Design: A Synthesis

then change. The change is not at random, but we should maintain fit in
response to new size or growth, new environments, new leadership, new
technology, a different strategy and a modified climate. As anyone of the
contingencies is likely to change within a short time, design is an ongoing
dynamic process, where management continues to search for the correct
organizational design and adjusts the configuration and particularly the
properties on an ongoing basis. This is the life of a viable organization.
The lifecycle model is a metaphor which views the organization like the
biologicallifecycle of birth, growth, maturity and death. There are a number
of variations on the general theme. (Banner, 1995, Chapter 17 presents a
review.) The lifecycle models are positive models which describe the
organization over time and find that organizations do tend to follow a
lifecycle. Here we want to examine the life cycle as a management problem.
The organization's design is then a matter of choice at each point and time,
and management must choose to re-design the organization over time if it
is to survive. So we posit that the organization has choice over its own
lifecycle: both the duration of the various stages, eg, maturity may be one
year or much longer; lifecycle stages are becoming shorter. So, the lifecycle
models are one path, perhaps the usual path, but not the only one.
Diagnosis and design help management with its choice.
Throughout this book, we have examined SAS and its evolution. It has
a lifecycle which follows the general pattern. The early days were very
focused on getting the job done of flying airplanes. The next phase involved
a consolidation of function and management where it both found its place
in the air service business and began to rationalize its organization and
processes. Given its regulatory environment and need for control, it quickly
moved to a more formalized organization, which did serve it well when
competition heightened. Since the early 80's, SAS has become a more
complex organization to address its more complex environment. SAS has a
lifecycle. At this point, we can see more clearly its beginning and early
development. Now in the maturity, it continues to evolve and adjust. Will
it decline, as suggested by the lifecycle? It may at some time in the future,
but it continues to strive and address its challenges. It is in a continuing
maturity phase.
There are a number of variations on the lifecycle model. Greiner (1972)
presented a five stage lifecycle: creativity, direction, delegation, coordina-
tion, and collaboration. He suggested different foci, structures, management
styles, control systems and rewards for the various phases. Cameron and
Whetton (1981) elaborated and confirmed the stages as: entrepreneurial,
collectivity, formalization and control, elaboration of structure and decline.
Each stage has its own requirements.
The entrepreneurial stage is a small organization, dominated by the
founder who does what is necessary with a prospector strategy and
9.7 Designing the Organization 329

nonroutine technology in an uncertain environment. He or she is involved

in most everything with a high microinvolvement, yet has a development
climate. As shown in Figure 9.2, a simple configuration is recommended
with low organizational complexity, low formalization, high centralization
where coordination is realized through direct supervision and meetings. The
incentives should be results based. This organization fits together around
the entrepreneur. He or she directs what is to be done; decision making is
centralized. There is no need for formalized rules or elaborate job complex-
ity. Rules would only get in the way of the entrepreneur and restrict the
flexibility that is needed to pursue the prospector strategy and adjust to the
uncertain environment.
In Figure 9.2, supporting proposition number are included in parenthesis
for the configuration and properties. These propositions are not the only
ones which go into the recommendations by the OrgCon. On the accompany-
ing CD, you can find the file Lifecycle.ocd which contains the input data for
Orgcon for the entrepreneur. You can review in more detail the data on the
management style, climate, size, environment, technology and strategy.
Using the OrgCon, you can then run the program and review the recommen-
dations for the Entrepreneur and the situational misfits. These files contain
more information than in Figure 9.2. And further, you can do sensitivity
analysis on the input data to investigate the nature of the entrepreneurial
stage of the organization.
For the entrepreneurial stage, the OrgCon found three situational
misfits: a nonroutine technology and a low level of training for the
workforce, a prospector and a few products, and a high microinvolvement
and a development climate.These misfits are then the signals that a more
revolutionary change may be needed. There are many ways to deal with
these misfits. First, they can be observed as management problems, but not
dealt with. Second, they may point to larger problems which cannot be
ignored. The entrepreneur may consider increasing the workforce skill level
for the nonroutine technology and the prospector strategy in order to
develop more products. At the same time, he or she will have to back off
from the high microinvolvement and high centralization to further develop
the climate. Frequently, the need for revolutionary change is observed that
the size becomes too large for the highly involved entrepreneur and he or
she cannot process the information. Here we can see other possible misfits
which can also lead to the need for more revolutionary change. A sensitivity
analysis on the environment indicates the effects and limits of the
environment on this stage. We suggest that the need for change goes beyond
size alone.
The second stage ofthe lifecycle is the collectivity stage. In Figure 9.2,
the contingencies and recommendations are given. It is a medium size
organization with a medium level of microinvolvement by management and
330 Chapter 9. Organizational Design: A Synthesis

a development climate. The environment has high complexity, medium

uncertainty, low equivocality and low hostility. The technology is medium
routine and the strategy is an analyzer with innovation. The recommended
configuration is a matrix or functional, with medium levels of organizational
complexity, formalization, and centralization. Coordination should be with
meetings and planning and the incentives should be results based.
This is a larger organization where the entrepreneur is giving way to less
direct involvement; there are the beginnings of more defined organizational
properties. There is a need for coordination which is obtained with planning
and meetings, rather than direct supervision. The misfits are of two kinds:
situational and design. The situational misfit suggests that the environ-
ment may not require an innovative strategy. The design misfits suggest
that the level of rules, procedures, etc. may not be sufficient, which hints at
a need for a more revolutionary change. For more detail, examine the file
Lifecycle.ocd on the CD.
The third stage of the lifecycle is the formalization and control stage in
Figure 9.2. Here the organization has become even larger, but the
management have a lower microinvolvement style and the climate is now
rational goal or internal process. The environment is complex with medium
uncertainty and low equivocality and low· hostility. The technology is
routine and non-divisible and the strategy is an analyzer with innovation.
The recommended configuration is functional with high organizational
complexity, high formalization and medium centralization. Coordination is
achieved through rules and the incentives are then procedural; we want the
rules and procedures implemented. The situational misfits suggest that a
routine technology and an uncertain environment as well as a low
equivocality with an analyzer with innovation strategy may not fit. The
organization may be too well defined to adjust to the changing environment.
The formalization organization is on the CD under the file Lifecycle.ocd.
The fourth stage of the lifecycle is the elaboration of structure stage in
Figure 9.2. The organization has grown even larger, the climate is a rational
goal, the management has a low microinvolvement. Uncertainty remains
high with medium uncertainty and medium equivocality and low hostility.
The technology is divisible and medium routine, and the strategy is an
analyzer with innovation. The recommended configuration is now divisional
with high complexity, medium formalization, low centralization. Coordina-
tion involves professionalism and the incentives should be results based. It
is a classic divisional organization with design fit and good total fit. The
organization can process the requisite information to deal with its
challenges. You can review the elaboration on the CD in file Lifecycle.ocd.
The fifth stage is the decline stage which is left largely blank. in Figure
9.2. The reactor strategy is not viable and we have not included any
recommendations.
 Contingency factors: Entrepreneurial Collectivity Formalization Elaboration Decline
and control of structure

Management style high medium low, medium low

(microinvolvement)

Climate developmental developmental internal process rational goal

Size small medium large large

Environment
Equivocality medium low low medium
Uncertainty high medium medium medium
Complexity medium high high high
Hostility medium low how low

Technology non-routine medium routine routine, medium routine

little divisible nondivisible divisible

Strategy Prospector analyzer with analyzer with analyzer with reactor

innovation innovation innovation

Misfits nonroutine technology and numerous situational and medium uncertainty and rou- rational goal climate and
low training level,high design misfits tine technology, low equivocali- low microinvolvement
microinvolvement and de- ty and analyser with innova-
velopmental climate tion, internal process

Recommendations
Configuration simple (3.25) (8.11) (5.15) matrix (4.23) (8.37) functional (6.11) (7.16) divisional (6.11) (8.25) (4.53)
Complexity low (3.14) (5.4) (8.9) medium (6.2) (4.24) high (3.15) (7.11) high (3.5) (4.54) (8.24)
Formalization low (5.10) (7.1) (4.25) (3.18) medium (5.11) (8.34) (6.2) high (4.40) (5,12) (6.2) (7.2) medium (4.55) (8.22)
(8.8)
Centralization high (3.19 (5.9) medium (8.36) medium (5.6) (6.2) medium (5.6) (4.56) (6.2)
Coordination meetings, planning, direct meetings, planning (4.28) rules (6.11) professionalization
supervision (4,28)
Incentives results (6.12) results (6.13) procedural (6.11) results (4.60) (6.13)

Figure 9.2 Lifecycle Designs, Cameron & Whetton Stages, ( ) Supporting proposition
332 Chapter 9. Organizational Design: A Synthesis

The lifecycle model is both evolutionary and revolutionary. The organization

must adjust continually within each stage; it must change more fundamen-
tally between stages. Yet, the identification of stages and stage changes is
problematic and easier to observe ex post than in the management process.
The time duration of each stage is not obvious to the manager. Triggering
mechanisms for stage change are more involved than size alone. From a
managerial view, the design of the organization is ongoing and constant.
Some changes are larger than others, but the managers themselves may not
know the distinction between evolution and revolution in design. Our
examination of SAS can be mapped onto a lifecycle and it is interesting to
identify the various stages. The lifecycle is an interesting metaphor, but at
the same time, an organization is not an animal; it is an artifact of our
creation, which can and should change in more complicated fashion. There
are many contingencies and many responses; the organization will evolve
in a very complicated manner. The traditional organizationallifecycle is one
path of many which the organization may take over time. The design
question is to choose from many possible paths on an ongoing basis-to
maneuver in the face of this multivariable and complex problem. This is the
challenge to keep the organization viable.

9.7.2 Design is Exploration and Exploitation

March (1991, 1994) developed the concepts of exploration and exploitation
as action frames for the future of an organization. Exploration means
"adaptiveness" and includes: "result, dynamic and long run" (1994, p.1) and
"search, variation, risk taking, experimentation, play, flexibility, discovery,
innovation" (1991, p.71); exploitation includes "efficiency, rules and
routines, static, short run," (1994, p.1) and "refinement, choice, production,
efficiency, selection, implementation, execution" (1991, p.71). For our design
goals, exploitation is closely related to efficiency and effectiveness;
exploration to viability Either approach to the exclusion ofthe other is likely
to lead to death of the organization. Exploration alone is costly without the
benefits of producing products or services efficiently. Exploitation alone
leads to a trap of short run efficiency, but long run mismatch with its
challenges and opportunities. Management is to find a balance of explora-
tion and exploitation.
In our terms, exploitation is very close to finding a "fit" for the organiza-
tional design. We want to find an organization which exploits the environ-
ment and is efficient and effective. Exploration is more concerned with the
viability of the organization and the creation of selected "misfits," mostly
situational misfits. Of course, viability can not be assured through the
creation of misfits alone and hence, viability also incorporates exploitation
9.7 Designing the Organization 333

as well. In Figure 1.3, the concept of fit includes three aspects: situation,
contingency and design.
Exploration can generate misfits in at least two ways: managerial and
natural. Managerial misfits are conscious and explicit creation of situa-
tional misfits by the adoption of a new management style, strategy or
technology for the organization. Eg, management may adopt a new
technology. which does not fit with the current strategy. In this book and in
the OrgCon, we suggest what the situational misfits are and where to
explore for resolution, if management chooses. Situational misfits can be
purposeful and part of the organization's exploration. Situational misfits
can also arise naturally outside the organization, without managerial
choice. One frequent natural situational misfits occurs when the environ-
ment for the organization changes and the organization's strategy does not.
Eg. a defender strategy which was appropriate for a low uncertainty
environment may not be appropriate for a highly uncertain environment.
In short, nature may explore itself and create misfits.
Exploitation is misfit reduction and establishing the design fit for the
organization. In Figure 1.3, the design fit criteria are: situation, contingency
and design. Here, the goal to find a design which is efficient and effective.
In the diagnosis and design of an organization, we could consider the design
question:

• Ifwe want an organization which will explore well and exploit less well,
what should the organization design be?

and,

• Ifwe want an organization which will exploit well and explore less well,
what should the organization design be?

There are very large numbers of possibilities here. We propose a two step
process: first, we will suggest likely strategies, management styles, size and
technology for exploration and exploitation, and then second, we will utilize
the design concepts to suggest a good design. The first phase is somewhat
exploratory itself. The second phase exploits what we know. Exploration
and exploitation are then intermediate design properties in this view.

• Ifthe organization is an explorer, then the strategy should be prospector,

management style should be low microinvolvement, size should be small
or medium, and technology should be nonroutine.

This organization is similar to the entrepreneur stage of the lifecycle.1t is

one possible way to be an explorer.
334 Chapter 9. Organizational Design: A Synthesis

• If the organization is an exploiter, then the strategy should be defender,

management style should be high microinvolvement, size should be
medium or large, and, technology should be routine.

This organization is similar to the later stages of the lifecycle.

These propositions are in part definitional.. Exploration is adaptiveness.
A prospector is adaptive in the extreme and creates its own environment.
An analyzer is also adaptive to its environment. Exploration requires a lot
of information to be processed from many sources. A highly microinvolved
management will become overwhelmed. A large organization usually will
not be able to explore and act quickly. And, a nonroutine technology is
required to explore. Exploitation focuses on efficiency and doing things well.
It need not adapt well or quickly; it can defend. Routine technology is more
efficient and management can be more involved and still manage the large
organization.

The second phase, the design phase is then:

• If organization is an explorer, then the organization should be (have) a

matrix, ad hoc configuration, low formalization, high decentralization,
medium organizational complexity, coordination by meetings, liaison
roles, committees, results based incentives, and high media richness.
• If organization is an exploiter, then the organization should have (be) a
functional, machine bureaucracy configuration, high formalization, high
centralization, high organizational complexity, coordination by rules,
procedural based incentives, and low media richness.

In many ways, an exploring organization is the mirror image of an

exploiting organization, and vice versa. The optimal balance between the
two extremes depends upon a more detailed statement of contingencies for
the organization as suggested in Figure 1.3. The exploration-exploitation
pair provide a way to think about the realization of the viability, efficiency
and effectiveness goals of the organization.
The exploration-exploitation pair can also be considered in two phases:
the diagnosis and design phase and the operations phase. In the above, the
diagnosis and design phase have been discussed. That is, given the
situation, what is a good design for the organization. Here we want to turn
to the second phase: the operations phase for the organization. For the
above organizations, we posit that:

• Ifthe organization has a matrix or ad hoc configuration, low formaliza-

tion, high decentralization, low organizational complexity, coordination
9.7 Designing the Organization 335

by meetings, etc,. results based incentives and high media richness, then
the organization can explore well.
• If the organization has a functional or machine bureaucracy configura-
tion, high formalization, high centralization, high organizational com-
plexity, coordination by rules, procedure based incentives and low media
richness, then the organization can exploit well.

We can generalize that an organization which explores well will not exploit
well and vice versa. There is a tradeoff and a choice. Although the
discussion above states the propositions in contrasts and extremes, we also
suggest the choice is more complicated; there are multiple intermediate
designs which strike different balances between exploration and exploita-
tion.

9.7.3 Design is Learning

March's exploration-exploitation is also a frame for learning, organizational
learning. AB discussed above, the exploration-exploitation requires a
balance and either extreme is not viable. March focuses on two other
aspects of the balance: risk and investment. Risk attitude is central to
striking the balance: greater exploration requires a greater tolerance for
risk and failure, which is then linked to the incentives and rewards for
success and failure. March posits that most new ideas, or explorations are
bad ideas, or will tum out to be bad ideas. Obviously, we do not know which
ones a priori, or we could avoid them from the beginning. Exploration then
requires activities which have unknown outcomes with a high probability
offailure. Iffailure is highly punished, then individuals will avoid explora-
tion and the organization will not learn, or learn very little and slowly. In
order to explore and learn, we must not only tolerate, but encourage risk
taking which necessarily involves failure. Sitkin (1992) argues that small
failures are preferred to large ones and that a learning organization should
encourage small failures, which permit the organization to learn for greater
success. This approach creates a balance of exploration with risk. Whether
an organization should encourage very large explorations and associated
risks will depend upon the desired balance.
Levinthal and March (1993) argue that many, ifnot most organizations
strike an improper balance, excessive emphasis on short run exploitation
to the determent oflonger run exploration; it is too myopic. In the short run,
individuals are rewarded for risk averse behavior, which puts a premium
on exploitation. Over the longer run, there is too little exploration and the
performance ofthe organization suffers. Ofcourse, exploration is not meant
to be random and without purpose; and further, success should be rewarded
336 Chapter 9. Organizational Design: A Synthesis

more than a failing but valiant attempt. There is an underinvestment in

exploration and an overinvestment in short run exploitations, which will
persist until exploration is rewarded by the organization.
They offer two mechanisms for learning and exploration: simplification
and specialization. Simplification involves the construction of buffers:
specialized individuals who are charged to learn with targets, search and
slack. Decomposition of problems into subproblems is the second approach.
Decomposition and its relation to the organization structure is the most
basic mechanism. Loosely coupled organizations are better for learning,
where tightly coupled are better for error detection and control (p. 92). As we
have discussed in Chapter 2, divisional configurations are more loosely
coupled than are functional configurations. The divisions are less connected
than are the functions. Further, low formalization with fewer rules and high
decentralization tend to spread decision making and decouple the organiza-
tion. Decomposition is one means to keep the risk offailure from exploration
small. And clearly, the incentives for risk taking are important. Galunic and
Eisenhardt (1996) found that divisional organizations can obtain new areas
to explore and develop through the assignment of charters to the division.
That is, a division will be chartered to initiate a new product or line of
business. The large corporation can learn by assigning learning tasks to the
smaller divisions, i.e., decomposing the problem into a smaller one.
Cohen and Levinthal (1989, 1990) develop the concept of absorptive
capacity of an organization as a measure of the organization capacity to
leanl. and innovate. They (1989) demonstrate that organizations which have
research capability leanl. more quickly than organizations which do not
have a research capacity. This capacity may not lead to new and fundamen-
tally new ideas, but it does facilitate the understanding and evaluation of
what others are doing, and then leads to a reasoned and timely adoption of
innovations and new technologies. The organization can exploit outside
information effectively and timely. Research capacity keeps the organiza-
tion at the edge of new technology and permits it to innovate, even ifit does
create the new edge of technology. Here, learning has two important
elements: evaluation of what is important and secondly, knowing early on.
Without the capacity, organizations tend to learn too late for the innovation
to be of value. It helps explain why an organization should invest in basic
research, even though most basic research is in, or will become part of the
public domain.
Another approach to learning is to examine the barriers to learning and
why organizations have difficulty taking on new challenges and ideas.
Probst and Buchel (1997, p.71-72) argue that the hierarchal organization
design itself can create barriers to information processing and leanl.ing.
Greater organizational complexity means that the organizational work is
partitioned and broken up into finer pieces and the individuals who are
9.7 Designing the Organization 337

assigned to the separate jobs can be further apart. Thus, there is a greater
need to process information for coordination. The focus is first on opera-
tions. Second, the "new" and different ideas needed for learning may be
driven out. A higher vertical differentiation creates more levels between the
top and bottom of the organization, where each level can facilitate informa-
tion flow, but can also impede it. The greater the number oflevels, the more
likely that distortion and impediment will occur. The organization is then
limited in its ability to learn from within itself, but also to share informa-
tion for learning from the outside. One solution is to increase the informa-
tion processing of the various layers. Another approach is to simply
eliminate some of the layers, typically middle management in what we call
"delayering." But will delayering increase organizational learning? Not
necessarily, the fundamental problem remains. New ideas may not be
generated, considered and adopted. We conclude that high organizational
complexity can be a barrier to learning, but simple delayering may not
generate learning by itself; it only makes it easier.
In similar manner, the horizontal differentiation of a functional organi-
zation can be barriers to learning. Learning can take place within the
functions, but it much more difficult to generate challenges and new ideas
which span the functions. Individuals within the functions observe problems
and opportunities more readily within functions and further they can deal
with them easier. Interfunctional problems, eg, a production problem across
marketing and operations, is more difficult to observe, define and mount the
necessary actions than issues within the function. A traditional approach
is to create liaison activities: boundary spanners, committees, taskforces,
etc. to solve problems, but also to learn in new ways. In brief, the argument
is that the greater the partitioning of the work, the more difficult it is to
learn. The challenge is to find mechanisms which facilitate learning.
As stated above, Levinthal and March (1993) offer two mechanisms for
learning and exploration: simplification and specialization. Simplification
involves the construction of buffers: specialized individuals who are charged
to learn with targets, search and slack. Liaison roles, committees, issue
champions, and boundary spanners are examples. Decomposition of
problems into subproblems is the second approach. Decomposition and its
relation to the organization structure is the most basic mechanism. Loosely
coupled organizations are better for learning, where tightly coupled are
better for error detection and control (p.92). Divisional configurations
partition the organization into to loosely coupled subunits. However, the
coupling should not be total such that each unit is independent. The
organization must be coupled enough to observe and learn from others and
from the interfaces, but at the same time, not coupled so tightly as to
impede experimentation and learning.
338 Chapter 9. Organizational Design: A Synthesis

Some organizational designs are more conducive to learning than others.

There is a needed balance between being coupled just enough to learn and
not too tightly connected to drive out experimentation and exploration. In
terms of configurations, we suggest that simple, ad hoc and divisional
configurations are more conducive to learning than functional and bureau-
cratic configurations. As discussed above, greater complexity is likely to
impede learning. Clearly, greater formalization will minimize experimenta-
tion and learning. Greater decentralization is more likely to enhance
learning. Higher media richness and general knowledge about the outside
as Levinthal and March (1989) argue should support learning. And finally,
results based incentives which are tolerant of failure should also help.
However, this discussion is primarily conjectural; organizational learning
remains a rich area for research, despite recent development and progress.

9.8 SAS
Since 1950, SAS has experienced four distinct phases in its history: the
relatively stable 1950-1975 period, the environmental threat of the
1975-1981 period, the initial Carlzon period of the 1980s, and today's new
challenges. Each phase has its own special characteristics, and we want to
examine each as well as the transitions. Situation, contingency, and design
parameter fits will be examined to illustrate the influence of misfits of the
organization and how one misfit ripples throughout to create problems
elsewhere. We begin with the first phase-1950-1975.
The 1950-1975 period for SAS has been described at some length.
Briefly, it was a relatively stable period. To review, SAS was a large organi-
zation with a few thousand employees. The environment was described as
low equivocality, low complexity, and low uncertainty. The technology was
highly routine and nondivisible. SAS had a defender strategy and a group
climate. The components of this situation fits well, and no misfits are
identified.
During this period, SAS had a dominant functional configuration
organization which overlay a geographical configuration. It had high organi-
zational complexity, and centralization was high. Formalization was high.
Abundant rules were utilized to achieve coordination and control. Incentives
were procedure based. This organizational design is consistent with
propositions of design. Proposition 5.16 gives wide latitude on the choice of
a configuration, which includes a functional configuration. The stable
environment (Proposition 6.10) further suggests a functional configuration.
And finally, a defender strategy uniquely requires a functional configura-
tion (Proposition 8.47). The group climate suggests a less rigid structure.
The organizational complexity should be high (Proposition 5.1) because of
9.SSAS 339

SAS's large size and its status as a nonpublic organization. Proposition 7.9
augments this recommendation for a routine technology, and this is
consistent with a defender strategy (Proposition 4.48). Clearly, organiza-
tional complexity should be high.
For formalization, Proposition 5.12 suggests that a large organization
should have high formalization. The stable environment (Proposition 6.1)
calls for high formalization, medium complexity, and medium centraliza-
tion. Routine technology (Proposition 7.2), modern information technology
(Proposition 7.6), and a defender strategy (Proposition 8.48) all require a
high formalization.
Considering the centralization, Proposition 5.5 indicates a large
organization should be decentralized, and Proposition 6.1 indicates a higher
centralization for this environment. These two recommendations are in
conflict. Routine technology (Proposition 7.8) indicates a medium centraliza-
tion, and the defender strategy (Proposition 8.49) indicates high centraliza-
tion. Centralization is driven in both directions with these conflicting
recommendations. Finally, the coordination and control (Propositions 6.10
and 7.21) should utilize rules, planning, and procedure-based incentives.
Generally, Propositions 4.1-4.7 suggest a looser structure with less
formalization and control than would otherwise be required.
The SAS organizational design is largely consistent with the recom-
mended design. There is an excellent fit for the functional configuration,
high formalization, high organizational complexity, and coordination rules.
Centralization has conflicting recommendations oflow, medium, and high.
This seeming conflict could suggest a centralized strategic decision making
and more decentralized decision making on operative issues. This would be
compatible with high formalization: using parameter fit rules, we note that
high centralization and high formalization are probably overkill and would
lead to a very inflexible organization. However, in the main, SAS was well
designed for its situation during the 1950-1970 time frame.
The second period-1975-1981 created numerous contingency misfits. In
general, its situation changed with the oil crisis of 1975, but the organiza-
tion did not change. In 1975, suddenly the environment changed from low
equivocality, low uncertainty, and low complexity to high on all three
dimensions. The crises also affected the organizational climate. The impli-
cations were great, and yet the adjustments were minimal. Propositions
4.23 and 6.15 indicates that the functional configuration is now question-
able, media rich meetings and integrators need to replace rules, and
incentives should be results based rather than procedural based.
The centralization, organizational complexity, and formalization
(Proposition 6.6) should now be unambiguously lower. Suddenly, contin-
gency misfits abound between the situation and the organizational design.
Configuration, formalization, complexity, centralization, and coordination
and control are out of balance. It is probable that SAS's management had
340 Chapter 9. Organizational Design: A Synthesis

a high preference for microinvolvement. The environmental change by itself

created a number of contingency misfits. Besides the contingency misfits,
the new environment created situational misfits as well. High equivocality,
high uncertainty, and high complexity are not compatible with a defender
strategy. For the most part, SAS maintained a defender strategy, perhaps
becoming more cost conscious than previously.
Therefore, in the 1975-1981 period, SAS had a serious contingency misfit
problem, a probable situation misfit, but the design parameter fit was fine.
Within itself, SAS was a consistent organization, but in misfit with its
environment, at least.
Jan Carlzon's appointment began the third period. Early in 1981,
Carlzon began to change a number of things. He brought a low microin-
volvement preference management style; he recognized the changed
environment; he changed the strategy from a defender to an analyzer with
innovation. This new situation now was consistent, and there were no
situational misfits.
The organizational configuration is recommended to be matrix or
divisional for analyzer with innovation (Proposition 8.37) but not a
divisional configuration for a nondivisible technology (Proposition 7.17). The
new environment (Proposition 6.15) further supports a matrix configura-
tion. As mentioned above, the new environment (Proposition 6.6) calls for
low formalization, low complexity, and low complexity. Yet SAS remained
a large organization, which suggests a high organizational complexity
(Proposition 5.1) as does the low microinvolvement (Proposition 3.15). The
routine technology (Proposition 6.9) similarly calls for a high complexity as
does the new strategy (Proposition 8.35). Overall, complexity should remain
high. All of the above is also consistent with the developmental climate.
Formalization is suggested to be high by the routine technology
(Proposition 7.2) and the large size (Proposition 5.12) but is driven to be low
by the low microinvolvement (Proposition 3.18) and developmental climate
(Proposition 4.25). Formalization is pulled in opposite directions.
In contrast, centralization should be low, due to low microinvolvement
of management (Proposition 3.20) and developmental climate (Proposition
4.26), large size (Proposition 5.5), new environment (Proposition 6.6), and
to a lesser degree the new strategy (Proposition 8.36). Centralization should
be low.
The environment (Proposition 6.15) calls for coordination and control by
group meetings, integrators with high media richness, together with
results-based incentives. This again fits the developmental climate
(Propostions 4.28-4.30). However, the routine technology (Proposition 7.21)
pulls in the opposite direction, with rules and procedure-based incentives.
Overall, the new SAS of 1981 fits the new situation well. The manage-
ment style of low microinvolvement, the large size, the dynamic environ-
9.9 BAS and the Lifecycle 341

ment, the routine nondivisible technology, and the analyzer with innovation
strategy fit together: there were no situational misfits. The routine,
nondivisible technology would require special attention in this environment.
The contingency fit was good but not perfect. The organizational configura-
tion of the recommended matrix is difficult to realize. SAS experimented
with various divisional forms based on region, customer, and so on, but none
were entirely satisfactory. Formalization remained high, due to the
technology, as did complexity. Centralization was made low, and many rules
were displaced. In brief, the contrasting dictates of the technology and large
size versus the dynamic environment pull SAS in opposite directions. It is
a delicate balance to obtain a well-performing organization.
The fourth period-the late 1980s and until the middle of the 1990s-has
been a difficult one for SAS. In the late eighties, Carlzon's measures to
renew SAS's offerings and find new ways for exploration turned SASs
strategy into more of a Prospector strategy. This however was not consis-
tent with SAS's design parameters (Propositions 7.8, 7.9, 7.10), and SAS
lost control of costs and finally the company's competitiveness deteriorated.
Additionally, the problems affected the climate negatively but creating a
higher level of mistrust and numerous conflicts.
Since 1994, Stenberg's attempts to change the company's focus on its
core business, allowing SAS to pursue an analyzer with innovation strategy
seems more appropriate. There is now situational fit as well as contingency
fit. And the company's configuration is consistent with the design parame-
ters (7.39,7.38,7.40). The environment remains uncertain, complex, and
equivocal. Technology in terms of routine and divisibility are largely
unchanged, despite enormous changes in technical details, particularly
computers. In many ways, the airlines pioneered with integrated computer
operations, and they continue to innovate. An analyzer strategy fits this
new world of continuing change, where mergers and alliances have created
even larger airlines but where also the small-niche airline has emerged.
SAS can be formally analyzed by Organizational Consultant for the four
periods informally discussed in this chapter. Input is stored in the files
"SAS.ocd."

9.9 SAS and the Lifecycle

SAS illustrates organizational change. The 1950-1975 period was a rela-
tively calm time when SAS was well organized to meet its challenge. In the
mid-1970s, the environment changed rather dramatically, and SAS did not
adapt quickly. This is not an unusual situation. Jan Carlzon became the
organizational change agent; he introduced a new management style, mani-
pulated the climate, and changed the company's strategy. SAS was changed
342 Chapter 9. Organizational Design: A Synthesis

to a new organization. Our informal analysis and Organizational Consul-

tant results provide recommendations for change as well as a useful guide
for a retrospective view of what happened. For the most part, SAS adjusted
appropriately and well; however, SAS continues to face difficult issues. Our
analysis suggests that major organizational changes are not required.
SAS is an interesting illustration, but is it general to the change process?
Most organizations face the challenge of a new environment, new technol-
ogy, or even a new management style. The requisite change is usually
difficult. SAS illustrates both well. However, SAS is not a good illustration
of the rate of change that many organizations are experiencing today. SAS
had five or six years to begin to change. Many organizations, particularly
smaller organizations, have five or six months, if that. Consequently,
organizational change is now a continuing process, where previously it was
more intermittent. Configurations and organizational characteristics are in
a continuing state offlux.

9.10 Summary
In this chapter, we put the pieces together for organizational diagnosis and
design. The design propositions have been grouped for design implications:
when would a particular design recommendation be made? A functional
organization, for example, is likely to be recommended when the organiza-
tion is medium or large, the environment is relatively certain, noncomplex,
and unequivocal, the technology is routine, and the strategy defender or
analyzer without innovation (Table 9.2). Each antecedent makes a func-
tional recommendation more likely but not certain. Even though some of
these conditions are met, the particular organization may have other
contingencies that are more important and lead to a different recom-
mendation.
Additionally, different views on organizational design were presented
including views on both product and process. The dynamics of the organiza-
tional design process were presented and illustrated by the life-cycle
approach to organizational design and redesign, exploration-exploitation,
and learning.
Design remains a difficult and complicated task. We further develop the
design approach by reexamining the criteria-situation fit, contingency fit,
and parameter design fit-together make up total design fit.
The design approach is applied and discussed for the Jones Company.
Certainty factors are reintroduced and illustrated. Design balance and
conflicting design implications are resolved using certainty factors. A final
design recommendation rests on a balanced and total consideration of the
situation.
9.12 Summary 343

Describe your organization:

configuration, complexity, formalization,
centralization, coordination and control.

Describe the contingencies:

management style, size, environment,
technology, and strategy.

Does the situation fit

together?

/
NO! YES! They are acceptable.

Develop recommeded changes, if any,

Devise a plan to change for the design of the organization.
them, or let them be? Are you satisfied?

/
NO! YES!

~ ~
Examine the con· Examine the sensitivity
tingencies for the analyzes for insights and
sensitivity analysis. understanding.

Develop implementation.
STOP

Figure 9.3. Checklist for Design Fit

The design process is mapped out in Figure 9.3. We begin with a description
ofthe organization: the current organization and the current contingencies.
We then ask whether there are situation misfits. A situation misfit can be
either resolved or lived with. Generally, we would prefer to resolve the
situation misfit. Next, design recommendations are developed. The design
approach has been illustrated throughout by a continuing reference to SAS
344 Chllpter 9. Organizational Design: A Synthesis

and other organizations. The design recommendations and charts can be

useful guides. Given a recommended design, it is always appropriate to do
a sensitivity analysis on selected inputs: inputs that are uncertain or can be
easily changed. Finally, with an acceptable design recommendation, an
implementation plan should be devised. This design process is systematic;
it considers comprehensively the relevant input data for a design
recommendation that fits the situation and an organization that fits
together.

References
Cameron, Kim and David A. Whetton. 1981. "Perceptions of Organizational
Effectiveness Over Life Cycles," Administrative Science Quarterly, 26,
Cohen, Wesley M. and Daniel A. Levinthal. 1989. "Innovation and Learning: The
Two Faces ofR & D." The Economic Journal. 99, 569-596.
Cohen, w., and D. Levinthal. 1990. "Absorptive capacity: a new perspective on
learning and innovation." Administrative Science Quarterly, 28, 223-244.
Cohen, Wesley M. and Daniel A. Levinthal. 1990. "Absorptive Capacity: A New
Perspective on Learning and Innovation." Administrative Science Quarterly. 35,
128-152.
The Economist. 1993. October 30.
Duncan, Robert B. 1979. "What is the Right Organization Structure?" Organiza-
tional Dynamics, Winter, 59-79
Galunic, D Charles and Kathleen M.Eisenhardt. 1996. "The evolution of
intracorporate domains: Divisional charter losses in high-technology, multidivi-
sional corporations." Organization Science, 7(3), 255-282.
Greiner, Larry E. 1972. "Evolution and Revolution as Organizations Grow." Harvard
Business Review, 50.
Khandwalla, P. 1973. "Viable and Effective Organizational Designs for Firms."
Academy of Management Journal, 16,481-495.
Lawrence, Paul R., and Jay W. Lorsch. 1967. Organizational and Environment.
Boston: Harvard Business Press.
Levinthal, Daniel A. and James G. March. 1993. "The Myopia of Learning. " Strategic
Management Journal, 14, 95-112.
March, James G. 1991. "Exploration and Exploitation in Organizational Learning."
Organization Science, 2(1), 71-87.
March, James G. 1994. Three Lectures on Efficiency and Adaptiveness in Organiza-
tions. Helsingfors: Swedish School of Economics and Business Administration
Research Report 32.
Mintzberg, Henry. 1983. Structures in Fives. Englewood Cliffs, NJ: Prentice-Hall.
Perrow, Charles. 1967. "A Framework for the Comparative Analysis of Organiza-
tion." American Sociological Review, 32(2), 144-208.
Petroski, Henry. 1993. The Evolution of Useful Things, New York: Alfred A. Knopf,
288.
9.12 Summary 345

Petroski, Henry. 1985. To Engineer is Human: The Role of Failure in Successful

Design, New York: St. Martin's Press, 247.
Probst, Gilbert J. B., and Bettina S. T. Buchel. 1997. Organizational Learning: The
Competitive Advantage of the Future. London: Prentice-Hall Europe.
Simon, Herbert A. 1981. The Science of the Artificial. Cambridge, MA: The MIT
Press, 247.
Sitkin, Sim B. 1992. "Learning Through Failure: The Strategy of Small Losses."
Research in Organizational Behavior, 14, 231-266.
Weber, Max. 1946. From Max Weber: Essays in Sociology, translated and edited by
H.H. Gerth and C. Wright Mills, New York: Oxford University Press, 490.
CHAPTER 10

Using the Organizational Consultant

10.1 Introduction
In a recent report the Danish Ministry of Industrial Affairs stressed that
the introduction of new technology may even be harmful to the corporation,
if it is not accompanied by the appropriate organizational alignment. This
is a practical statement that "fit" is necessary for organizational success-
the main theme of this book.
In this chapter we summarize how the theory of strategic organizational
design is useful and discuss how Organizational Consultant (OrgCon) may
be used within the following areas:

• Organizational diagnosis and design

• Alignment of the understanding of the organization's situation
• Learning by doing

The Organizational Consultant systematically addresses complex issues,

which involve large number of variables and trade offs, as suggested in
Figure 1.3. The input information for the OrgCon depends upon the
particular organization and its situation. In this chapter, we outline how to
obtain appropriate input data. In the next few chapter we illustrate the
process with cases. The Organizational Consultant facilitates a low-cost way
to investigate numerous scenarios and enrich our understanding of practical
as well as theoretical issues. The Organizational Consultant software
program helps diagnose the current situation and the current organiza-
tional structure and evaluate different organizational designs. The OrgCon
aids managers, consultants, researchers, and teachers.
The installation and use of the program are described in the Installation
Guide and Help system on the accompanying CD-ROM. A complete
description of the theory and methodology behind the Organizational
Consultant was provided in chapters 1-9.
Basically, Organizational Consultant is an expert/knowledge base
system, which incorporates knowledge from a large body of literature on
organizational theory as well as knowledge derived from experts into an
information processing framework for organizational design. This know-
ledge base has been transformed into a system of decision rules. The
348 Chapter 10. Using the Organizational Consultant

development process for the "if-then" rules is presented in Chapter 1. The

content ofthe rules is presented in Chapters 3-9. Once a user describes his
or her current organizational situation, the Organizational Consultant
compares that situation with the knowledge base and notes any "misfits."
After analyzing different parts of the organization, Organizational Consul-
tant provides recommendations with explanations. The explanations relate
the recommendations and conclusions to the theory of organizational
design.
It is thus important to understand that the Organizational Consultant
is a decision support system which requires the individual to contribute
information to the decision making process. Organizational Consultant is
not a "black box," where data input are magically transformed into the best
organizational design for a given corporation. Moreover, decision makers
may choose to accept all or part of the Organizational Consultant's
conclusions based on their own understanding of the organization's
particular situation.

-
A
n

<
I
n
a WrittenRepor1
I
p -+ y
u

\
S Presentation
t
I
s

I I

Interpretation
01 facts
~
Interpretation
01 results

Figure lO.1.The Organizational Consultant is a Decision Support system

(DSS)

In applying Organizational Consultant, the individual plays an important

role in at least two separate stages. First, the individual must interpret the
facts of the current situation for input into Organizational Consultant.
Some facts are easy to decide as they are based on quantifiable information
10.2 Practical Organizational Diagnosis and Design 349

(e.g., the number of employees). Other facts are not so straightforward;

determining environmental complexity for instance or environmental
equivocality, may require an individual to judge the situation based on their
typical, day-to-day interactions within the corporation. Or, if a student is
trying to analyze an academic case scenario, he may have to make some
reasonable assumptions or "educated guesses."
Second, the individual must interpret the results and recommendations
of the Organizational Consultant. The Organizational Consultant makes
recommendations with assigned certainty factors. The higher the certainty
factor, the more likely it is that the recommendations apply to the organiza-
tional situation. However, as with any expert system, recommendations
need to be carefully assessed before being implemented.
Figure 10.1 shows a graphic representation of the role the individual
plays in using Organizational Consultant.

10.2 Practical Organizational Diagnosis

and Design
Consider these managerial situations:

• A manager in a firm who is considering restructuring her business unit

and does not know where to begin.
• A loan manager who is trying to determine whether to make a loan to a
small business that is planning a major expansion.
• A consultant who has responsibility for re-engineering several major
processes of an organization and wants to know whether these changes
will be consonant with the existing organizational design.
• A strategic planner in a company whose job is to identify and evaluate
potential take-over targets.
• A venture capitalist trying to determine the effects of replacing a start-
up company's founding management team with professional managers.
• Another strategic planner who is looking for exploitable weaknesses in
the organizational design or strategy of a competitor or a new market
entrant.

The information available for each situation is very different. The manager
who has to restructure her business unit usually has in depth information
about the functioning of that unit. Additionally, she will have access to
appropriate internal as well as external data. However, she is also herself
part of the unit that is being analyzed and her interpretations may thus be
biased.
350 Chapter 10. Using the Organizational Consultant

A consultant working for the corporation may get access to appropriate

internal and external information but has to set up a procedure to compile
these data. Appropriate data may be subjective in nature. Different people
in the organization may view the organizational situation quite differently.
Such differences may be attributed to differences in job, knowledge,
education, values, etc. The consultant has to choose the point of view from
which he should analyze the organization or facilitate a change process. It
may be a diagnosis from the point of view of the top management or the
owners. It may also be as a guide to the organization to help groups to
better understand their different views as to better group decisions. In
working with the organization the time dimension is also important. Is the
analysis for making short term corrections or is the analysis part of a long
term strategic adjustment? Different scenarios may be drawn. The
Organizational Consultant input dimensions make it easy to create profiles
for persons, situations or scenarios. The Organizational Consultant analysis
provides a ready tool to evaluate differences in the profiles and thus
evaluate the implications of the different scenarios.
Success or survival of an organization cannot be estimated only from
past performance: balance sheets, accounting information, and budgets. The
success depends on the organization's ability to adapt to the particular
situation that the organization faces. Information on leadership, manage-
ment, environment, strategy may be very useful for a loan manager or
venture capitalist who has to decide whether to grant a loan or invest in a
specific firm. However, such information is usually considered hard to get
and costly to obtain. A loan manager may get some information from the
specific organization, but most likely she will have to complement such data
with data from independent external sources. The Organization Consultant
provides a structured and low-cost query process
Competitor analysis can be developed from external sources(Chapter 11).
Merger and takeover analysis may be silent and quick as well.
Organizational Consultant can be used continually to perform a check-up
or audit ofthe organizational management and design as well as evaluate
major implications of new technology or new strategies. In the lifecycle view
of organizational development discussed in chapter 9, the organization has
to make major changes as it moves from one state to another. With the fast
pace in technological development and competition such changes may come
within short intervals. Organizational Consultant provides a means to
compare both the past with the present but also to evaluate the future,
allowing for learning both from success and failure.
Chapter 11 provides an analysis of the firm Medtronic based only on
publicly available data.
In Chapter 12, the Oticon case shows how to compile internal and
external data for analysis of a specific company.
10.4 What Can We Learn By Doing? 351

10.3 Alignment of the Understanding of

the Organization's Situation
In previous chapters the determination of the situational factors has
discussed. Many factors rely on the subjective view of a particular
individual. For example, different people in the organization may view the
competitive situation quite differently. This may be based on the way the
particular information about the environment is interpreted, or it may be
due to differences in the type or amount of information each individual
possess. One individual may follow the environment very closely and be up-
to-date with respect to new government regulations, the development of
new technologies, and the actions taken by the nearest competitors.
Another individual in the same company may view the competitive
situation differently; either because she interprets the information on the
environment differently or because she has very different information. The
management style will also effect how the management views the other
variables. One individual may see situational misfits, where the other
individual sees no situational misfits at all. Both are right from their
perspective and make decisions on that basis. But if they both have to
agree, some alignments are necessary.
Even individuals that work very closely together may view their
organization and the organizational situation very differently. One may
wonder if these people actually work in the same organization. Organiza-
tional Consultant and its theory of organizational design provide an easy
tool to accesses differences in views of the organizational situation and sort
out their consequences. For each individual an input profile is created. Such
profiles are easy to compare and thus enable the analysis to focus on the
important dimensions where differences are present.
Differences in views most often create disagreements and conflict.
Bringing a number of people from the same company together and using
Organizational Consultant provides an easy way to identify differences and
find a way to resolve conflicts. Most often conflicts are not only resolved, but
each individual has a better understanding of the organizational situation
and its implication for their organization.

10.4 What Can We Learn By Doing?

The Organizational Consultant provides a new way of learning about
complex organizational relationships. A manager, consultant, or a student
is not only provided with results but forced to be systematic. The explana-
352 Chapter 10. Using the Organizational Consultant

tions that follow any recommendation teach organization theory and design.
Additionally, you are asked to understand and work with general and
difficult concepts not only at the abstract, but also specifically related to the
cases and real situations. Teaching organizational theory and design in the
context of a specific industry using particular cases in that industry has
proven very helpful in providing a better understanding of both the abstract
and the specifics.
Academic case studies demonstrate both how to use Organizational Con-
sultant as well as what information is needed when performing an
organizational audit or checkup. They offer a glimpse of reality, since most
cases are drawn from the real business world. The cases can provide a
context for understanding some of the theoretical concepts and definitions
used in the Organizational Consultant program. Finally, case studies also
provide handy academic assignments for analysis. Real world projects can
be helpful to executives and academics alike.
While case studies can be very useful for understanding theoretical
concepts, it is important to realize that cases are not perfect representations
of reality. No matter how a case may resemble a real-world business
environment, there will inherently be gaps. Furthermore, ifthe Organiza-
tional Consultant program is part of an academic class, some gaps may
exist in your case study on purpose. The fact that a gap exists may be an
effective way of demonstrating the importance of one particular organiza-
tional design concept. Despite their potential drawbacks, case studies
remain an important part of how people learn and can provide strong
illustrations ofthe interrelationships between organizational design theory
and practice. Other readers may decide that the best way to learn about
Organizational Consultant is to use the system to analyze their current
organization. Ultimately, it is our hope that all readers will be able to apply
Organizational Consultant to their organizations and make positive
changes in their organizational designs.
To learn from case studies, the reader must adopt a critical perspective
while first reading the case. Most scholars suggest several readings of a
case. The first reading should simply skim for general ideas. A second
reading should be done with the goal oflooking for key pieces of information
for entry into Organizational Consultant. A final reading may be needed
prior to interpretation of the Organizational Consultant's results to help
test the feasibility of the proposed recommendations.
One analytical tool that may help during the second reading of a case is
a key issue template. Highlight the key information needed to run the
Organizational Consultant analysis.There may be areas which are
immediately clear and verifiable from the data supplied in the case; but
there may also be issues which are not so clear cut. In these instances, a
reader may decide to run the analysis several times with different values,
10.4 What Can We Learn By Doing? 353

or may choose the "Unknown/No Answer" option provided by Organizational

Consultant.
As you read the case or think about your organization, jot down your
thoughts about the current organization.
A description of the situation and the organization is generated by
answering the questions. The user focuses on some aspects of the organiza-
tion; ignores others. The first run with Organizational Consultant is only
a beginning for further exploration. Did the results make sense? Why did
these results obtain? What should be changed to obtain a better under-
standing? The sensitivity analysis facility serves these purposes well where
several scenarios are analyzed by changing systematically important
variables.
In searching through the results, new questions arise. Three issues are
important. First, the user should examine the values for the contingency
factors, i.e., a feedback of results to obtain a better description. Second,
variables may not only change in the certainty factor and the reasoning has
to be reviewed. Finally, recommendations that remain unchanged suggest
robustness of these recommendations.
It is important that the user employs Organizational Consultant as a
decision support tool. Organizational information has to be interpreted from
the basis for the answers to the questions. Additionally, the results from the
expert system are somewhat general and not phrased in relation to the
particular organization. The user must interpret the results and generate
specific recommendations. For example, if the expert system recommends
that decentralization should be high, the user must determine in more
detail which decisions to decentralize.
Thoughtful change (sensitivity analysis) and re-examination will
illustrate conceptual relations, tradeoff's and enhance an intuitive under-
standing. To summarize, doing a case using Organizational Consultant is
as follows:

• ENTER INPUT
Note issues on which you were not sure about the answer as possible
candidates for a sensitivity analysis.
• RUN SUB-ANALYSES (Size, Management style, etc.)
Carefully review. Change sub-analysis results if you do not agree. If
you change, note the change as a possible candidate for a sensitivity
analysis.
• RUN ORGANIZATIONAL ANALYSIS
Carefully review recommendations and explanations. Make sure that
you agree. Note the certainty factor to see how confident Organiza-
tional Consultant is. Carefully review the situational misfits. If there
354 Chapter 10. Using the Organizational Consultant

are misfits, can or should they be removed? If so, how? Note it as a

possible candidate for a sensitivity analysis.
• RUN SENSITIVITY ANALYSIS
Go through the previous steps for an appropriate number of changes.
Stop when you are satisfied with the results. Use the Delta Analysis
Template.
• REVIEW THE DETAILED RECOMMENDATIONS; INTERPRET
WHAT THE RECOMMENDATIONS MEAN FOR THE PARTICULAR
CASE
Evaluate if the recommendations are appropriate for the case. Make
the recommendation specific to the case. Use the vocabulary of the
case.
• REVIEW THE CHANGE PROCESS
Create a plan for implementing the changes.
• REPORT THE RECOMMENDED CHANGES IN ORAL OR WRITTEN
FORM

Table 10.1. Sensitivity Analysis or Delta (a) Analysis Template

Base Situation Variation

Strategy Change in one of the variables on the

Size/Ownership left hand side of the multi contin-
Technology gency model (Figure 1.3)
Environment
Management preferences
Climate
II II

Organizational Organizational
Recommendations Recommendations

Configuration Configuration
Properties Properties

10.5 Cases
In this chapter we have provided guidelines for using Organizational
Consultant in a number of different situations. In the following chapters
more specific discussions of these topics will be provided.
The Metronic case in Chapter 11 focuses on publicly available, external
input data for the Organizational Consultant. A methodology to gather and
10.6 Organizational Consultant Setup 355

interpret external data is developed in depth. The input file for the Med-
tronic cases is available from the input directory. The reader can thus verify
the case discussion and do sensitivity analysis on questions that you find
interesting to investigate.
Chapter 12 presents the Oticon Case, the Danish hearing aid producer,
which has introduced a new and highly discussed new organizational form-
the spaghetti organization.
Four additional cases-ABB Electromechanical Meters, Duke University
Press, GTE Government Systems Mobile Subscriber Equipment, and
Bluestone Group, Inc.-are included in Chapter 13.
Input files for Medtronic, Oticon and the four cases in chapter 13 are
included in the input directory. The complete discussion ofSAS is included
in the case directory. A complete discussion ofdata input and how to use the
OrgCon in the case Microlink can also be fund in the case directory.

10.6 Organizational Consultant Setup

The educational version of the Organizational Consultant knowledge base
system is in included with this book on the accompanying CD. The CD
contains the system, sample input files, case write-ups, and the Adobeco
reader.1 The Adobeco reader is needed to read the case files. Run "setup"
from the accompanying CD and the installation program will guide you
through the installation. The Adobe co reader is installed by clicking on the
Adobe exe file (16 or 32 bit versions).
Activate Organizational Consultant by clicking on Orgcon7.exe. The
userid and password for the educational version are: Userid: OrgCon,
Password: Ecomerc.
For further instructions click on the orgcon.hlp file and go to the topic
Contents. Check updates on www.busieco.ou.dklorgcon.

Notes
1.© Adobe Systems Incorporated
CHAPTER 11

Using Publicly-Available Information

for Strategic Organizational Diagno-
sis and Design
by Starling D. Hunter, III

11.1 Introduction
The purpose of this chapter is to assess the feasibility of using publicly-
available information, rather than inside or first-hand knowledge, for
strategic organizational diagnosis and design. The chapter provides a
detailed example of how business and popular press accounts on the leading
U.S. biomedical firm, Medtronic, were identified, collected, analyzed, and
transformed into inputs for the Organizational Consultant (OrgCon).
Section 11.2 of the chapter describes the data sources and the collection
and analysis methods. Section 11.3 provides a brief overview of Medtronic
Inc. Section 11.4 contains a detailed discussion of how answers to the
OrgCon questions were derived and includes extensive extracts from the
news articles and other information sources. Section 11.5 provides a
discussion ofthe OrgCon's recommendations for Medtronic with section 11.6
consisting of a more detailed discussion of the situational and design-
parameter misfits which the analysis uncovered.

11.2 Data Sources, Collection and

Analysis
11.2.1 Description of Data Sources
Sources of information on organizations are numerous and with current
advances in information and computing technology, new sources are being
created at a rapid pace (Orenstein, 1998). Sources can be free or commercial
(fee-based), topic-specific or comprehensive, cross-sectional or longitudinal,
single or multi-sourced, and have varying levels of user friendliness and
customization options (Orenstein, 1995).
358 Chapter 11. Using Publicly-Available Infonnation for Strategic
Organizational Diagnosis and Design

One challenge facing OrgCon-users investigating organizations about

which they have little or no direct knowledge is knowing what information
is useful, where to obtain it, how to collect it, and how to analyze it. In this
study four on-line full-text sources-Lexis-Nexis (Elsevier), ABIIInform
(University Microfilms Inc.), Infotrac (Information Access Company), and
Wall Street Journal On-Disc (Dow Jones)- and one world wide web (WWW)
search engine, Yahoo, were employed. In the table below, these sources and
databases are compared along several dimensions.

Table 11.1. Comparison of On-line, Full-text Sources

Lexis- ABII Infotrac Wall St. WW

Nexis Inform Journal (Yahoo)

Comprehensiveness Very High Very High High Moderate Low

Longitude High High Low Moderate Low

(dO yrs.) (>10 yrs.) «5yrs.) (5-10yrs) «2 yrs.)

Search Flexibility Very High High Moderate Moderate Very Low

Subscriptions Required Required Required Required None

Number of Sources Very High Very High High Very Low High

Abstracts! Excerpts Some High Moderate High Low

User Friendliness Moderate Low Moderate High High

Full-Text Available High Moderate Moderate High Low

Comprehensiveness: Full-text sources (FTS) may cover a broad or narrow

range of topics or subject matter. Lexis-Nexis (LN), Dialog, Infotrac, and
WWW all cover business, legal, regulatory, financial, international, political,
and cultural news. Infotrac and WSJ contain primarily business and
financial news.
Longitude: FTS vary on how many years of past coverage they contain.
LIN and Dialog typically exceed 10 years while Infotrac and WSJ are less
than eight in most instances. The WWW is by far the "youngest" with
coverage typically extending back only 2-4 years.
Search Algorithms: All FTS can be queried by some sort of search engine
or algorithm. The key considerations are the degree to which the users may
define their own search strings, how many characters they may contain,
and how "smart" they are (i.e., whether they employ Boolean logic). Dialog
and LN have very flexible and powerful search engines.
11.2 Data Sources, Collection and Analysis 359

Subscriptions: All of these FTS require access through paid subscrip-

tions. The world wide web (WWW) is the only exception.
Multiple Sources: FTS vary on the number of sources they contain.
Dialog and LN provide access to several thousand newspapers, journals,
newswire, magazines, etc., while WSJ has, obviously, only one.
Abstract Format: The results of search terms used in on-line databases
may appear with or without abstracts. Determining the relevance of an
article by reading its abstract rather than its full-text is often preferable.
Dialog, Infotrac, and WSJ typically provide abstracts of 50-150 words as a
standard feature. LN and WWW, on the other hand, provide only the article
title, word length, author name, and publication name and date.
User Friendliness: The databases vary in their ease ofuser-friendliness,
defined here as the ease of use of menus and interfaces, means of and effort
required for article retrieval, and accessibility of contents (how many
distinct sub-files have to be searched). In general, web-based browsers and
interfaces are much more user friendly than the DOS-based and PF-key
interfaces of ABI, WSJ, LN, and Infotrac.
Full-text availability: The proportion or articles available in full-text
differs across and within various FTS (Orenstein, 1995). LN and WSJ had
the highest proportion of articles available in this format.
Although articles were collected from all ofthe above full-text sources,
the WWW included, the remainder ofthis section describes only how Lexis-
Nexis ® was utilized. 2
The "All News" library in Lexis-Nexis® was queried with the following
search string:
Atleast3 (Medtronic) and date aft 12/31 /89 and date bef 1/1/96.
The "atleast3 (Medtronic)" indicates that only articles where the term
"Medtronic" appeared at least three (3) times would be selected. This
"atleastN" option can be adjusted by the user and, obviously, the higher "N"
is, the smaller the number of articles selected.
A total of 1363 citations were obtained for this search string, 511 for the
years 1990-1992 inclusive and 852 for the period of 1993-1995. Examples
of citations appear below.

• Star Tribune, August 27, 1992, Metro Edition, Marketplace; Pg. 2D,
1407 words, Medtronic is well aware ofthe long list of talent that it has
let slip away, Tony Carideo; Staff Writer.
• Business Wire, August 26, 1992, Wednesday, 648 words, Medtronic
releases "Blueprint for Health Care Reform", MINNEAPOLIS.
• Star Tribune, August 26, 1992, Metro Edition, Marketplace; Pg. 3D,
179 words, Medtronic pump's use with drug OK'd, Steve Gross; Staff
Writer.
360 Chapter 11. Using Publicly-Available Information for Strategic
Organi%IJtional Diagnosis and Design

• Business Wire, August 25, 1992, Tuesday, 715 words, Medtronic

releases new drug/pump combination for spasticity affiicting multiple
sclerosis and spinal injury victims, MINNEAPOLIS, MN.

The list of citations was reviewed and ninety-one (91) articles were
identified and retrieved for further investigation. These articles were chosen
because their headlines suggested that there was some information
contained which addressed at least one of the OrgCon questions. Articles
reporting strictly financial (e.g. quarterly earnings, dividend increases,
stock-splits, etc.) or purely scientific/technical information were also
excluded. Examples of citations of articles that were subjected to further
review are provided below:

• Corporate Report Minnesota, December, 1995, Vol. 26; No 12; Sec 1; pg.
64,767 words, The heart ofMedtronic-R&D, Eric J Wieffering, Minneap-
olis; MN; US; Midwest, 9610645, CRMN.
• Health Line, December 1, 1995, INSIDE THE INDUSTRY, 654 words,
MEDICAL TECHNOLOGY: COMPANIES DEPART U.S. TO ESCAPE
FDA.
• Business Wire, November 8, 1995, Wednesday, 504 words, FDA clears
new family of Medtronic pacemakers for United States marketing,
MINNEAPOLIS.

11.2.2 Content Analysis Approach

Each of the selected articles was submitted to an, iterative multi-step
content analysis. The first step entailed re-evaluating the "relevant"
articles. The lead paragraph of each article was read and the article was
coded as "Relevant" or "Non-relevant." The articles were reviewed in
sequential order, from 1990 through 1995. The most frequent reason for
coding an article as Non-relevant was because of redundancy, that is, it
reported information already contained in another source(s). Ten (10) of the
selected articles were eliminated in this first step.
The remaining eighty-one (81) articles were then read in their entirety
and pertinent passage (sentence, quotation, paragraph, figures, etc.) were
assigned to one or more of the following categories ofOrgCon questions: (1)
Current Configuration (2) Current Complexity (3) Current Formalization
(4) Current Centralization (5) Size (6) Age/Ownership (7) Product Diversity
(8) Technology (9) Environment (10) Management Preferences (11) Strategic
Factors and (12) Climate Factors. Many articles contained information on
more than one category. An example of a pertinent passage is given below:
11.2 Data Sources, Collection and Analysis 361

"(The) price (of medical equipment) is a secondary issue and there are
minimal price differentials among the major manufacturers. The key
purchasing issues are technology, service and reliability." (Rosenbaum,
1991).
This passage was assigned to the "Strategy" group, particularly to the
question which concerns price levels of an organization's products relative
to that of its competitors.
The last step in the content analysis was the transformation into the
appropriate format for use in the OrgCon. This was necessary because the
information required to answer many OrgCon questions rarely appears in
the format (i.e. measurement scale) that answers require. Consider the case
of several articles listing and describing the location and functions of
Medtronic's foreign and domestic operations and facilities. That information
was used to answer the three OrgCon Current Complexity questions that
appear below:

• Including the main center, how many geographic locations are there
where organization members are employed?
• What is the average distance of these outlying units from the organiza-
tion's main center?
• What proportion of the organization's total work force is located at these
outlying units?

The OrgCon uses the answers to these three questions to determine an

organization's spatial differentiation. The articles listed the locations and
employment levels ofMedtronic's major units and subsidiaries. Answering
the first ofthe three questions was simply a matter of counting the number
of distinct units and facilities listed.
The second question asks for more precise data on the location of the
Medtronic units. As noted above, the articles did indicate the city, state, and
country of each facility. The required transformation for answering the
second question involved using a geographic atlas to calculate the distance
in miles of each facility from the headquarters and then dividing by the
number of total facilities.
The third question also asks for very precise data, this time concerning
the percentage of employees working at the outlying locations. The relevant
information in the articles appeared as statements about employment levels
at certain locations, regions, or countries. The required transformation was
to determine the number of employees working at each non-headquarters
facility and then divide this amount by the total number of employees.
The final step in the content analysis was to assign a certainty factor (cf)
to each answer. Certainty factors are used in the OrgCon to indicate the
strength of belief concerning an input or a design recommendation. Where
362 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

answers were relatively clear and unambiguous, a certainty factor of 100

was used. Examples of this case included the current configuration which
is undoubtedly a matrix, and the size and age/ownership questions. The
lowest certainty factor used was 70 and this was reserved for questions that
required a great deal of inference or "best-guessing" on the part of the
investigator. Certainty factors of 85 were used for answers falling between
these two extremes.

11.3 Company Profile3

Medtronic is the world's largest medical technology and therapeutic device
company specializing in implantable and invasive therapies for improved
cardiovascular and neurological health. Primary products include implant-
able pacemaker systems, mechanical and tissue heart valves, perfusion
systems including blood oxygenators and centrifugal blood pumps, balloons
and guiding catheters used in angioplasty. Established in 1949, Medtronic
has since grown to become a Fortune 500 business.
Medtronic was founded in 1949 by Earl Bakken, then a graduate student
in electrical engineering and a part-time employee at Minneapolis's
Northwestern Hospital. Bakken's duties entailed the repair of centrifuges,
electrocardiograph machines and other electronic equipment at the hospital
where his wife was also employed. Bakken eventually quit graduate school
and with his brother-in-law formed a medical equipment repair service they
dubbed "Medtronic." In 1950 the partners contracted as sales representa-
tives for several large medical equipment manufacturers. As Bakken
increased his experience with and -exposure to medical professionals and the
equipment that they used, he was often asked to assist or advise in their
experiments. During its first 10 years in operation, Bakken built nearly 100
customized-sometimes single-use-devices for medical research and
experimentation. Becoming a full-fledged medical products manufacturer
in its own right was the logical next step.
Medtronic was soon manufacturing several medical research products
such as defibrillators, forceps, animal respirators, a cardiac rate monitor
and a physiologic stimulator. The manufacturing processes were nowhere
near current standards-"parts were handmade or surplus, quality control
was accomplished by visual examination, and products were packed in
newspaper and shipped in recycled boxes" (Kepos, 1994: 351).
Medtronic earned a reputation as a leaderin the bio-medical engineering
devices in the late 1950's when Medtronic's external pacemakers began
being used at prestigious medical institutions around the country and the
world. The close affinity between electronics and cardiac research continued
11.3 Company Profile 363

to advance open-heart surgery and the fortunes ofMedtronic for many years
to come.
The growing product line and increasing demand both enabled and
required that the company move to an expanded facility in the early 1960's.
Research on and introduction of several new products continued throughout
the 1960's and helped sales surge to over $12 million by the end of the 1967-
68 fiscal year. Medtronic established a European Service Center in 1967 to
provide technical support for a region that generated approximately 80% of
overseas sales. By 1970 the company had divided into four geographical
regions: Europe/AfricalMiddle East, Canada, Latin America, and Asia!
Pacific. Direct Sales offices were established in 19 countries.
The 1970's brought several leadership changes for Medtronic. Bakken
gave up day-to-day responsibilities of the presidency in 1974 to become the
company's chairman of the board. Two others held the post of presidency in
the remainder of the decade. This period also saw Medtronic make its first
acquisitions.
Computer technology helped Medtronic develop pacemakers that were
tailored to patient's requirements. The company's developments in the
1980's and early 1990's included cardiovascular and cardiopulmonary
products, neurological and drug delivery devices. With a number of
acquisitions and new product introductions in the late 1980's and early
1990's, Medtronic reached $1 Billion is sales in 1991.
The successes in the 1980's came with some inevitable setbacks. A
government review of pacemaker implants culminated in Senate hearings
that revealed a pattern of overuse of pacemakers in the state of Maryland.
In 1983 the company voluntarily recalled a new pacemaker lead when it
was discovered that it had an unacceptably high failure rate in clinical
tests. Medtronic also became embroiled in expensive patent litigation with
large medical manufacturers like Eli Lilly and Seimens AG during this
period.
By the early 1990's, Medtronic had developed six primary areas of
expertise: bradycardia pacing tachyarrhythmia management, cardiopul-
monary, heart valves, interventional vascular, and neurological. In 1992,
the company's international sales contributed 40% of total revenues. This
was partly responsible for the development of new facilities and expanded
operations in Japan, China, and Eastern Europe. Currently headquartered
in Minneapolis, Minnesota, Medtronic now does business in more than 120
countries, with operations organized into three global areas: Americas,
EuropelMiddle East/Africa, and AsialPacific. There are four businesses
(divisions)-Pacing, Vascular, Cardiac Surgery, Neurological, and Developing
Business & Ventures-encompassing 14 strategic business units and nine
ventures. 4 The company had almost 13,000 employees and $2.2 Billion in
sales in the 1996 fiscal year.
364 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

11.4 Discussion of Answers to OrgCon

Questions
This section provides detailed rationales for each of the twelve categories of
OrgCon Questions: Environment, Strategy, Technology, Product Diversity,
Climate, Management Preferences, Size, Age/Ownership, Current
Configuration, Current Complexity, Current Formalization, and Current
Centralization.

11.4.1 Environment
Environmental Complexity: "Medtronic's environment is complex (cf 100)."
There is a great deal of evidence suggesting that Medtronic faces a highly
complex environment. Environmental complexity was defined in Chapter
5 as the number of environmental variables and their interdependence. Low
organizational complexity indicates that only a few variables describe the
environment while high complexity indicates that the environment has
many important variables to consider. Daft's (1995) description of the
environmental sectors was used to determine indicate the large number and
variety of variables with which Medtronic must contend. Table 11.2 rates
various environmental sectors on the basis of their complexity, uncertainty,
equivocality, and hostility.

Table 11.2. Characteristics of Environmental Sectors

SECTOR Complexity Uncertainty Equivocality Hostility

Industry High High High High
Raw Materials Low Low Low Low
Human Resources Moderate Low Low Low
Financial Resources 5 Low? Low? Low? Low?
Technology High High High High
Market High High High High
Government High High Very High High
Socio-cultural Moderate Low Low Low
International High High High High
Economic Conditions Low Low Low Low

The most complex sectors in Medtronic's environment are probably the

Governmental, Technological, and International.
11.4 Discussion ofAnswers to OrgCon Questions 365

Environmental Uncertainty: "Medtronic's environment has a high level

of uncertainty (cf 100)." In this context, "uncertainty" refers to the degree
of knowledge the organization has about the level or "value" of environmen-
tal variables that are known to exist (Burton & Obel, 1995: 152-3).
Medtronic has high levels uncertainty concerning many variables in its
environment. In the market sector high uncertainty is associated with the
determination of the rate and timing of new product introductions, which
product lines to expand or cut-back, and which markets to enter or exit.
In the government sector, uncertainty stems from the timing and impact
of proposed changes to FDA approval processes (Hamburger, 1994; Health
Line today, 1995) and the success of the firm's lobbying efforts in Congress
(Collins, 1995).
In the technology sector, there is high uncertainty related to whether
existing patents can be protected domestically and abroad (Feyder, 1992;
Health Industry Today, 1991), which technological advances to incorporate
into new product designs and how quickly, technological developments in
related (e.g., bio-medical, pharmaceutical) and unrelated (e.g., micro-
electronics, information technology) fields that may improve or obviate
existing products or that may create demand for new ones (Rosenbaum,
1991).
Environmental Equivocality: "Medtronic's environment has high
equivocality (cf100)." The environment ofMedtronic also has high levels of
equivocality. While it exists to some degree in every environmental sector,
it was perhaps highest in the Government sector for the six-year period
under study. The major source of equivocality for Medtronic, and for the
health care industry as a whole, was the Clinton Health Care Reform
proposal which failed to gain Congressional approval in 1994.
While many different versions of this plan were discussed during the
1992 Presidential campaign and during Clinton's first year in office, the
centerpiece of the proposed legislation was to provide universal access to
healthcare (Riley, 1993). The effects of adding up to 30 million previously-
uninsured Americans to the healthcare system, of limiting the options and
coverage levels of as many more, and of possible changes to reimbursement
schedules were difficult, if not impossible, to anticipate. Medtronic, like
many other firms in the healthcare industry, opposed this initiative,
advocating private sector solutions instead.
Environmental Hostility: "Medtronic's environment is hostile (cf 100)."
Sakson (1995) described the market for defribillators as a "hotly competi-
tive" area. Nissen (1992b) labeled as "hardball tactics", the highly-
publicized raid of a Minnesota-based biomedical company called Bio-Plasty,
Inc. undertaken at the behest of FDA Commissioner David Kessler.
Hamburger & Meyers (1994) quotes the CEO of Minntech Corp. of
Minneapolis as saying that "the attitude of the FDA is (so) anti-business
366 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

and negative that any prudent CEO (in the medical equipment industry)
has to be looking outside the U.S. for a safe haven."
Other forms ofhostility are described by Borger (1993) who suggests how
patent infringement lawsuits are used as a "strategic weapon" by some
companies and Pollack (1994) who details Medtronic's claims of "discrimina-
tion" against its products by the Japanese health insurance system. Meyers
(1995) tells of efforts by Medtronic and other firms who have had to develop
elaborate and sophisticated computer security measures to repulse attacks
from computer "hackers."

11.4.2 Strategy
Capital requirements: "Medtronic operates in an industry with a medium
capital requirement (cf 85)." According to Rosenbaum (1991) there are a
number of small, niche-focused companies in the biomedical industry. Some
of them are start-ups and others have grown steadily despite offering a
limited range of products and services. Medtronic competes with (Health
Industry Today, 1994), has acquired (Nissen, 1992) and has entered into
strategic alliances (Health Industry Today, 1995a) with many firms that are
much smaller than itself. Some of these firms, particularly the strategic
alliance partners, are making products which are comparable or even
superior to those offered by Medtronic. These facts suggest that scale-large
size and heavy R&D spending-may be important factors in the industry but
that smaller, less resource-rich organizations can enter and prosper along
side their larger rivals.
That said, major players in the industry such as Medtronic, St. Jude, and
Siemens spend tens or even hundreds of millions of dollars on R&D each
year, and the amounts show no sign of decreasing. Considering the
following statistics for the R&D spending of Medtronic from 1986 to 1995.
There are a number offactors in the industry that act to increase capital
requirements. Virtually every one, if not all, of Medtronic's products
requires the approval of government regulators in the country where the
product is sold and or manufactured. The principal regulatory body in the
U.S. is the Food and Drug Administration or FDA and its approval process
is widely recognized as being lengthy (Meyers, 1992), bureaucratic (Health
Line Today, Dec 1995) and slow (Riley, 1993). Additionally, recalls of
defective products (Rosenbaum, 1991; Colman, 1995), lengthy clinical trials
(Business Wire, 1995a) patent infringement fights (Feyder, 1992; Gross,
1992; Health Industry Today, July 1991), and product liability claims
(Marsh, 1995) are extremely_common and costly.
11.4 Discussion ofAnswers to OrgCon Questions 367

2ro .-------------------~--__. 12

200 11
10
...
/I)

.!!
1ro
~ R&D spending
9
'0 (rrillions)
o 100
8 _ % of revenues

0 +---+--+---+--+--+--+--+--+--+ 6
r.ofo ,,~~ ,,~r.ofb ,,~r.o~ ,,~~~ ,,~~" ,,~~~ ,,~~~ ,,~~ ,,~~~
,,~

Year

Figure 11.1. R&D Spending at Medtronic: 1986-1995

Source:AJexander(1995)

Furthermore, the rapid pace of change in the industry, particularly the rate
of new product introductions, has increased dramatically in the past decade
(Gross, 1991). In 1992, Medtronic's stated goal was to have the shortest
time-to-market in the industry, a goal that would require a reduction ofits
pacing products development cycle from 18 to 12 months (Machine Design,
1992). The design cycle time of one ofits pacing products has been reduced
by 50% in the last decade and Medtronic has managed to turn out four
generations of its implantable defibrillator in two years in its efforts to stay
ahead of its major rivals (AJexander, 1995b).
What one might infer from the all of above is that while barriers and
initial capital requirements may be low in this industry-a start-up with a
good product can be successful-the need for heavy R&D spending, the
lengthy and uncertain product approval process (Hamburger and Meyers,
1994; Solberg, 1994; Sakson, 1995) and the rapid pace of innovation
(Borger, 1993) create conditions which are likely to push capital require-
ments upwards over time.
Product innovation: "Medtronic has a high product innovation (cf 100)."
Medtronic thrives on being an innovator (Business Wire, February 1993;
Carideo, 1992; Ray, 1993; George, 1995; Fisher, 1993). It has pioneered
368 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

many technologies or product lines, particularly in its flagship pacing pro-

ducts. Medtronic introduced the first pacemaker in 1957 and was a leader
in the introduction of rate-responsive pacing, as well (Rosenbaum, 1991).
The company has adopted a policy similar to that ofIntel (Grove, 1996) and
other hypercompetitors (D'Aveni, 1995) of obsoleting its own products at a
rapid pace (Alexander, 1995b). Says Glen Nelson, Medtronic Vice Chair-
man: "We have to be prepared to essentially antiquate any of our products
with our own developments. If we can produce substantial improvements in
a product's performance, then anyone could." (Nelson, 1992)
Discussing the reason for Medtronic's unrelenting pace of new product
introductions, Dennis Salke, former Vice-President and General Manager
of Medtronic's heart valve division said: "For a great big company, they get
up scared every morning and that's part of their strength. They've got people
in the company who saw market share decline in the 1970s and early 1980s.
They felt the trauma and know why it occurred. They weren't ahead in
technology. A lot ofpeople at Medtronic have vowed its never going to happen
again. The ante to the game in the (bio}medical field is to be there first with
a quality product. You can't get left behind." (Alexander, 1995b).
These same sentiments were echoed by Art Collins, Chief Operating
Officer at Medtronic, during his testimony to a Senate subcommittee
hearing on FDA control of the export of unapproved medical devices
(Collins, 1995). He exclaimed: "The key to the global leadership ofMedtronic
and the U.S. medical device industry in general can be summarized in one
word-innovation. Our strength lies in our ability to develop innovative
therapies that meet patient needs, and to deliver those therapies to the
market. Today, medical technology advancements are happening at
lightning speed compared with the pace just five or ten years ago. And this
innovation is occurring throughout the world. To remain competitive,
Medtronic and other manufacturers must conduct product development and
manufacturing activities in an environment that fosters innovation. If such
an environment cannot be found in the United States, we have no choice but
to look elsewhere."
Process innovation: "Medtronic has high process innovation (cf 100)."
Medtronic has as strong concern for process innovation-how it makesjts
products-as it does for product innovation-what it makes. Some ofthe ways
in which this has been accomplished are through the following means:

• The use of computer-aided design, manufacturing and engineering

systems such as ProIENGINEER, which resulted in an 80% reduction in
the time required to create and model a complex pacemaker and a 90%
reduction in the time required to produce product design variations
(Zimmerman & Drake, 1995).
11.4 Discussion ofAnswers to Orgeon Questions 369

• The use of a Cray supercomputer for product development and design-an

industry first-as described in the section 11.4.3.
• The implementation of process improvement, quality and just-in-time
(JIT) programs designed to eliminate product defects, cut costs, and
improve responsiveness to market demand (Wise, 1990, Business Wire,
Feb 1995b).
• The upgrading of marketing processes such as technical support,
customer education initiatives and product line expansion and bundling
(PR Newswire, 1995a).

A primary motive for both product and process innovation includes the
lowering of manufacturing costs (George, 1995).
Concern for quality: "Medtronic has a high concern for quality (cf 100)."
Medtronic has an extremely high concern for quality. This concern is rooted
in the company value system and instantiated in the numerous quality
initiatives that have been and are still being undertaken at the company
(George and McKibbin,1995). The attention to quality can be attributed to
a combination of internal and external factors. One of the external factors
deals with the FDA approval process. In the biomedical industry, quality
and safety are very closely related and the FDA's traditional role is to
ensure that only safe, high quality products reach the American consUmer
(Alexander, 1994). Low quality can result in otherwise useful products
failing to be approved, a lengthening or delaying the approval process, and
in the worst case scenario, recall of products already on the market
(Sakson,1995; Hamburger, 1994). CEO William George has been quoted as
saying that the most risky thing his company could do would be to put
unsafe (i.e. low quality) products on the market (George and McKibbin,
1995).
Within Medtronic there has long been a recognition that its products are
life-sustaining (George and McKibbin,1995; Collins 1995). Morale-boosting
visits from people who were helped or even whose lives were saved by
Medtronic products are a frequent occurrence (Slovut,1994; Inskip, 1992).
At Medtronic, concern for quality has been achieved through its tight
coupling not only to the value system (George, 1993; Lee- & Zemke,1993),
but also to on going efforts to achieve cycle-time reductions and process
improvements. The company has kept product quality as a major concern
through many process re-designs in such areas as order fulfillment,
manufacturing, and new product introduction and development (Alexander,
1995a). This has led to the adoption of a "do it right the first time" mindset
according to Chief Operating Officer Arthur Collins. He continues, "I think
cycle time reduction, if properly managed, enhances overall quality of the
product as measured by the customer" (Alexander, 1995b).
370 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

Price level: "Medtronic's price level is moderate relative to its competitors

(cf 100)." (Bio)medical technology is expensive. In 1992, Medtronic's smart
implantable defibrillators sold for about $20,000 each while "standard"
pacemakers were priced from $1,500 to $9000 (Gianturco,1992). While, the
prices do decline as the products move through their life-cycle, prices have
increased as the trend toward miniaturization has picked up (Sakson, 1995;
Business Wire, 1992). Medtronic's Jewel brand of miniature defibrillators
was selling for over $53,000 when first introduced in early 1995.
Although the products are expensive, price has not traditionally been of
great importance. Price differences among major competitors has been
minimal and the key issues have revolved around technology, after sales
service, reliability and the total cost of care (Rosenbaum,1991; Sakson
1995). Price of products is rarely, if ever, mentioned as a factor in discus-
sions about Medtronic's products. Thus, we are left to conclude that
Medtronic's price level compared to that of its competitors is moderate-that
is to say, neither significantly higher or lower.

11.4.3 Technology
Major activity: "Medtronic's major activity is categorized as production (cf
100)." The production technology is unit (cf70). Medtronic describes itself
as "the world's largest therapeutic device company, manufacturing
biomedical devices for improved cardiovascular and neurological health"
and as the world's leading medical technology company specializing in
implantable and invasive therapies."6
The list of options provided by the OrgCon for describing a company's
major activity are (1) production (2) service (3) retail and (4) wholesale and
Medtronic was assigned to the first category. Several articles attest to the
fact that Medtronic is not merely a retailer or wholesaler of biomedical
products, but is a producer of these goods (Wiefi'ering,1995; Borger,1993;
Engen,1993). It conducts its own research and development, designs, and
manufacturers most ofthe products that it sells (Meyers,1992).
Classifying Medtronic's production technology is not an easy task. We
don't know how many pace-makers-Medtronic's best-selling product and
oldest product line-the company manufactures and sells in a given year.
The larger that number, the more likely it is that the production technology
is a mass production-large batch or mass-produced technology. Further-
more, it is unlikely that Medtronic uses any process production technolo-
gies-controlled, standardized and continuous processing. Unit produc-
tion-where units are custom made and work is non-routine-seems to
provide the best description.
11.4 Discussion ofAnswers to Orgeon Questions 371

What would perhaps be better is to describe its production technology as

mass customization-a production process made possible computer-
integrated manufacturing (CIM) technologies such as robots, numerically
controlled machine tools, computerized software for product design,
engineering analysis, and remote control of machinery (Daft, 1998;
Meredith 1987; Adler, 1988). In the manufacturing systems studied by
Woodward (1965) batch manufacturing (large or small) was used to achieve
customization. Continuous processing was employed to produce a single
standard product in very large, if not unlimited, quantities. The advent of
CIM enables the best of both types of production-unlimited batch sizes with
high levels of customization-and has been shown to decrease scrap rates
and increase machine utilization, product variety, and customer satisfaction
(Meredith, 1987). Given that mass customization is not an option, we have
chosen to select unit production and give it a low certainty factor.
Routine technology: "Medtronic does not have a routine technology (cf
100)." There are several indications that Medtronic is not utilizing a routine
technology. Among them are: the increasing sophistication, variety, and
number of product offerings (Business Wire, 1995b; Alexander, 1995a), its
historical emphasis on path breaking R&D (Fisher, 1993; Zimmerman and
Drake, 1995), and its numerous and ongoing process re-designs and quality
initiatives (Zimmerman & Drake, 1995; Rosenbaum, 1991). Also to be
considered are its passion for acquiring leading-edge companies-particularly
innovative start-ups (Couretas,1994; PR Newswire, 1995b; Business Wire,
1995b) and its increasing participation in joint ventures and strategic
alliances (Health Industry Today, Sept. 1995a; Business Wire, 1994; UPI,
1991).
Divisibility: "Medtronic's technology is a somewhat divisible (cf 100)."
Divisibility is the degree to which tasks can be divided into smaller,
relatively independent tasks and is an important consideration in the
determination of appropriate organization structure. Technologies may be
divisible along a number of dimensions such as functional area (i.e. R&D,
production, marketing, etc.), product lines, or geographical regions.
The determination that Medtronic's technology is somewhat divisible is
based on the recognition that while some of its major product lines share
core technologies and processes, the differences are among them are
significant. Pacing products, which include Medtronic's flagship pacemakers
and its newer tachyarrhythmia (rapid hear rate) management devices,
employ integrated circuit (lC) technology. Neurological products include
tremor control devices which also rely on IC technology. The Cardiac
Surgery products, however, rely on a very different set of technologies,
principally on bio-compatible materials, bio-resistant fibers, mechanical
pumps, and tissue valves. These products appear to share a few characteris-
tics with the Vascular products such as catheters, stents, and balloon
372 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

angioplasty devices, none of which are particularly electronidelectrical or

mechanical in design. It does not appear that these technologies are
divisible across functional areas within the organization but they do seem
to be highly divisible across international boundaries.
Processes and technologies which are common to and divisible across all
four product lines are those related to computer-aided design and manufac-
ture (CAD/CAM), product data and information systems which support the
design and development processes (Zimmerman & Drake, 1995) and other
information technologies such as hardware and software (Health Industry
Today, 1995).
Technology Dominance: "The technology dominance is weak (cf 100)."
Medtronic does not have a dominant technology. The company employs
routine technologies in the production and manufacturing of its products
and non-routine technologies in much of its R&D (Alexander, 1995a;
Goldwasser & Moreau,1994). The marketing, sales, and distribution of the
products employ both technologies in varying proportions. According to
Wieffering (1995), Medtronic's "core competence" (Prahalad & Hamel, 1990)
is its R&D.
Advanced information system: "Medtronic has several advanced informa-
tion systems (cf 100)." Managerial information systems are employed to
gather, process, analyze and distribute information necessary to operate an
organization (Laudon & Laudon, 1996; Lucas, 1997). Daft and colleagues
(Daft, 1995, Daft & Macintosh, 1978) have argued that different organiza-
tional technologies require different information systems. In other words,
information systems should be compatible with the information processing
requirements ofthe technology or task that it is intended to support. Thus,
routine technologies, which rely heavily upon explicit tasks and procedures,
would have different information systems (IS) needs than would a non-
routine technology which is likely to be characterized by high task variety
and information richness. The former might have a greater need for
transaction processing systems and the latter a need for executive
information or decision support systems.
Medtronic employs many advanced information systems across many
levels of the organizational hierarchy, product lines, and functional
departments. Examples include:

• A product data management system for the management of design and

manufacturing information that allows users access to current and
archived design documents throughout the entire company. It also
ensures that only pre-approved designs are used in the manufacturing
process, provides electronic appending of all relevant technical docu-
11.4 Discussion ofAnswers to OrgCon Questions 373

ments to designs, and maintains an audit trail of all design changes

(Zimmerman & Drake, 1995).
• A Cray J916 supercomputer for its Center for Biomedical Research
which is used to perform "complex computational fluid dynamics and
structural analysis simulations to develop new cardiovascular medical
devices (Health Industry Today, 1995c). This is reported to be both
Medtronic's first supercomputer and the first Cray system in the
biomedical industry to be used in medical device development and
manufacturing (PR Newswire, 1995a).
• The ProlENGINEER CAD/CAE system installed on a network of Sun
Microsystems workstations which links all Medtronic design depart-
ments in the Minneapolis headquarters to the company's Netherlands
facilities, thereby permitting concurrent engineering. The system is used
to support Medtronic's automated design-to-manufacturing process for
pacemakers (Machine Design, 1992; Zimmerman and Drake, 1995).
• A relational database with advanced graphical software that allows its
technical support specialists to track more than 1 million possible
pacemaker and connector combinations (Verity, 1992).

Medtronic is also the lead-user of a $700 million audio-teleconferencing

center established by AT&T in Minneapolis (Alexander, 1993).

11.4.4 Product Diversity

Different products: "Medtronic has many different products (cf 100)."
Medtronic is the world's leading medical technology company specializing
in implantable and invasive therapies. Primary products include implant-
able pacemaker systems used for the treatment ofbradycardia, implantable
tachyarrhythmia management devices, mechanical and tissue heart valves.
Other products include balloon and guiding catheters used in angioplasty,
stents, implantable neurostimulation and drug delivery systems, and
perfusion systems including blood oxygenators, centrifugal blood pumps,
cannula products, and autotransfusion and blood monitoring systems. 7
According to David (1996), product diversity "was not always
(Medtronic's) mantra." In 1985, 85% of Medtronic's sales came from
pacemakers. By 1994, that figure had been reduced to an estimated 57% of
sales and a comparable amount of earnings. This was due primarily to "19
acquisitions in the intervening years, coupled with relentless R&D
spending, which have brought the firm into new, higher-growth areas such
as neurological devices and blood pumps" (David, 1996: p.23).
Different markets: "Medtronic has many different markets (cf 100)." The
1996 Medtronic annual report states that" our vision for Medtronic is to be
374 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

the world's leading medical technology company specializing in implantable

and invasive therapies. "S While the global medical community comprises the
primary-if not sole-market for Medtronic's products, this market clearly
contains many segments or niches but this is not the same as multiple
distinct markets and differs markedly by country and region.
Eifferent countries: "Medtronic operates in more than one country (cf
100)." Medtronic does business in more than 120 countries, with operations
organized into three global areas: Americas, EuropelMiddle East/Africa, and
AsialPacific. 9
Activity level abroad: "The activity level abroad is over 25%." The United
States is Medtronic's largest market, accounting for about 56 percent of
total revenues. 10 Western Europe is second to North America as Medtronic's
largest market and company officials believe that when combined with the
Middle East and Mrica, this region could eventually become the world's
largest market for medical devices.
These figures seem to indicate that more than 25 % of the sales occur in
overseas markets. But this is not the same as "activity" which would also
include activities such as production, research & development, distribution,
marketing, etc. Given the trajectory defined by its globalization strategy, it
is certain that if Medtronic is not yet conducting 25% of its activity overseas
it is very near this level and will soon exceed it. The table below lends
support to this contention:

Table 11.3 A Selected List of Medtronic Locations around the World

Facility Type & Function Number Location

Sales Offices & Distributorships 24 Europe Middle East South Mrica
Manufacturing Sites 5 Netherlands, France, Germany
Research Center 1 Netherlands
Education Centers 5 Netherlands, UK, Germany; Spain
European HQ, R&D, Manufacturing 5 Switzerland
Customer Education Center 1 Venezuela
Manufacturing Facility 1 Canada
Customer Education Center 1 Beirut, Lebanon
South Mrica Institute of Cardiovascu- 1 Johannesburg, SA
lar Medicine
Pacemaker Manufacturing Plant 1 Shanghai China
New R&D & Customer Education 1 Japan
Centers

Source: Medtronic Homepage, http://www.medtronic.comlannuall1etter.html

Different products in the foreign market: "Medtronic has many different

products in its foreign markets (cf 100)." Medtronic makes different
11.4 Discussion ofAnswers to OrgCon Questions 375

products as well as different versions of "domestic" products for its foreign

markets. Furthermore, the company conducts R&D (Moffat & Mendes,
1991) and customer education (Koska,1991) in and for each of its major
markets.

11.4.5 Climate
Level of trust: "The level of trust is high (cf 100)." The level of trust at
Medtronic is considered to be high although it is unclear as to how much
this varies across and within different levels, departments, geographical
regions, etc. Speaking to this issue, CEO William George gives an example
of how trust operates within Medtronic, its relationship to the company's
values, and how violation of trust is handled: "We were tested at Medtronic
by a subsidiary we had in Europe. The former president ofMedtronic Europe
challenged the values statement. He said, 'You don't like these promotional
funds we have.' And we said, 'No, we don't.' His response was, 'You're trying
to impose US. values on Europeans, and they operate differently.' We
reminded him, 'No, these are not US. values. They are Medtronic values. We
know the culture is different in Europe, but you still have to operate from the
same values and set of standards. You violated what we consider to be a
sacred trust.'
This person had to leave the company because he didn't get it. I know that
sounds a little brutal, but you have to be prepared to take that kind of step.
If you violate the trust, you can't be part of the organization. You can make
mistakes; everyone makes lots ofmistakes. But you can't sacrifice the values. "
Conflict: "The level of conflict is low (cf 100)." None of the articles
reviewed indicated that there was any appreciable degree of conflict-
opposition of goals, forces and beliefs-within Medtronic, either between
major departments or functions, between management and labor, or
between top company officials.
One key distinction is between the negative or "destructive" form of
conflict-characterized by competitive, inter-functional or interdepartmental
rivalries-and the "constructive" or beneficial variety (Coser 1956). Conflict
of both varieties seems to be negligible or non-existent at Medtronic. This
is consistent with the stroI).g influence of the charismatic founder, Earl
Bakken, and the unambiguous value and mission statements referred to
previously.
Morale: "The employee morale is high (cflOO)." The morale at Medtronic
is very high by all accounts. One way in which this was achieved and
maintained to this day is through visits by people who were helped by
Medtronic products. Slovut (1994) describes the visit coma to a Medtronic
plant where its Bio-Pump® heart and lung machine is produced by a three-
376 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

year-old hypothermia patient, her parents, older siblings, and the doctor
who revived her. The visit is described as "the kind of morale-building visit
that the company has scheduled annually since the 1970's ... letting
employees meet the patients whose lives were saved or improved by the
medical devices they produced."
Reward Equity: "Rewards are given in a highly equitable fashion (cf85)."
One compensation problem that has plagued Medtronic over the years is
that related to its handling of "intrapreneurs" Carideo (1992) describes how
Medtronic's inability to balance its efforts to nurture new ideas and
products with the legitimate aspirations and compensation demands of
product developers resulted in an exodus of top talent in the 1970's and 80's.
CEO William George is quoted as saying: "I don't think we managed the
process as well as we could have. During that period. Medtronic had
problems with sporadic R&D funding, the lack of 'real-world' incentives to
keep new ideas in-house, and not enough milestones with which to measure
progress of research projects. "
Underscoring the importance of this issue to Medtronic's future, George
also stated that getting top talent to "stay put" is the only way for the
company to "re-invent itself' over the coming decade. The company is, thus,
attempting to "create an atmosphere in which would-be entrepreneurs can
pursue those projects inside Medtronic."
To tackle this problem, Medtronic has created the "Venture Management
Group" under the direction of Vice-Chairman Glen Nelson. Clearer
benchmarks of progress, such as the submission of required FDA applica-
tions, are being established, as are reward systems that are more similar
to what designers would receive in the "real world."
Resistance to Change: "The resistance to change is medium (cf 100)."
Slow responsiveness is suggestive of high levels of inertia and resistance to
change (Hannan and Freeman, 1984; Gresov, Haveman, Oliva, 1993). The
prior discussion on rewards might also be taken as evidence that portions
ofMedtronic are slow or resistant to change. While many factors could have
contributed to the "brain drain" that Medtronic experienced, it is unclear as
to why it took the company so long to recognize the problem or to formulate
an effective response to it.
Leadership Credibility: "Leader credibility is high (cf 100)." The
leadership credibility-respect, inspiration, acceptance, etc.-is extremely
high. There are many reasons for this and the one mentioned most
frequently is the influence of Medtronic founder Earl Bakken. Since 1976,
Bakken has traveled the world meeting new employees. In that time, he has
met almost every new employee in the company. "Generally he meets them
in small groups in sessions that last about 90 minutes. At the meetings he
tells the new employees the history ofthe company and discusses Medtronic 's
11.4 Discussion ofAnswers to OrgCon Questions 377

mission, which is printed in eight languages on the inside of the company's

annual report. Then he distributes a card with the mission statement on it
and a medallion with an abbreviated version." (Carideo, 1992).
Scapegoating: "The level of scapegoating is low (cf 85)." The level of
scapegoating-shifting of responsibility at Medtronic is low. This can be seen
quite clearly in the remarks made in an address by CEO William George to
an audience at the Liberal Arts and the Law conference in 1992 (Storm,
1992). Commenting on the need for companies to take responsibility for
their actions and mistakes, such as defective products, he said: "You're
always better off going public, telling the truth, taking the hit and getting
it behind you," he said.
He continued stating that while all companies have problems and have
made mistakes. "Bad products", he said, "should be taken off the market,
fixed and returned to the market." Too much ofthe top management's time
is "often wasted handling problems because a company won't face up to
them." He said management's time could be better spent fixing the
problems.
George's earlier statements to the effect that "everyone makes mistakes"
is suggestive of culture which does not seek to punish or single out
employees for failures. He tells in an interview of his personal experience
with an organization where employees felt that compliance statements were
cynical attempts by upper management to protect themselves from the
consequences of their own actions. "I worked for a company once where... no
one believed them (standards of values). People felt that the compliance
statements were there to protect top management legally-to get themselves
off the hook, " said George (George and McKibbin, 1995).

11.4.6 Management Preferences

Kinds of Decisions: "Top management prefers to make policy and general
decisions (cf 100)." CEO William George makes it clear that the top
management prefers to delegate operational decisions. "... (W)e are able to
decentralize and delegate quality and ethical decisions to people because
there is a framework from which they can make good decisions. We say, 'You
don't have to make a trade-off between cost and taking care of the patient.
You always take care of the patient, and the cost will take care of itself.' That
makes an easy framework for people." (George and McKibbin, 1995).
Decision Time Horizon: "Top management primarily prefers to make
both long-term and short-term decisions (cf 100)." The top management
prefers to make long-term decisions but because of the "rapidly changing
environment" it is being required to make many more short term decisions
than it would ordinarily. This tendency toward more short-term decisions
378 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

has been spurred by changes in the regulatory environment, shifts in the

purchasing power from doctors to administrators and from individual
hospitals to purchasing cooperatives, the proliferation and increasing
sophistication and influence of supplier groups, and the rapid reduction in
the number of channels for and the "commodification" of Medtronic's
flagship product-pacemakers (Weiffering, 1995).
Level of Detail: "Top management has a preference for aggregate
information when making decisions (cf 85)." While Medtronic officials
frequently discuss their policy of decentralizing decisions, they say much
less about the level of detail it prefers to use when making decisions. The
lack of discussion on the use of information systems and technologies (e.g.
decision support or executive information systems, databases, or data
warehouses) which might support, improve, or enable highly-detailed
decision-making might indicate that there is either a low or medium
preference for detailed information by top management.
Pro-activity / Reactivity: "Top management has a preference for some
proactive actions and some reactive actions (cflOO)." Medtronic can best be
described as "pro-active" in its decision-making. The numerous strategic
and marketing plans, lobbying of Congress and other legislative bodies,
customer education efforts, and environmental scanning attest to the fact
that the firm's managers attempt not only to presage but to mold its
environment wherever possible.
Risk Aversion: "Top management can be characterized as risk neutral (cf
100)." Medtronic believes that the greatest risk to itself and the medical
device industry is the sale of unsafe products. Ai?, such, its supports a
rigorous enforcement process by the FDA, including thorough factory
inspections using existing international standards, such as ISO 900l.
However, Medtronic, along with its competitors, faced what they
considered to be an increasingly hostile and anti-business regulatory
environment. The scrutiny by the FDA of Medtronic centered on ensuring
patient/consumer safety. The aim was reducing the risk to patients from
faulty or unproven therapies and products. Their means of achieving this
was to increase their oversight of the product development process.
According to CEO George (George and McKibbin, 1995): "Unfortunately,
over the years, the FDA's prescribed product approval process has been
broadened from evaluation of the safety and efficacy of the device to the
evaluation of the efficacy of the therapy. This broadened role is not a part of
the original legislation, nor is it part of the legislative intent."
Results of these tactics included a lengthening of the product approval
process and the discouragement of the development ofinnovative therapies
and products (George and McKibbin, 1995). Rather than opting for the less-
risky course of developing only those products that would have the greatest
11.4 Discussion ofAnswers to OrgCon Questions 379

chance FDA approval, Medtronic chose to move much ofits R&D to Europe.
The "hands off' approach of the Europeans and the recognition of medical
device development as an "inexact science" is apparently more favorable to
Medtronic.
Motivation and Control: "Top management has a preference for
motivation by inspiration (cf 100). "
The accounts from Medtronic are filled with examples of how the
company inspires its employees with its mission of restoring people to full
life and health, its code of ethics, and its "non-negotiable, universal,
mission-driven values." (George and McKibbin, 1995).
The importance of Medtronic's preference for motivation and how it
factors into the top management's decision-making and strategic planning
was made evident in an interview with CEO George. Below is his descrip-
tion of a potential Medtronic acquisition. "We considered acquiring another
company in a field outside the cardiovascular area that we very much
wanted to get into. They were serving a clear, unmet patient need in our
strategy, and the company was a leader in its field. Its stock price had gone
down because it had hit some rocky times, so it would have been a great
financial move for us to acquire this company.
Medtronic 's earnings per share would have gone way up if we had bought
them. But when we really started looking at the company, we saw that their
values were fundamentally different from ours. They got results from people
by driving them. Either you make it, or you get fired. Our sales force has a
turnover of about 2 percent a year. Theirs has a 30 percent turnover. They do
a great job of training their people, but they have a whole new sales force
every three years.
Theirs is a totally different style of doing business that starts right at the
top. How do you change that? You can't; it's ingrained in the culture. It's
who they are. It's the way they do business. That just didn't fit for Medtronic.
So we walked away from it. The price we'd have had to pay in cultural terms
would have been too great. People would have said, "We thought Medtronic
had clear values, and now you're acquiring this company?"
Interviewer: Was your board behind that decision?
Yes, very much. They were unhappy we even brought it up. We said, well,
it's such an important field that we almost have to bring it up. They are the
leader in the field." (George and McKibbin,1995).

11.4.7 Size
Number of Employees: "Medtronic has approximately 12500 employees (cf
100)." Medtronic has been growing steadily. In 1990 it had fewer than
10,000 employees and by 4130/97 employment had risen to 12,500 employ-
380 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

ees. ll The number might be larger were it not for Medtronic's use of its
temporary workers, what it calls "overload employees" (Solomon, 1993).
While this practice is not unique, what is different at Medtronic is its
keeping temporary workers on the payroll and allowing them to become
eligible for paid time off and holiday pay. "(This) is a clear effort to attract
and maintain an overload work force that can move around the company
and fill in for business units that are having an increase in workloads."
These employees would be the same as temporary employees from outside
firms, however, they only work for Medtronic. Most of the people provide
administrative services. "

11.4.8 Age and Ownership

Organization Age: "Medtronic's age is mature (cf 100)." Medtronic was
founded in 1957. 12
Type of ownership: "Medtronic's ownership status is incorporated (cf
100)." Medtronic is incorporated. It is a public company and its shares are
traded on the NYSE, ticker symbol MDT.

11.4.9 Current Configuration

Configuration. "Medtronic has a matrix configuration (cflOO)." Medtronic's
operations are organized into three global areas: Americas, Europe/Middle
East/Africa, and AsialPacific. The two dimensions of the Medtronic matrix
structure are geography and business unit or venture. There are five
businesses (divisions)-Pacing, Vascular, Cardiac Surgery, Neurological, and
Developing Business & Ventures-encompassing 14 strategic business units
and nine ventures.

11.4.10 Current Complexity

Job Titles: "Medtronic has a very large number of different jobs (cf 100)."
Medtronic's homepage on the world-wide web maintains a partial listing of
current employment opportunities. 13 On April 18, 1998 there were openings
for sixty-three (63) distinct job titles divided into six categories: Engineer-
ing! Scientist (26), Finance (5) ,Human Resources (2), Information Systems
(7), Regulatory/Quality/Clinical (15), Technician (2), and Other Professional
11.4 Discussion ofAnswers to OrgCon Questions 381

(6). The large number jobs described in the first category alone is sufficient
to establish that Medtronic has a high degree of specialization.
The openings in the Engineering/Scientist category were:

Senior IC Design Engineer, Senior Principal IC Design Engineer,

Associate IC Design Engineer, Materials Sourcing Specialist, Sr.
Electrical Engineer, Senior Software Engineer (RES C), Design Automa-
tion Engineer (IC Physical Design), Digital Design Engineer, Principal
RF Design and Test Engineer, Principal Firmware Engineer, Principal
Software Engineer, Sr. Product Assurance Engineer, Senior Software
Engineer, Manufacturing Engineer, Senior Staff Scientist, Principal
Design Engineer, Senior Firmware Engineer, Principal IC Designer,
Principal Firmware Engineer, Firmware Engineers, Software Engineers,
Sr. Electrical Engineer / Electrical Engineer, Senior Software Engineer,
Principal Software Quality Engineer, Sr. Software Engineer (Firmware),
Principal System Verification Engineer, Packaging Engineer, Principal
Electrical Engineer.

Given that this was only a listing of its current openings, it is likely that
Medtronic has many times this number ofjob titles. The openings described
in the other categories, while fewer in number, appeared to be almost as
finely delineated.
Advanced degrees or specialized training: "Of the employees at Med-
tronic, 21 to 50 % have an advanced degree or many years of special
training (cf 100)." This concerns the level or degree of professionalism-the
level of formal education and training of employees-in an organization
(Daft, 1995). Medtronic has a very large number of managerial, profes-
sional, and technical employees including several varieties of engineers and
scientists, medical doctors, professional managers, legal and sales staff
(Mackay,1995; Solomon, 1993; Ray, 1993; Geber,1993).
This large number is due, in large part, to the "high-tech" nature of its
products and the high levels of education required to not only design such
products but also to market them. Decision-makers for the purchase of
much of Medtronic's equipment include cardiologists, cardiovascular
surgeons and electro-physiologists (Rosenbaum,1991). It follows that
products produced by and sold to such highly educated users would require
Medtronic to have both a very highly educated and trained workforce and
a high proportion of such people in its organization. While is quite possible
that the percentage is over 50% but it is not possible to tell given the
information examined. It is almost a certainty that the percentage is over
20%, however.
Number of Vertical Levels: "Medtronic has an unknown number of
vertical levels separating top management from the bottom level of
382 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

organization." Average # of Vertical Levels: "The mean number of vertical

levels is unknown." These two questions are measures ofvertical differenti-
ation. As a detailed organization chart was not available these questions
remain unanswered. There are, however, a few clues evidence to support
the contention that Medtronic does not a very flat hierarchy. One indication
is the number of grades or levels described in the listing of job openings for
engineers and scientists. A careful examination of the openings for one
grouping of positions, those for integrated circuit engineers, reveals that
there are at least four grades of engineer: associate, principal, senior
principal, and senior. The presence of so many grades within the same
position suggests, but does necessarily demonstrate, that there may also be
a tendency in the organization toward creating hierarchical distinctions
between levels in the chain of command.
Geographically distinct locations: "Medtronic has more than 30 separate
geographic locations (cf 100)." Average distance from headquarters:
"Medtronic has an average distance of separate units from organization's
headquarters of 501 to 3,500 miles (cf 100)." These two questions are used
to measure the organization's spatial differentiation. Medtronic does
business in over 120 countries (Collins, 1995; Comline, 1990) with major
R&D facilities in EUrope and Japan, sales office on several continents and
subsidiaries in several states of the U.S. including Arizona, Colorado,
Massachusetts, California, Texas, and New York. 14 Given that many of the
major subsidiaries are several thousands of miles away, it seems likely that
the average distance of the outlying units is in excess of 500 miles.
Proportion working at outlying locations: "From 26 to 60 % ofMedtronic's
total workforce is located at these separate units (cflOO)." This question is
yet another measure of spatial differentiation. As of 4/30/97, the date of the
filing of Medtronic's 1996 Annual Report, the company had 12,500
employees. No figures were given on exactly how many employees work
outside of the Minnesota headquarters. The number of employees at major
U.S. subsidiaries are shown in Table 11.4.
Medtronic has maintained a significant presence in Europe for over 30
years and it currently has four manufacturing sites in the Netherlands,
France, and Germany. Its Bakken Research Center, where most of its R&D
is conducted is in the Netherlands and it also maintains several customer
education centers throughout Europe. Its sales and technical service
organization in Europe includes a network of over two dozen country and
regional offices and several new facilities have been scheduled for develop-
ment in Switzerland. Medtronic's total European employment in 1995 was
estimated at 1500 employees (PR Newswire, 1995b). Employment levels in
Japan are also said to exceed 1000 employees (Pollack, 1994). This places
11.4 Discussion ofAnswers to OrgCon Questions 383

the number of employees outside of the Minnesota headquarters at nearly

5000 employees, well within the 26-60 % range.

Table 11.4 Location, Employment Levels: Medtronic Non-HQ, U.S.

Operations

Name Location Number of Employees

Micro-ReI Tempe,AZ 980 (Alexander, 1995)
Metronic Parker, CO 500
Medtronic, Neurological Div. Columbia Heights, MN 490
Medtronic, Bio-Medicus Eden Prairie, MN 175
Medtronic Interventional Danvers, MA 180
Vascular
Medtronic Cardio Rhythm San Jose, CA 40
Carbon Implants Austin TX 50
Cardio Care Forest Hills, NY 50
TOTAL <2400 or 20 %

11.4.11 Current Formalization

Written Job Descriptions: "Job descriptions are available all employees,
including senior management (cf 100)." The Medtronic homepage for
employment opportunities provided detailedjob descriptions for operational
employees, as well as low and middle management. Typical jobs at the
operational level included Engineering Technician, Polisher, Mailroom!
Copy Center Clerk, and Office Services Assistant. Lower level management
positions described include Production Planner and Associate Accountant.
Middle management positions described included Senior Financial Analyst
and Senior Regulatory Affairs Manager. A Senior Management position
that was described was Information Technology Director. This job
description appears in its entirety below: "Medtronic's US Cardiovascular
Information Systems Department seeks an exceptional Information
Technology Director at our Minneapolis, MN location. This individual will
direct the development and implementation ofInformation Technology plans
for The Americas and ensure plans are aligned with area business objectives.
Responsible to manage 1/ T personnel and resource planning. Influence
1/ T direction by working with the functional areas and positioning 1/ T
standards in alliance with business requirements. Review, manage, and
coordinate the development of business applications using current IIT
technologies for The Americas, balancing the long-range plans and
384 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

technology needs and strategies of both the area and Corporate I I S. Plan
and implement measurements and benchmarks to measure cost effective
technologies. Serve as consultant to all operations within The Americas, to
develop solutions using information technologies. Manage the overall I IT
Resources for The Americas. Maintain knowledge of I IT and the I IS
discipline and keep abreast of emerging technologies.
Candidate Qualifications: Bachelor's Degree in Business, Computer
Science, etc. Master's Degree preferred. Ten year's experience in information
systems. At least three years in a supervisory and I or management role.
Business experience in relevant user areas and I or Medtronic experience is
desirable. (Source: www.medtronic.comlemployment) ."
Non-Managerial Employees: "From 81 to 100 % of non-managerial
employees are given written operating instructions or procedures for their
job (cf 85)." Medtronic provides very detailed job descriptions for a wide
variety of positions-managerial and non-managerial alike. It is very likely
that a high percentage of non-managerial jobs, perhaps all ofthem, having
written operating instructions or procedures as well.
Rules, Percentage in Writing: "From 41 to 60 % of all the rules and
procedures that exist within the organization are in writing (cf85)."
Given the above, it is a safe assumption that at least half of the rules
and procedures at Medtronic are in writing. While this could clearly vary
across departments, functions andjobs, the strong concern that the firm has
on regulatory compliance and quality-a concern that is reflected in a high
degree of specialization-would necessitate that there be a high percentage
of rules in writing.
What it is not possible to tell from the either the posted job
openings/descriptions or from the reviewed articles is how closely the
position incumbents are supervised to ensure compliance with the
standards set in the descriptions. Given Medtronic's preference for
delegation, we would think that the supervision might be loose. However,
with the high concern for quality we would expect it to be a bit higher. As
a compromise, the middle ground is chosen for each of the following four
questions.
Compliance with Written Standards: "Where written job descriptions
exist, moderately close supervision of employees is used to ensure compli-
ance with standards set in the description (cf 85)."
Latitude from Written Standards: "The employees are allowed a
moderate amount of deviation from the standards (cf85)."
Compliance with rules by managerial employees: "The written instruc-
tions or procedures given are followed to some extent (cf 85)."
11.4 Discussion ofAnswers to OrgCon Questions 385

Managerial Employees: "Supervisors and middle managers are to some

extent free from rules, procedures, and policies when they make decisions
(cf 85)."

11.4.12 Current Centralization

Medtronic takes great pride in its policy of decentralizing its operational,
quality and ethical decisions. The company officials see this as an outgrowth
or consequence of the company's clear and consistent mission and values.
Below, Medtronic CEO William George makes clear the connection between
them: "... (W)e are able to decentralize and delegate quality and ethical
decisions to people because there is a framework from which they can make
good decisions. We say, 'You don't have to make a trade-off between cost and
taking care of the patient. You always take care of the patient, and the cost
will take care of itself.' That makes an easy framework for people.
We operate from a Medtronic standard of values-not a US. standard, not
a Saudi Arabian standard, not a Japanese standard. We explain the
standard and ask our employees to sign a statement each year saying that
they have complied with that standard and are committed to complying with
it next year. The most important thing, ofcourse, is that they understand the
values. But if they say they're committed, then we can delegate and
decentralize. There's no way I can ever know what's going on throughout the
organization." (George and McKibbin, 1995).
This statement would seem to indicate that, overall, the level of
centralization at Medtronic is not high, and may even be low. There was
very little information in the accounts, however, which spoke directly or
indirectly to the full extent to which Medtronic decentralizes decision-
making. However, given the apparent ubiquity of and compliance with the
Medtronic mission statement, we conclude the following:
Top Management Direct Involvement: "Top Management is to some
extent involved in gathering information they will use in making decisions
(cf 85)."
Top Management Interpretation: "Medtronic has a moderate level (41-
60%) of participation in the interpretation of the information input (cf85)."
Top Management Direct Control of the Execution of Decisions: "Top
management would seem to directly control quite little (0 to 20 %) of the
decisions executed (cf 85). "
Based on the above we further assume that typical middle managers
have some discretion in: establishing their budget, selecting evaluation
criteria for their units, hiring and firing of personnel, setting personnel
rewards, purchasing equipment and supplies, establishing a new project or
386 Chapter 11. Using Publicly.Available Information for Strategic
Organizational Diagnosis and Design

program, and deciding how work exceptions are to be handled. A certainty

factor of 85 is assigned to all of these answers.

11.5 Total Organizational Analysis

Based on answers about the organization and its situation, the OrgCon has
characterized the organization's size, climate, management style, strategy,
current structural characteristics, and design misfits. These are discussed
in subsections 11.5.1 through 11.5.6, respectively.

11.5.1 Size
"Based on the answers provided, it is most likely that Medtronic is large (cf
100)." The size of the organization-large, medium, or small-is based upon
the number of employees, adjusted for their level of education or technical
skills. The adjusted number of employees is still greater than 2,000. Thus,
Medtronic is categorized as large.

11.5.2 Climate
"Based on the answers provided, it is most likely that the organizational
climate is either a group (cf 78) or a developmental (cf 77) climate." The
organizational climate effect is the summary measure of people and their
behavior. The group climate is characterized as a friendly place to work
where people share a lot of themselves. It is like an extended family. The
leaders of the organization are considered to be mentors and, perhaps even
parent figures. The organization is held together by loyalty or tradition and
commitment is high. The organization emphasizes the long-term benefit of
human resource development with high cohesion and morale being
important. Success is defined in terms of sensitivity to customers and
concern for people. The organization places a premium on teamwork,
participation, and consensus. Employees with a high morale, along with
high leader credibility and high trust are all elements of a group climate. An
organization with low levels of conflict and scapegoating may indicate a
group climate.
11.5 Total Organizational Analysis 387

The developmental climate is characterized as a dynamic, entrepreneur-

ial and creative place to work. People stick their necks out and take risks.
The leaders are considered to be innovators and risk takers. The glue that
holds organizations together is commitment to experimentation and
innovation. There is an emphasis placed on being on the leading edge and,
accordingly, readiness for change and meeting new challenges are
important. The organization's long-term emphasis is on growth and
acquiring new resources, success is defined as having unique and new
products or services and being a product or service leader. The organization
encourages individual initiative and freedom. High morale is frequently one
element of a developmental climate as is high leader credibility. An
organization with low levels of conflict and scapegoating can be categorized
as having a developmental climate and moderately equitable rewards in the
organization drives the climate towards a developmental type.

11.5.3 Management Style

"Based on the answers provided, it is most likely that Medtronic's manage-
ment profile has a medium preference for micro-involvement (cf73).1t could
also be the case that Medtronic's management has a low preference for
micro-involvement (cf66)." The level of management's micro-involvement
in decision making is the summary measure of management style. Leaders
have a low preference for micro-involvement; managers have a high
preference for micro-involvement. The management of Medtronic has a
preference for taking actions on some decisions and being reactive toward
others. This will lead toward a medium preference for micro-involvement.
Since the management has a preference for medium detailed information
when making decisions a medium preference for micro-involvement
characterization is also appropriate. Management is risk neutral and this
is one of the characteristics of a manager with a medium preference for
micro-involvement.
Medtronic's preference for motivating people through inspiration rather
than control is consistent with a low preference for micro-involvement. Its
preference for delegating decisions points toward a low preference for micro-
involvement as well. Management has a long-term horizon when making
decisions, which characterizes a preference for a low micro-involvement.
388 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

11.5.4 Strategy
"Based on the answers provided, it is most likely that Medtronic has a
prospector strategy (cf 84). It could also be an analyzer with innovation (cf
82)." Following the Miles and Snow (1978) typology, the organization's
strategy is categorized as one of either prospector, analyzer with innovation,
analyzer (with or without innovation), defender, or reactor.
An organization with a prospector strategy is an organization that
continually searches for market opportunities and regularly experiments
with potential responses to emerging environmental trends. Thus, the
organization is often the creator of change and uncertainty to which its
competitors must respond. However, because of its strong concern for
product and market innovation, a prospector usually is not completely
efficient. Medtronic has numerous products and a prospector is constantly
seeking new product opportunities to serve both existing and potentially
new customers. A non-routine technology is likely to be-costly for Medtronic,
and a prospector strategy of new product development where margins are
likely to be high is very reasonable. Given its high concern for quality, a
prospector strategy is a likely one for Medtronic.
An organization with an analyzer with innovation strategy is one that
combines the strategy of the defender and the prospector. It enters a new
market after viability has been clearly shown. But in contrast to an
analyzer without innovation, it has innovations that run concurrently with
the regular production. It has a dual technology core. With its high concern
for quality, Medtronic could be an analyzer with innovation. Additionally,
an organization with a medium capital investment is likely to have some
rather fixed capabilities, but can also adjust.
The analyzer with innovation which seeks new opportunities but also
maintains its existing profitable markets. Moderate prices create some
flexibility but also require attention to efficiency. An analyzer with
innovation strategy where Medtronic both attends to existing profitable
markets and seeks selective new opportunities is appropriate.

11.5.5 Current Organizational Characteristics

"The current organizational complexity is high (cflOO): horizontal differen-
tiation is high (cf 100), vertical differentiation is not determined but is
probably high and spatial differentiation is high (cf 100). The current
centralization is medium (cf 85) and the current formalization is medium
(cf89)."
11.5 Total Organizational Analysis 389

11.5.6 Situation Misfits

"A situation misfit is an unbalanced situation among the contingency
factors of management style, size, environment, technology, climate, and
strategy. Medtronic has no situation misfits (cf 100)."

11.5.7 Organizational Consultant Recommenda-

tions
Based on answers about the organization, its situation, and the conclusions
with the greatest certainty factor from the analyses above Organizational
Consultant has derived recommendations for the organization's configura-
tion, complexity, formalization, and centralization. There are also recom-
mendations for coordination and control, the appropriate media richness for
communications, and incentives. More detailed recommendations for
possible changes in the current organization are also provided.

11.5.8 Organizational Configurations

"The configuration that best fits Medtronic's situation has been estimated
to be a matrix configuration (cf79). It could also be: an adhocracy (cf69). It
is certainly not a professional bureaucracy (cf -100), a machine bureaucracy
(cf -100), a functional (cf -100) configuration."
A matrix structure assigns specialists from functional departments to
work on one or more interdisciplinary teams that are led by project leaders.
Permanent product teams are also possible. A dual hierarchy manages the
same activities and individuals at the same time.
Since Medtronic has many products or markets, a matrix configuration
is prescribed. Medtronic's environment has neither low equivocality nor low
complexity and this also points towards a matrix. The fact that Medtronic
is large makes a matrix configuration an acceptable choice but by no means
a requirement. The matrix configuration is also a more likely configuration
when an organization has a unit production technology. When a firm has a
prospector strategy, the organizational configuration can be a matrix. The
matrix structure may be a trans-national structure. When foreign productJ
service diversity is high and international involvement is high, Medtronic
should have a multidimensional, global configuration.
390 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

An adhocracy organization is normally an organization with high

horizontal differentiation, low vertical differentiation, low formalization,
decentralization, and great flexibility and responsiveness. An adhocracy
configuration is appropriate when neither the environmental equivocality
ofMedtronic nor the environmental uncertainty is low. Medtronic has many
products or many markets which indicates that an adhocracy is an
appropriate configuration. However, the size of the organization is very
important for the choice of an adhocracy configuration. A prospector like
Medtronic could be configured as an ad hoc organization. An organization
with a group climate may also have an ad hoc configuration.
Because Medtronic does not have a routine technology, it is quite
unlikely that a professional bureaucracy would be an efficient organiza-
tional configuration. Medtronic's prospector strategy also militates against
a professional bureaucracy.
The configuration cannot be a functional when the technology is
nonroutine. When the organization has a nonroutine technology, it is not
likely that a machine bureaucracy is an efficient organization. When the
organization has a prospector strategy, it cannot be a machine bureaucracy!
When the organization has high hostility, it is unlikely to be a machine
bureaucracy. A machine bureaucracy will prevent it from acting appropri-
ately when unexpected events occur.

11.5.9 Organizational Characteristics

"The recommended degree of organizational complexity is high (cf 56) but
it could also be low (cf 46)." Since Medtronic has a prospector strategy, the
organizational complexity can be either low or high. The large size indicates
that it should have high organizational complexity. Medtronic has a non-
routine technology which implies that the organizational complexity should
be high. Large size combined with a nonroutine technology calls for high
complexity, particularly vertical differentiation and because Medtronic has
an advanced information system, organizational complexity can be greater
than it could otherwise.
A prospector strategy permits the organizational complexity to be either
low or high. Not much is known about the environment since both the
environmental uncertainty and the environmental equivocality ofMedtronic
are high. In this situation, the organizational complexity should be low,
allowing the organization to adapt quickly. When the environmental
hostility of Medtronic is high, organizational complexity should be low. A
group climate in the organization requires a low level of complexity with a
low level of vertical differentiation.
11.5 Total Organizational Analysis 391

"The recommended degree of formalization is low (cf64)." When the

organization is in the manufacturing industry and it does not have a routine
technology, its formalization should be higher than if it was in the service
industry. Medtronic has a prospector strategy so low formalization is
required so that the organization can react quickly. Low formalization is
also required because of the need for innovations. Organizations with
nonroutine technology should have low formalization. Since the set of
variables in the environment that will be important is not known and since
it is not possible to predict what will happen, no efficient rules and
procedures can be developed. This implies that Medtronic's formalization
should be low. When environmental hostility is high, formalization should
be low. A group climate requires a low level of formalization as well.
"The recommended degree of centralization is medium (cf36) but could
also be low (cf35). There is evidence against it being high (cf -17)." Medium
centralization is recommended when top management has neither a great
desire nor very little desire for micro-involvement. Because Medtronic has
an advanced information system, centralization can be greater than it could
otherwise.
Large organizations should have low centralization. Medtronic has a
prospector strategy and a low centralization will help the organization react
and innovate quickly. There are many factors in the environment that affect
the organization but Medtronic does not know which factors are or will be
important. Thus, centralization should be low. A group climate in the
organization requires a low level of centralization.
"With a non-routine technology, Medtronic should obtain coordination
and control via group meetings (cf 96)." Coordination is a major issue
requiring a lot of time by functional managers and product or project
managers. An environment with high equivocality, high uncertainty, and
high complexity requires that coordination and control be obtained through
the use of integrators and group meetings.
Managers should make frequent adjustments in order to maintain
project and product goals and use scarce functional resources and personnel
efficiently. In an international firm, matrix dimensions will likely include
country or region and may include either product, customer, or function.
Project or product managers will likely be required to champion new
innovations in customers, products or technologies. When the organization
has a group climate coordination should be obtained using integrators and
group meetings.
"Incentives should be based on results (cf 95)." Incentives could be
results based but with a group orientation. An organization with a group
climate will likely have to process a large amount of information and will
need to use information media with high richness.
392 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

"Medtronic's span of control should be narrow (cf50)." This recommenda-

tion is an outgrowth of the fact that Medtronic has a nonroutine technology.

11.5.10 Organizational Misfits

Organizational misfits compares the recommended organization with the
current organization. Medtronic does have organizational misfits (cf 100):
in particular, current and prescribed formalization do not match (cf 89).

11.5.11 More Detailed Recommendations

Medtronic should consider decreasing the number of positions for whichjob
descriptions are available and having fewer written job descriptions.

11.6 Detailed Discussion of Organiza-

tional and Situational Misfits
The OrgCon diagnosis of an organization's design typically describes two
types of "misfits"-organizational (design parameter) and situational. The
former is said to exist whenever the level of an organization design para-
meter differs from the level recommended by the OrgCon. For example, the
current differentiation might be high when the recommendation is that
differentiation be low. Situational misfits occur when the design situations
are not internally consistent. The OrgCon's recommendations for a highly
equivocal environment are not consistent with those for a routine technol-
ogy. There are no recommendations for situational misfits.
The OrgCon diagnosis for Medtronic determined that there was one
organizational design parameter misfit related to the level offormalization.
Specifically, the current and prescribed formalization do not match. Current
formalization is medium (cf 89) while the recommended level which is low
(cf 64). The table below summarizes all recommendations for low formaliza-
tion.
The configuration recommendation is probably the weakest of the group.
The OrgCon found that Medtronic's current configuration was most likely
a matrix but could also be an adhocracy. Adhocracy's should have low
11.6 Detailed Discussion of Organizational and Situational Misfits 393

formalization. The other five recommendations must be seriously consid-

ered, however.

Table 11.5 Recommendations for Formalization

Contingency Variable Condition Recommendation

(for Formalization)
Configuration Adhocracy Low
Strategy Prospector Low
Technology Non-routine Low
Environment High Uncertainty Low
Environment High Hostility Low
Climate Group Low

When we re-examine the conditions within Medtronic that contribute to

the misfit, we can conclude that it is probably the following factors which
make the formalization higher than the OrgCon recommends:

• The existence ofjob descriptions for virtually all employees-operational,

lower and middle management, and some senior management.
• The high percentage of non-managerial employees that are given
written operating instructions for their jobs.
• The large percent of rules and procedures that exist within the
organization are in writing.

The OrgCon's recommendations for elimination of this misfit are a

reductions in the number of written job descriptions and in the number of
positions for which job positions are available. The latter recommendation
may be achieved by different means. It may be as simple as discarding the
job descriptions for existing positions or it may mean eliminating positions
completely, possibly through recombination or redesign. It should not be
interpreted as a directive to fire the incumbents ofthe positions!
Remarkably, the OrgCon found no situational misfits for Medtronic.
This indicates that a high level of "internal consistency" exists between the
contingency factors for organizational structure, i.e. strategy, size,
ownership, technology, environment, and management preferences. How
this has been achieved by Medtronic is an interesting question to consider.
It seems to suggest that even in the face of what we assume is a lack of
formal, theoretical knowledge of the principles of organization design,
managers at Medtronic have discovered, possibly through trial-and-error
experimentation, how to achieve a fit between the organization and its
multiple contingency factors.
394 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

Notes
1. The Lexis-Nexis® news service contains over 100 "libraries" each
containing dozens of "files". Each file may contain from one to a few
thousand publications. For example, the Lexis-Nexis Directory of On-line
Services, 1995 Edition describes the NEWS library as consisting of "more
than 2300 full-text information sources from U.S. and overseas newspapers,
magazines, journals, newsletters, wire services and broadcast transcripts."
The NEWS library contains several files, such as English, Non-English,
Dutch, French, German, and Italian Language News, Magazines, All News,
Current News, Major Papers, Newsletters, Newspaper, Transcripts, Wire
Services, U.S. and Non-U.S. News Sources, U.S. Regional Sources, Business
Dateline, and many others.

2. This profile relies heavily on material contained in the International

Directory of Company Histories, vol. 8, Paula Kepos (Ed.).

4.http://www.medtronic.com/annual! profile.html.

4. While very little information was collected or reviewed that related to the
Financial Resources sector, the absence of any mention of scarcity of funds
for expansion or other company initiatives suggests that this sector
presents few major problems for Medtronic.

6. http://www.medtronic.com/annual/msumm.html

7. http://www.medtronic.com/annual!msumm.html

8. http://www.medtronic.comlannuallletter. html

9. http://www.medtronic.comlannual!

9. http://www.medtronic.comlpubliclmedtronic /annual/geo/amer.html

11. Medtronic Annual Report

12. Medtronic Homepage, htt.://www.medtronic.coml

13. Source: www.medtronic.cocmlemploymentlrestrak.html.

14. Infotrac company profile, 1996

References 395

References
Adler, P. 1988. "Managing Flexible Automation." California Management Review,
30(3):34-56.
Alexander, S. 1993. "AT&T opens a national teleconferencing center in Minneapo-
lis; Medtronic is its first customer." Star Tribune. Minneapolis, MN, October 20.
Alexander, S. 1994. "Medtronic to shift some work on its new products to Europe;
Company wants to sidestep U.S. regulatory delays." Star Tribune. Minneapolis,
MN.
Alexander, S. 1995a. "Medtronic updating chip factory." Star Tribune.
Alexander, S. 1995b. "New Product Machine; Medtronic picks up pace of medical
device development to stay ahead of competition and ensure success." Star
Tribune. Minneapolis, MN.
Borger, J. Y. 1993. "3 racing to win by a heartbeat." Minneapolis-St Paul
CityBusiness. August 27, 11.
Burton R. and B. Obel. 1995. Strategic Organizational Diagnosis and Design.
Boston: Kluwer Academic Press.
BusinessWire. 1992. "Medtronic introduces Elite II pacemaker offering advanced
capabilities in smaller size." Business Wire. Minneapolis, MN.
BusinessWire. 1993. "New patient-friendly implantable cardiac device puts
Medtronic at forefront of global cardiac stimulation market." Business Wire.
Monte Carlo: Monaco.
BusinessWire. 1994. "Neurex and Medtronic to collaborate on the development of
products for analgesia." Business Wire. Minneapolis, MN.
BusinessWire. 1995a. "FDA clears new family of pacemakers." Business Wire.
Minneapolis, MN.
BusinessWire. 1995b. "Medtronic plans $26 million investment in Arizona for
advanced microcircuit fabrication technology." Business Wire. Minneapolis, MN,
February 16.
Carideo, T. 1992. "Medtronic is well aware of the long list of talent that it has let
slip away." Star Tribune, August 27.
Collins, A. D. 1995. "Prepared Testimony of Arthur D. Collins, Jr. Chief Operating
Officer, Medtronic, Inc. before the .Senate Committee on Labor and Human
Resources, Subcommittee on Aging and FDA Approval of the Export of
Unapproved Medical Devices." Washington, D. C.: United States Senate,
Federal News Service.
Colman, P. 1995. "Pacemaker flaw to cost taxpayers $ 12 million 2,640 patients in
U.S. may need surgical removal of faulty wire that has killed two." Rocky
Mountain News. Denver, CO.
COMLINE. 1990. "Medtronic Japan to Construct Technical Center in Hokkaido."
COMLINE Daily News Biotechnology and Medical Technology. June 19.
Coser, L. 1956. The functions of social conflict. New York: Free Press.
Couretas, J. 1994. "Medtronic Acquires Molder DIp Inc. For $128 Million." Plastics
News.
Daft, R. 1995. Organization Theory and Design. St. Paul, MN: West Publishing.
396 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

Daft, R. and N. Macintosh. 1978. "A new approach to design and use of manage-
ment information." California Management Review. 21,82-92.
D'Aveni, R. A. 1995. Hypercompetitive rivalries : competing in highly dynamic
environments. New York: Free Press.
David, Gregory E. 1994. "Heart to Heart." Financial World, 163 (18), 22-23.
Engen, J. R. 1993. "Earl Bakken's Hawaiian experiment: The inventor of the
pacemaker attempts to create a pace-setting hospital." Corporate Report
Minnesota. 24.
Feyder, S. 1992. "Medtronic, Siemens settle patent infringement fight; Fridley
pacemaker firm to get $ 75 million now, annual royalty payments over 10
years." Star Tribune. September 12.
Fisher, L. 1993. "Finding Ways to Deliver a Life-Saving Jolt." The New York Times.
February 24.
Geber, B. 1993. "A fresh start for sales training." Training. 30.
George, W. and S. McKibbin. 1995. "The soul of a corporation." Hospitals and
Health Network. 69.
George, W. W. 1993. "An open letter to Tom Peters on spirituality and motivation."
Star Tribune. Apri119
George, W. W. 1995. "It's time to restore innovation in U.S. medical technology."
Medical Marketing & Media. 30.
Gianturco, M. 1992. Electronic lifesavers. Forbes.October 12.
Goldwasser, J. and D. Moreau. 1994. "New Chips Medtronic Has Hearts Aflutter."
Kiplinger's Personal Finance Magazine.
Gresov, C., H. Haveman, et al. 1993. "Organizational Design, Inertia, and the
Dynamics of Competitive Response." Organization Science 4(2),181-208.
Gross, S. 1991. "Medtronic estimates sales of up to $ 100 million during first year
for its new heartbeat product; Firm discloses potential market for PCD device."
Star Tribune. October 10.
Gross, S. 1992, September. "Medtronic stock rises after settlement of suit; Market
continues to react positively to deal with Siemens." Star Tribune. September 15.
Grove, A. S. 1996. Only the paranoid survive : how to exploit the crisis points that
challenge every company and career. New York: Doubleday.
Hamburger, T. 1994. "Those Fda Blues; New technology quickly approved in
Europe." Star Tribune. June 26.
Hamburger, T. and M. Meyers. 1994. "Losing the edge; Minnesota medical firms
say they must export technology, jobs." Star Tribune. June 27.
Hannan, M. And J. Freeman. 1994. "Structural Inertia and Organizational
Change." American Sociological Review, 49,149-164.
Health Industry Today. 1991. "Medtronic and Lilly settle patent disputes on ICDs."
Health Industry Today.July, 54.
Health Industry Today. 1994. "Medtronic outbids St. Jude in Electromedics
purchase." Health Industry Today. 57.
Health Indstry Today. 1995a. "Mallinckrodt, Medtronic form seven-year alliance."
Health Industry Today. 58.
References 397

Health Industry Today. 1995b. "Medtronic moves ventures group to Europe; could
others follow? Medtronic Inc.'s Corporate Ventures organization." Health
Industry Today. 58.
Healt Industry Today. 1995c. "Medtronic, Inc. acquisition of its first supercom-
puter." Health Industry Today. 58.
Health Line Today. 1995. "Medical Technology Companies Depart U.S. to Escape
FDA." Health Line Today.
Inskip, L. 1992. "Employees meet some of those they've helped." Star Tribune.
December 27.
Kepos, P. 1994. International Directory of Company Histories. New York: St. James
Press.
Koska, M. T. 1991. "Restructuring cardiac services: MDs are key to continued
success; KPMG Peat Marwick and Medtronic hospital survey results. " Hospitals.
56.
Laudon, K. C. and J. P. Laudon. 1996. Management Information Systems:
Organization & Technology. Upper Saddle River, NJ: Prentice-Hall, Inc.
Lee, C. and R. Zemke. 1993. "The search for spirit in the workplace." Training. 30.
Lucas, H. C. J. 1997. Information Technology for Management. New York: McGraw-
Hill.
Machine Design. 1992. "Designs for broken hearts; Medtronics creates extraordi-
nary devices routinely with Pro/ENGINEER; cardiovascular and neurological
devices." Machine Design, 64 (21),142.
Mackay, H. 1995. "Bosses Urged to Practice More 'One-on-one'." Sacramento Bee.
August 7.
Marsh, B. 1995. "The Product Liability Morass; Complications Set In; Big Suppliers
Pulling out of Medical Market." Los Angeles Times. Irvine, CA.
Meredith, J. 1987. " The Strategic Advantages of New Manufacturing Technologies
for Small Firms." Strategic Management Journal 8(3):249-258.
Meyers, M. 1992. "Growing abroad; Companies find European market to be
rewarding as well as daunting." Star Tribune. July 6.
Meyers, M. 1995." Dangers and Opportunity on the Net; Firms focus on security as
the expand offerings." Star Tribune. Minneapolis.
Miles, Raymond E. and Charles C . Snow. 1978. Organizational Strategy, Structure
and Processes. New York: McGraw-Hill.
Moffat, S. and J. Mendes. 1991. "Picking Japan's Research Brains." Fortune.
Nelson, G. D., M.D. 1992. "Compassion, technology could reform health care." Star
Tribune. September 7.
Nissen, T. 1992a. "Medtronic Snaps Up Fledgling Firm for Handsome Price."
Minneapolis-St Paul CityBusiness. 9.
Nissen, T. 1992b. "FDA Chiefs Upcoming Visit Has Med Tech Firms Buzzing."
Minneapolis-St Paul CityBusiness. 10. October 23.
Orenstein, R. M. 1995, 1998. "Fulltext Sources Online." Needham Heights. MA,
BiblioData.
Pollack, A. 1994. "Medical Companies Berate Japan." The New York Times. New
York. March 28.
398 Chapter 11. Using Publicly-Available Information for Strategic
Organizational Diagnosis and Design

Prahalad, C. K. and G. Hamel. 1990. "The core competence of the corporation."

Harvard Business Review. 68(3), 79-91.
PRNewswire. 1990. "Medtronic to Break Ground for New Technology Center in
Japan." PR Newswire. Minneapolis, MN. June 14.
PR Newswire. 1992. "Medtronic Reports Regulatory Action on Tachyarrhthmia
Device." PRNewswire. Minneapolis, MN. September 16.
PR Newswire. 1992. "Medtronic Names Collins to Head Operations Outside The
United States." PR Newswire. Minneapolis, MN. May 14.
PR Newswire. 1995. "Cray Research announces first supercomputer order from
Medtronic; First Cray System in biomedical industry to be used in medical
device development, manufacturing." PR Newswire.Minneapolis, MN.
PR Newswire. 1995. "Medtronic acquires Micro Interventional Systems, Inc.,
Expanding neuro business into minimally invasive brain therapies." PR
Newswire. Minneapolis, MN. November 2.
Ray, G. 1993. "Have tools-Need Blueprint; Careful planning and user buy-in, not
tools, are the key to client/server success, experts say." Computerworld.
Riley, K. 1993. "Health care reform plan worries medical technology companies."
The Washington Times. Washington, DC. June 4.
Rosenbaum, S. 1991. "Sensor technology will spur growth in pacemaker market;
development of rate-responsive pacemakers with multiple sensors." Industry
Overview. Health Industry Today. 54.
Sakson, S. 1995. "Heart-starting technology rivalry jolted; Regulators OK changes
in device, giving edge to Medtronic Inc." Austin American-Statesman.
Slovut, G. 1994. "Girl shares heart-warming tale at Medtronic plant." Star Tribune.
Solberg, C. 1994." Med-tech firms paralyzed." Minneapolis-St Paul City Business.
12.
Solomon, C. M. 1993. "Working smarter: how HR can help." Personnel Journal. 72.
Storm, S. 1992. "Corporations Need Responsibility, CEO Says." Successful Business.
8,1.
UPI. 199. "Mallinckrodt offers new test: Medtronic in China venture." United Press
International. June 11.
Verity, J. W. 1992. "Pacemaking Software." Business Week. June 8.
Wieffering, E. J.1995. "The heartofMedtronic-R&D." Corporate Report Minnesota.
26,64.
Wise, R. 1990. "Why aren't quality programs boosting industry profits. Accounting
standards cannot measure all gains form innovations in electronics industry
production." Electronic Business. 16.
Woodward, J. 1965. Industrial Organization: Theory & Practice. London: Oxford
University Press.
Zimmerman, G. and J. Drake. 1995. "Fast-paced design cycle." Computer-Aided
Engineering. 14.
CHAPTER 12

Oticon
by Mikael S0ndergaard and Dorthe D0jbak

12.1 Introduction
The Oticon story consists of four cases each describing a period in the
evolution of the company. Oticon has attracted much attention in the
media both because its management style, its technological development,
and its development of a new organizational form-the spaghetti organiza-
tion.

12.2 Case 1: The Top Management Group

That Got Fired .... by Itself
On 20th May 1988 B~rsen the leading Danish business newspaper ran an
article containing an interview with Bent J. Simonsen, the managing
director of Oticon. In the article, Bent J. Simonsen said that Oticon had
decided to replace the entire management with one absolute leader:
"Absolute monarchy in top management can be used to create a better
feeling of community spirit. And that is exactly what Oticon is preparing
itself for. "(Simonsen, B~rsen, 1988)
Lars Kolind, who was vice managing director of the "Radiometer-
koncern," was sitting in his garden reading B0rsen. He showed the article
to his wife and said: "If the headhunters do their homework properly they
will find me. " (B~rsen, 1995)
And, so they did. The following day the President of Oticon' s Board of
Directors, Palle Rasmussen, called Lars Kolind and invited him for a job
interview. On October 1, 1988, Lars Kolind was appointed CEO at Oticon.
Lars Kolind first task was to create an overview of Oticon's historical
background and current situation.
400 Chapter 12. Oticon

12.2.1 The Demant Era

Oticon was founded in 1904 when the Danish manufacturer Hans Demant
became the sole agency for the American Hutchinson hearing aids.
In 1904 Hans Demant had gone to London to buy a hearing aid for his
wife who was severely hearing impaired. After his return to Denmark,
people heard about the hearing aid and were interested in buying a similar
one. Seeing how large demand for hearing aids was, Hans Demant decided
to contact the American hearing aid manufacturer, Mr. Hutchinson, and
ask him for the sole agency for his products. This was how Oticon, the first
company of its kind in the world, was founded.
In 1910, Hans Demant died and his son William Demant took over the
company (Morsing, 1995:19). With William Demant heading Oticon, the
company started expanding rapidly. Even before World War One, retail
sales covered all of Scandinavia and Saint Petersburg.
Due to the scarcity of goods during the Second World War, William
Demant started his own production of hearing aids in 1944, and in 1946 his
company was the first to produce and sell an all-Danish hearing aid: the
Oticon model TA.
In 1957, William Demant converted his company into the William
Demant, Ltd., later renamed Oticon, Ltd. The share capital was DKK 1
million, and all shares were placed in the Oticon Foundation, which was
established at the same time. The purpose of the foundation was to:
"Assure that all future profits are invested partly in research and develop-
ment related to the compensation of hearing loss, and partly for the benefit
of present and future employees and their families." (Thygesen Poulsen,
1993)
William Demant was an enthusiastic salesman, and he never gave up
an opportunity to sell Oticon products. The aging William Demant, on a
trip to Russia, was so eager to do business that he was selling hearing aids
in his hotel room, with the help of a Russian interpreter to enable
communication (Thygesen Poulsen, 1993).
Employee and customer well being was important to Demant. Personal
contact was the cornerstone of his way of doing business, and his philoso-
phy was that public relations and marketing was a function of friendships
and personal contact (Morsing, 1995): "When he had made a deal he shook
hands on it. No papers and no contracts. His principle was a 'gentleman s
agreement.' 'If you want to understand Oticon you need to know about
Demant. His personality is still an important part ofthis business. '" (Oticon
employees, Morsing, 1995).
As William Demant grew older, market trends were changing. An
increasing number of hearing disabled patients started buying hearing aids
12.2 Case 1: The Top management Group That Got Fired ... by Itself 401

through The Danish Hearing Care, which became an increasingly

important customer of Oticon. The Danish Hearing Care became a
middleman between the actual hearing aid user and Oticon. Then technical
skills, rather than personal skills, became important in sales relations.
In 1958, William Demant decided to delegate the day-to-day operations
to a top management group composed offour employees whom he trusted
to have the necessary technological knowledge combined with personalities
that he knew and respected. However, Demant continued to have his office
at Oticon until he died in 1983 (Morsing, 1995:21).
The president of the new top management group was Bent J. Simon-
sen, an experienced salesman whose personality closely resembled the
Demant's. The other top managers were the three professional business-
men and technically highly skilled persons: Bengt Danielsen, Henning
M~nsted and Torben E. Nielsen.

12.2.2 Growth and Expansion in The Seventies

Oticon expanded dramatically under the new management's direction.
From 1958 until 1985 the company established subsidiaries in 14 countries
(Appendix 1) and independent distributors in 80 countries. Expansion was
driven in part by two factors: first a wish to transform Oticon from a
craftsman's company into an internationally well known company, and
second by market trends leading towards an internationalization of the
hearing aid industry.
The three Vice Presidents headed one department each. Danielsen was
in charge of the administration and accounting functions, research, as well
as long term planning. M~nsted headed the International Department
which included sales and marketing, logistics as well as the 14 subsidiar-
ies. Nielsen was in charge of production and product development, called
the Electronics department.
The management style at the head office was very hierarchical.
Decisions were made democratically by the top management group.
President Bent J. Simonsen had the final say. To a large extent, manage-
ment information was monopolized, and only a minimum of information
and knowledge circulated horizontally. Work processes and means of
coordination were highly formalized. There were fixed working hours and
meetings were held according to pre-established plans. Work tasks were
well described. Within the departments there were great opportunities for
being promoted. All departments were constantly striving to improve and
specialize their skills.
402 Chapter 12. Oticon

Top managemeln group

Pres ident Simonsen
VP Oaniselsen. VP Musted. VP Nielsen

Figure 12.1: Old Oticon's Head Office

Ot,icon's key activities were development and production of superior quality

hearing aids (Thygesen Poulsen, 1993). The product development process
was carried out in an integrated way and the product development function
was organized in teams of experts from the relevant departments. These
teams typically consisted of employees from the departments of research,
development, marketing and production. Final decisions were made by the
parent department and not by the project teams.
Because of lack of space in the mid seventies, management decided to
move the Electronics division into new office facilities, while the Interna-
tional department and the Administration department stayed at the head
office. The geographical expansion of the different departments allowed
management to separate accounting measures and compare results, and
additionally enabled the different departments to concentrate on specializ-
ing within their individual areas. Things went well. The high quality of
Oticon's products and the company's extensive audiological knowledge
base were serving to differentiate it from competitors and enabling Oticon
to charge premium prices well above those of competitors' products.

12.2.3 Technology Changes in the Eighties

Oticon dominated the Danish market for hearing aids from the beginning
of the sixties until the end ofthe seventies (Figure 12.1). Sales revenues
were derived from a variety of countries, and in 1979 the company was the
world's leading hearing aid manufacturer with a 15% share of the global
12.2 Case 1: The Top management Group That Got Fired ... by Itself 403

market. Oticon had grown from being a smaller hearing aid manufacturer
and sales company into being one of the world's leading edge companies
and experts in mass production of technically advanced hearing aids.
Oticon, based upon its advanced technological and audiological knowledge,
was one of the first companies to convert pocket aids into Behind The Ear
aids: "We were an internationally highly regarded company back then.
When one of the four managers showed up at the annual fair in Germany
people would gather around him and ask him about the newest tendencies
and the development in the industry. " (Oticon employee, Morsing, 1995:22).
But in the eighties, new developments were threatening to undermine
Oticon's success: "In the beginning of the eighties everything started to go
really quickly and we didn't wake up at all because we thought our solution
was the only right one. But then Starkey came around and they did well
thanks to their really good marketing. And quite a long time passed before
we finally realized that" (Oticon employee, Morsing, 1995:31)
In the early eighties, the American company Starkey caught world wide
attention when the American President Ronald Reagan wore one of its
new, modem In The Ear hearing aids on an American TV show. In The Ear
aids were of inferior quality to the Behind The Ear aids. This was why
Oticon had avoided producing the aids, believing that their clients would
continue to give priority to quality rather than appearances. However,
sales of In The Ears in the USA quickly increased to 80% (Thygesen
Poulsen, 1993) of the total sales of hearing aids.

12.2.4 The Difference Between Behind The Ear

and In The Ear Aids
As opposed to the traditional Behind The Ear aid, the In The Ear contained
both the power supply and the amplification system in a more compact
casing that could be inserted into the ear and thus much more discrete.
The switch from the production of Behind The Ear to In The Ear aids
meant that the production process became more complex due to the small
size of the In The Ears.
Behind The Ear aids consisted of a small device that was placed behind
the ear with a fine tube that lead the signal into the auditory canal by
means of an earplug. The In The Ears were state-of-the-art products
incorporating flexprint, thick film techniques, integrated chips and surface-
mounted components. The In The Ears were smaller than the Behind The
Ears; the smallest of them so small that it could be placed inside the inner
ear. The so-called "full size" aids were a little larger and therefore also
filled the outer ear.
404 Chapter 12. Oticon

With the introduction of In The Ear aids, production methods started

to change. In The Ear aids required an individual fitting to the customer's
ear and at the same time they also had to fulfil the wishes and the needs
of the individual client. Therefore, hearing aids became just as much
service products as production products, and they could no longer be mass-
produced.
In the Ear aids had great impact on many areas. Customer needs had
changed, hearing quality was no longer the only priority, b~tter appear-
ance was also required. Smaller competitors could now enter the market.
With the increased number of competitors came an increased focus on
prices. The clientele also began to change; private hearing clinics became
more important customers than hospitals and other public hearing centers.
(Fivesdal & Bakke, 1993)

12.2.5 External Influences Leading to Internal

Weaknesses
By 1988 the hearing aid industry had experienced intensified competition.
The hearing aid industry was a good example of an industry in which the
production was concentrated into a few countries. There were more
suppliers who were all able to produce nearly identical products. The
development was fastest on the American market which was also the
biggest market for hearing aids.
The switch from the production of Behind The Ear to In The Ear aids
meant that the production process became more complex due to the small
size of the In The Ear's. In Denmark Oticon's two most important
competitors were Widex and GN Danavox.
Widex, started by former Oticon employees, is a family undertaking well
known for guarding all economic information. Widex is a rather closed
company only discloses information when it is absolutely necessary.
However, there is no doubt that Widex was at the front edge of the
technological development, giving Oticon keen competition. (Figure 12.2).
GN Danavox is a subsidiary ofthe Danish company GN Store Nord with
its substantial net capital. GN Danavox has a very centralized organization
headed by its very forceful director, Ole Lund. (Politiken, Feb. 10, 1993).
AI? for Oticon' s foreign competitors Siemens, the German company was
tough-just as was the American company, Starkey.
All hearing aid manufacturers were dependent on the ability to exploit
technology in an innovative way. The introduction of In The Ear aids had
underlined that necessity. Hearing aid manufacturers were forced to
12.2 Case 1: The Top management Group That Got Fired ... by Itself 405

produce state-of-the-art products continually, and at an ever-increasing

rate. Earlier they were able to live on an invention for ten years, they now
had a maximum period of five years to make a profit.
In the eighties Oticon lost market share. It had been dominant with a
market share of 15% of the global sales of hearing aids (Figure 12.2). In
1988, Oticon had a global market share of 10%. Widex and GN Danavox
had a combined market share of 12%.

Company 1960-1979 1980-1990

Oticon 15% 10%

GN Danavox 5% 5%

Widex 5% 7%

Foreign Competitors 75% 78%

Figure 12.2. The Distribution of Global Sales of Hearing Aids

Source: Thygesen Poulsen, 1993

Oticon's prior success was grounded in solid engineering work, engineers

had become very powerful, and the Electronics department received by far
the most resources. "The heroes were technical heroes. But they were
theoretical and skilled heroes, and not characterized by action and new
ideas. " (Oticon employee, Morsing, 1995:25).
Work processes at the head office were affected by the engineer's
decision on what developments had to be made and when. The geograph-
ical separation gradually alienated the different departments of Oticon.
Communication suffered and diffusion of responsibility was common. They
began to compete for resources. "There was no common organizational
culture. Rather there was a combat culture between Electronics and Inter-
national. It was not even competition ... ifonly it had been... but each ofthem
were just striving to fulfil their own individual goals." (Oticon employee,
Morsing, 1995:24).
Consequently, Oticon was slow to develop and produce new products.
When Oticon finally launched a new product it followed the competitor's
product. Oticon's competitiveness began to deteriorate. These problems
were compounded by the fact that the dollar exchange rate fell drastically
in 1985. With 93% ofits production exported this had serious consequences
for the Oticon's actual earnings (Thygesen Poulsen, 1993).
406 Chapter 12. Oticon

12.2.6 Need for Changes

From the mid eighties Oticon's management realized that things needed
to change in order for Oticon to catch up with competitors' moves and
customers' demands. Employees started feeling insecure about Oticon' s
future; they were impatiently waiting for new initiatives from the
management. "The old management was dragging on. It wanted to leave
the company while it was showing upward going tendencies. To leave a
company going downhill was not their idea, and therefore they weren't good
for the company-they were thinking too short termed. In my opinion they
should have left the company in 1979-ten years earlier. " (Oticon employee,
Morsing, 1995:32).
Several initiatives were taken to improve the situation. The Interna-
tional department started a major campaign in order to restore Oticon's
image. A new MIS (Management Information System) was established.
Some organizational restructuring was implemented. Even though
Oticon's managemen t was still highly respected, employee enthusiasm was
limited: "You do not create a new organization by making a new slogan and
saying a lot of fine words. The old management had no more power, and
this influenced the whole organization by a certain slackness" (Oticon
employee, Morsing, 1995:32).
In 1988, the top management group realized that it had lost its impetus;
it became clear to the management that it had run out of ideas. Oticon had
always been managed in a very hierarchical way followed by a lot ofrespect
from employees. By now the company was so tradition bound that it was
difficult for the current management to convince employees that changes
were for the better.
The board of directors and the top management agreed that it was
necessary to infuse some fresh blood into Oticon to ensure the survival of
the company. They decided that the best thing would be to replace
management with a single leader. It would be easier for one person to get
an overall feel of the organization, and to understand what was going on.
Oticon's top management group had already heard about Lars Kolind.
He had a degree in mathematics, a Diploma in Economics (organization),
teaching experience at the University of Copenhagen (politics), and
experience as director of the research center "Ris0." Currently he was vice
managing director of the "Radiometerkoncern." They wanted to meet him
in person. Lars Kolind was therefore invited for ajob interview. After a few
supplementary interviews, he was appointed CEO at Oticon. He started
work on October 1st, 1988.
12.2 Case 1: The Top management Group That Got Fired ... by Itself 407

Lars Kolind spent his first months meeting employees, competitors and
customers and analyzing the company's financial and competitive situation
(Morsing, 1995). This was what he came up with:

12.2.7 One Must Take the Good with the Bad

Oticon's employees were well educated. Thanks to Oticon's own research
center for psycho-acoustic research-Eriksholm, employees had access to
the latest scientific information on the connection between physical sound
and the experience of hearing-both with respect to healthy patients and
hearing-disabled patients. This meant that employees at Oticon had the
special know ledge required for producing the new, modem In The Ear aids
that customers were demanding.
In the fiscal year 1987, Oticon' s deficit was DKK 41 millions-out of an
annual net turnover of 430 millions (Morsing, 1995: 107). This deficit was
partly caused by the costs associated with the manufacturing of products.
These costs exceeded selling prices by DKK 26 per hearing aid.
Oticon had been losing market share during a couple of years, and could
not update its products. This situation had peaked with Oticon's belated
production switch to the In The Ear aids. "In October 1988 I attended
WHA shearing aid fair in Germany to get an overall picture ofthe industry,
and I was completely overwhelmed by presentations of remote controlled
and digitally programmable hearing aids, especially from Siemens and
Widex, who invested large amounts in the promotion with laser shows and
exhibitions. And during 1989 I became even more nervous while competitors
came up with marketing, seminars, magazines etc. There I was, CEO at
Oticon, and we had nothing to offer. " (Kolind, Thygesen Poulsen, 1993: 62).
By 1988 Oticon had regained its third place, but this was not enough to
secure the future existence of the company. "Oticon was once the world s
leading hearing aid manufacturer. And we have the ability to regain that
positioning. We are going to be the economically healthiest, the technologi-
cally best and, in the end, also the biggest. We are not going to content
ourselves with a number three positioning in the world, we are going to
become number one!" (Kolind, Thygesen Poulsen, 1993).

12.2.8 Kolind's Style

Very quickly, it became clear to Lars Kolind that the situation needed to
change in order to assure the survival of the company. From the beginning
408 Chapter 12. Oticon

he had some ideas on how changes could be implemented. Many ofthese

ideas seemed rather unusual to most employees.
Employees got their first surprise when they met their new manager.
Employees had had a lot of respect for Mr. Demant and his successors
because they were very authoritarian. Lars Kolind was not authoritarian.
He officially pronounced his religious faith and his devotion and close
relationship for his wife and four children. Because his wife was working
full time, he shared all duties with her. Sometimes he would have to leave
early to pick up his children from the kindergarten. His dress style was
relaxed, short shirt sleeves and casual shoes. As a boy scout he had met his
wife and he still looked the part.
Lars Kolind was met with a certain skepticism from the employees and
he did not fit into the distinguished and aristocratic organization of Oticon.
''And it was almost symptomatic at the first reception we held with Lars
because it was Lars who was walking around with the orange juice while
the old sales manager was walking around with the champaign. "
However, probably the biggest surprise came at a meeting shortly after
Kolind's' appointment when he called attention to the unacceptably high
costs associated with the manufacturing of products. Earlier only top
management discussed finance. Now it was put out in the open: "He told
us what was wrong with the economy. In plain words, so that everybody
would understand it." (Oticon employee, Morsing, 1995:111).
He explained clearly that it was necessary to introduce considerable
reductions. No one should feel too sure about their job, regardless of their
position. ''Actually that was the only thing I wanted to make sure during my
job interview with the top management group: That the foundation s
purpose to assure a good working environment would not disable me to
make the necessary firings." (Lars Kolind, Thygesen Poulsen, 1993:49).
This was a completely new concept for Oticon employees, who,
previously, had virtually been secured lifelong employment.
To immediately reduce the unacceptably high costs, Lars Kolind fully
centralized the economical decision making authority and demanded to
sign every cost disposition leaving the company. Additionally, all large
capital investments were to be presented to him with calculations of the
expected earnings. "I chose the economy as a means to get control, and after
a few days I controlled everything. No one could move a finger without
having my accept to do so. " (Lars Kolind, Morsing, 1995).
Even though Lars Kolind formally took over the full responsibility of
Oticon, three of the former managers remained. The former technical
director was asked to reorganize and "clean up" the component factory at
Hejrevejen. The sales director was appointed director of Oticon's subsid-
iary in France. And, the former CEO became president of the Oticon
foundation.
12.2 Case 1: The Top management Group That Got Fired ... by Itself 409

"Of course there was resistance, and his ideas were met with a certain
scepticism from the former directors who were still there and the board of
directors that was still there. But I think that most employees found his
ideas exiting. " (Interview with Torben Petersen, Oticon May 12, 1997) .

12.2.9 Reductions, Cut Backs and the New

Oticon
The centralization of the economical decision making authority was
followed up by several reductions. Ten percent of the employees in the
administration department were dismissed, and the department of long
term planning was closed down (Thygesen Poulsen, 1993). There were also
cuts and budget reductions in the departments oflogistics, sales, market-
ing, and accounting. This was not popular among the employees, But, the
results were spectacular.
After the first quarter of 1989, Lars Kolind had slimmed down the
organization and profits were positive. To show his appreciation, Lars
Kolind gave employees a box of chocolate with a covering letter saying:
"Now we earn more money in than we spend~ongratulations to all of
us!-Lars." (Thygesen Poulsen, 1993).
By December 1989, after one year, gross sales had risen by 4% to DKK
444.6 million and profits to DKK 22.3 million against a loss of DKK 6.2
million in 1988. It was the first positive profit in three years.
Actually 40% of the annual net turnover in 1989 was constituted by
products that had been introduced in 1987, one year before Lars Kolind's
appointment. What Lars Kolind had done was to reduce manufacturing
costs so that sales now contributed with positive figures. These cost
reductions had primarily been carried out by running faster.
In Denmark, Christmas is a time for relaxation, joy and reflection. A
time where you take the time to think deeply about things. Lars Kolind
spent his Christmas vacation 1989 trying to conceptualize a new organiza-
tional structure of Oticon that would enable the company to take on
competition and regain its positioning as a leading edge company.

12.2.10 Questions for Consideration

1. Describe the current situation for Oticon in 1978, 1980, and 1988.

2. Discuss situational misfits and how they should be resolved.

410 Chapter 12. atican

3. Recommend an organizational structure for Oticon for each of the three

periods.

4. Discuss the strategic and organizational change process.

What kind of change was needed?
Why hire Lars Kolind?

12.3 Case 2: The Spaghetti That Became

Revolutionary
When Lars Kolind, CEO at Oticon, returned from his Christmas vacation
in January 1990, he may have looked a little more enthusiastic and self
assured than usual.
Lars Kolind had spent his vacation trying to conceptualize an idea he
had gotten during his first year at Oticon; he wanted to create a company
that was open, transparent and mobile. Kolind believed this was the only
way for Oticon to regain its positioning back from the seventies as the
world's leading hearing aid manufacturing company. Because Lars Kolind's
way of creating such a company would involve drastic changes, he had
coined his idea: "Think the Unthinkable."

12.3.1 What Made Kolind Think of Spaghetti?

The overriding idea ofKolin d's "Unthinkable Thoughts" was to promote the
co-operation between the different departments at Oticon.
In the seventies, Oticon was the world's largest hearing aid manufactur-
ing company. The company's success had turned Oticon into a proud and
tradition bound company, dominated by an engineer's culture. The
company had used its audiological know ledge to develop ever more
advanced products without considering whether clients were actually
interested in the improvements. At the end of the eighties technology
changes had made it imperative to integrate the individual customer's
demands into the hearing aids, and service had become at least as
important as production itself. Oticon had fallen behind, and the com-
pany's internal communication problems had made coordination difficult
and new product launchings rare. As a result, Oticon had overlooked the
importance of the new, modern In The Ear hearing aids that had become
very popular among customers.
12.3 Case 2: The Spaghetti That Became Revolutionary 411

Hoping to change the situation, a new CEO, Lars Kolind, was appointed
in 1988. At the end of 1989, Lars Kolind had turned Oticon's deficit into a
profit, but he was still not satisfied. Kolind wanted to regain Oticon's
positioning back from the seventies and this could not be done with
Oticon's current structure.
Oticon's employees were all well-educated and had the required
knowledge for producing the new In The Ear hearing aids. But coordination
was difficult in Oticon's functionally defined organizational structure
where each department was striving to fulfil its own goals rather than the
organization's. According to Kolind, Oticon had to integrate its audiological
knowledge base into the work of all employees, and at the same time co-
ordinate the work of its employees so that everyone would become more
customer oriented and service-minded (Thygesen Poulsen, 1993). "The
competitiveness of companies will depend on their ability to generate
knowledge, to utilize knowledge in new products and services and to turn
such products and services into profitable business. " (Kolind, 1993).

12.3.2 Holly, Ivy and Unthinkable Thoughts

Lars Kolind was happy. During his Christmas vacation 1989, he had
managed to conceptualize his idea of turning Oticon into a knowledge
creating organization. Kolind's idea was to promote the co-operation
between the different departments by creating a firm that was open,
transparent and mobile (Thygesen Poulsen, 1993). "The vision was to
develop a truly knowledge based company, which could make a difference
in the form of a breakthrough in user satisfaction with hearing aids. This
not only requires a creative combination of technology, audiology and
psychology, but also a much closer co-operation between the different
professionals involved in the actual selling, fitting and fine-tuning of the
hearing aid to the needs of each individual user." (Kolind, 1994).
Part of the credit for Lars Kolind's "Unthinkable thoughts" probably
goes to Lars Kolind's older brother Peder Kolind who is known internation-
ally for several successes with entrepreneurial firms (BlIlrsen, 1995). For
several years, the two brothers had toyed with the idea of designing the
"company of the future." Lars Kolind had also stated how important it was
for Danish firms to understand how to use new technology intelligently
(Kolind, 1994).Kolind did not see new technology as a threat but, on the
contrary, as a valuable opportunity (Kolind, 1994).
The mission of Oticon would still be to compensate for hearing loss, but
seen in a broader perspective than the technological products as such. A
new important part of the company's profile would be to increase service
412 Chapter 12. Oticon

towards a new group of collaborators that had become increasingly

important over the eighties: the private hearing clinics.
Paramount to Oticon's new organization structure was that the whole
company, and not just the individual department, should benefit from each
individual's knowledge and qualities. Technology would not only have to
be built into Oticon's new products but also be used daily by employees
who should all be connected by a large computer network. In order to
become more service and customer oriented, employees should in the future
work across the organizational, and all assignments should be carried out
in project groups. In this way it would be easier to integrate the different
parts of the firm, thereby developing and implementing product concepts
faster and in better compliance with the demands and needs of the
customers (Thygesen Poulsen, 1993). Kolind believed this would enable
Oticon to improve its competitiveness: "Competitors will always be capable,
without much ado, of copying products. But if we create an organization
that can react very quickly and change to comply with demands of
customers and competitors' moves, Oticon will be hard to beat. " (Kolind,
1993).
A flexible and an adaptable organization would require a quicker
decision making processes. Titles would be abolished. Decisions were to be
made by those influenced by the decisions. "When in doubt-then do it. Ifit
works it's good, if not you're forgiven" and "It's easier to be forgiven than
to get approval" (Industry Week, July 19, 1994). These new mottos made
clear that decision making authority would be delegated to employees.
Kolind described the meaning by using the following metaphor: "The
employees hidden resources are brought out so they can play together like an
efficient football team where each individual knows the objectives and the
strategy and can kick the ball immediately without having to look up the
rules and ask why." (Thygesen Poulsen, 1991).
Lars Kolind was not afraid to give employees responsibility as he had
seen, through his many years as a boy scout, that good results could be
achieved through co-operation and delegation of responsibility: "No one can
resist confidence. That is the greatest thing a manager can give an employee,
and it most certainly activities creativity." (Lars Kolind, Weekendavisen,
January 1995).

12.3.3 Realizing the Unthinkable

This time, Lars Kolind did not aim at the company's two factories and their
total of 570 employees, as he had with many of the cost reductions during
Lars Kolind's first two years. Nor were the changes aimed at the 500
12.3 Case 2: The Spaghetti That Became Revolutionary 413

employees in Oticon' s foreign subsidiaries. 145 employees at the company's

head office were effected.
On April 18, 1990, Lars Kolind sent out an informal eight page letter in
which he introduced employees to his new project:"Think the Unthinkable."
Everybody would need to think a little non traditionally to make the idea
come true. Kolind called his project "Project 330;" it aimed an improvelaent
of the efficiency by 30% over 3 years.
Later in April 1991, the public was at a press conference arranged
together with the Ministry of Industry and Business. Oticon had been
chosen by the Industry's Employees as one of five companies to participate
in a project: "Company of the Future." Oticon received DKK 2,5 million
from the Ministry ofIndustry and Business. The idea ofthe project was to
create Danish role models and experiences by implementing new technol-
ogy and efficiency improvements of 20-30%. For economic support, Oticon
agreed to publish results of the project. (Morsing, 1995)
The press conference was covered by most Danish newspapers, and from
then on, Oticon was frequently mentioned in the Danish press.

12.3.4 Implementation of the Undefinable

For all changes employees were invited to take part in the discussion of
how changes should be implemented. Lars Kolind set the general
framework. Then employees were encouraged to fill out the frame
themselves. "He is extremely ... that is one of the really nice things about
him ... he is extremely good at letting people do the things themselves."
(Torben Petersen, Oticon May 12,1997).
As an example, the employees were to find a way to implement the
Unthinkable Thoughts. Two overlying projects were established: one group
with responsibility of the physical arrangements and information flow
between top management and employees, and another group, the so-called
IT group, with responsibility for the company's computer system. Mr.
Torben Petersen was appointed at Oticon in 1990, and his first role was to
function as leader ofthe IT group. Torben Petersen describes his first year:
"I started in 1990, and of course I had talked to Lars prior to that, and
people had told me that he traveled a lot. But I did not expect him to be
absent three weeks from the day I started! But, anyway, we had talked
about me having to make this office paper less. And he had told me how it
had to be, that we needed a computer system so that the office could be made
paper less. So, when I arrived, people showed me where my desk was, and
then it was all up to me what to do with the office. And then, two or three
weeks later, I heard somebody whistling in the corridor, back then Lars
414 Chapter 12. Otkon

always whistled, he doesn't 'really do that now. But he came into my office
and said: 'Oh hi, good to see you re here. ' And we talked a little and then he
said: 'There 'sjust one thing you need to remember, no matter what, I don t
want you to say that this isn't feasible' And then I didn't see him for the next
couple ofmonths ... I did ofcourse have some meetings with him but during
that first year but only very few -and I spent DKK 25 million! That's pretty
much Lars 'management style I think. "(Torben Petersen, Oticon, May 12,
1997).
Lars Kolind's way of managing the changes was indeed different from
what employees had seen before. By practicing "management by walking
around," and by moving his own desk away from the CEO's office and into
the middle of the organization, Lars Kolind gave employees a chance to
talk to him in an informal way rather than walk all the way up to the top
floor to the CEO's office to ask a question: "That's the idea of Oticon's
management-that the management works with the employees at the same
level as the employees. " (Kolind, Drl/lmmejob, April 24, 1997).
A great effort was made to inform employees of the new ideas and
visions. Messages and letters were regularly sent out, keeping everybody
up to date. Lars Kolind held a lot of informal meetings and speeches in
order to communicate his goals to employees.
Because some ofthe employees had never even worked with a computer,
Oticon's new information system was expected to be difficult for them.
Therefore, to familiarize everybody to computers, everyone was offered a
home computer. Eighty percent ofemployees accepted this offer, and many
of those who did not probably already had a computer at home. (Thygesen
Poulsen, 1993)
Another element in the new strategy was the idea of co-ownership. The
capital stock was increased from DKK 20 to 50 million in 1990 and another
capital infusion saw Kolind subscribe DKK 26 million shares, and the
employees DKK 2,5 million. This was not only done to rise the capital of
the company but also because Kolind believed that: "People playing with
other people's money do not feel the same responsibility as those who have
something stuck financially. It is not tenable if e.g. a manager comes to a
point where he is more interested in his own carrier and less in the survival
of the company." (Lars Kolind, B(Jrsen, 29. January 1993).

12.3.5 The Physical and Psychic Influences

To break old habits Lars Kolind decided that Oticon's entire head office
had to move into a new location (Thygesen Poulsen, 1993). Lars Kolind
proposed that the entire company, including the company's factories and
12.8 Case 2: The Spaghetti That Became Revolutionary 415

Eriksholm, should move to Thisted, where he had found an old castle big
enough to house the entire organization. After much deliberation, this idea
was abandoned because Oticon's employees, particularly the middle
managers, strongly protested. Later, it was finally decided that Oticon's
head office should move into Tuborg's closed mineral water factory on the
Strandvejen in Hellerup.
At 8 o'clock on August 8,1990, Oticon's new head office opened its doors
for the first time.

12.3.6 The Computer Network and the

Paperless Office
"The "moving in day" was an amazing experience. Everybody had been
really down and no one felt really safe. But when finally we had packed our
boxes and were sure to be "in ", we almost got euphoric. Everybody was on
their toes, it was really a fantastic experience. " (Oticon employee, Thygesen
Poulsen, 1993:92)
If there were any employees left who had not yet understood the
comprehensiveness of the changes, the new working environment must
have helped. The new organizational structure was very much visualized
through the company's new office facilities. Two big pillars ran the length
of the ground floor. One of them was engraved with the words: "Cogitate
Incognita" as a constant reminder of the importance of thinking the
unthinkable. The other pillar was made of glass and was as a powerful
symbol of the new paperless office. It functioned as a chute for all paper
that had been shredded on its way to the incinerator in the cellar.
Departmental barriers had been abolished. Oticon had been made into an
open office landscape without individual offices, departments andjob titles.
The physical arrangements of the company were designed to keep lines
of communication open. The company's lift had been closed to encourage
people to use the spiral staircase, because you get to talk to more people on
a staircase than in a lift. Further a cafe and several small cafeterias had
been established for employees to hold small, informal meetings over a cup
of coffee. PC's were found on every rolling table that constituted the desks
of employees. Rolling tables were easily rolled back and forth in the
company, depending on where employees worked.
Oticon had been made transparent; information was made equally
available to everybody. Increased communication, increased use of com-
puters and reduced use of paper. Less paper was necessary, as mobility
could only be attained once employees were no longer dependent on the
416 Chapter 12. Oticon

information stored in their ring binders behind the desks. However, Oticon
was not completely paperless and employees could still work with paper,
as long as everything that needed to be saved was digitally scanned into
the network. All PC's were connected to a central computer network which
was custom-built for Oticon by Hewlett Packard.
Employees could find the information they needed onOticon's computer
network. They could also find information on what was going on in the rest
of the organization. Anybody could log into the management's computer
and seek information. On the other hand, the computer network also made
it possible for management to login to the employees' computers and see
how their work was progressing.

12.3.7 Spaghetti Sometimes Gets Sticky

Ironically, only six months after the introduction of Oticon's new company
structure, the informal and flexible structure turned out to be too
centralized. With no middle managers, all decision making was put onto
the CEO who, in turn, was overloaded with information.
All the same, this gave Oticon its first chance to prove its ability to
change. Oticon sorted out the problem by quickly integrating a new level
of management, the so-called management group, into its structure. The
idea of the management group was to create a forum for discussion and
decision making that could relieve the pressure from top management.
With the management group being installed, Oticon became a two level
organization with top management and project groups. In between these
two levels, a more fluid level was acting: the management group.
Members of the management group were going to be two CEO's
(discussed later), many ofthe project owners, project leaders, professional
managers, personnel managers, the project co-ordinator and some project
members. With Oticon's new working structure, all of these titles filled out
more than one function/role at a time, and could very well participate as
project members in one project while being project leaders in another. This
gave members of the management group a cross organizational overview
which, combined with their professional background, enabled them to
decide what projects were important and needed special resources or
attention, and also, to keep an eye on the overall situation ofthe company.
Some replacement of members ofthe management group had to take place
every now and then to avoid putting too much decision making authority
with a few, particular people. "But because the CEO s are a part of this
group it became an organ not an intermediate link (between project groups
and top management). Not a three layer system with the CEO, the
12.3 Case 2: The Spaghetti That Became Revolutionary 417

management group and the projects, but a kind of two layer system, with
the CEO at the table send (of the management group), there is no doubt
about that, .. but more like a level of management and then the projects. "
(Torben Petersen, Oticon, May 12, 1997).
A second element of change were the project recruitment methods.
These were changed as it turned out that Oticon's internal communication
methods worked better than planned!
Originally, the idea was that Oticon's computer network should contain
a job exchange, where all projects were going to be registered with an
explanation of their objectives. Employees were then supposed to apply to
the various projects over the job exchange. Nevertheless it turned out that
employees did not sign up on the projects. The job exchange did not work.
"People never signed up on the projects, it did not work. We had to go out
and tap them on their shoulders. Maybe people were a little afraid. The job
exchange has now been replaced by informal job search. " (Torben Petersen,
Oticon, May 12,1997).
Instead, project recruitment took place in an informal way: Employees
all keep an eye on what projects are being established, and when inter-
ested, they would contact project leaders informally and ask them for a
project membership. When the word of mouth recruiting started replacing
the job exchange, it became harder to keep an overview on who was
participating in the different projects. To make sure that there was the
right mixture of people and skills in the project groups, project leaders
needed approval from the management group for the composition ofhislher
group before starting work.
A third element of change was the constitution of top management.
Oticon's restructurings had been connected with a substantial money
outflow, and consequently of this, the board of directors had urged Lars
Kolind to accept a co-director with financial responsibility. "I was dead
scared of the idea that during such an important restructuring process we
only had one man at the CEO level. Imagine if one day he would drop
dead." (Sven F. Thomsen, now president of Oticon's board of directors,
Thygesen Poulsen, 1993).
Niels Jacobsen was appointed CEO at Oticon on January 1st, 1992.

12.3.8 Spaghetti in Action

From the hearing impaired patients' perspective the first tangible result
of Oticon 's new organizational structure was Multifocus, launched in 1991.
Multifocus automatically adjusted, without any volume control, to the
noises ofthe environment. Actually, Multifocus had already been developed
418 Chapter 12. Oticon

under the old Oticon under the name E36. In 1978, researchers at
"Eriksholm" had developed the concept and the product. But due to lack of
communication, no one had been asked to finish and market it. It was
abandoned and forgotten. In 1991, it was "rediscovered" and its develop-
ment completed. Only four months later it was introduced on the market
with great success.
The launching of Multifocus was only one out of many future episodes
at Oticon to catch the public's attention. Much of this attention has been
focused on Oticon's economic results since the restructuring (appendix 2).
Lars Kolind's personality and untraditional management methods have
been ardently discussed: he was the originator of the Unthinkable
Thoughts. Nevertheless, not everybody recognized the media's description
of Lars Kolind. In 1994, changes were made at Oticons component factory
in Copenhagen. Sixtythree employees were fired. Lone Jensen, Oticon's
employee representative, publicly expressed her frustration with Lars
Kolind. She did not recognize the medias' description of Lars Kolind as the
boss she knew: "One night I dreamt that the management had to chase
employees around in the departments to get them to accept their redundancy
notice. " (Det fri Aktualt, November 24, 1994).
Shortly after this episode a meeting was arranged, open to public, where
Lars Kolind and Lone Jensen discussed their different point of views. At
the meeting Lone Jensen wished to discuss Oticon's pointless firings and
stingy retirement remunerations. After more than an hour, the two had not
reached accord, and the meeting was ended.
This meeting did however not change brokers' positive view of Oticon..
In May 1995 when Oticon was introduced on the Copenhagen Stock
Exchange its share price rose from 500 to 591 in a single day.
Oticon's introduction at the Stock Exchange was shortly afterwards
followed up by another success: in November 1996, Oticon launched a
product which, in the world of hearing aids, was considered no less than a
sensation. Oticon's new hearing aid had been developed at Oticon's head
quarters, and in Cupertino with international researchers and audiologists.
How Oticon managed to co-ordinate its actions across the world may have
seemed like a bit of a mystery to some people not familiar with the
spaghetti structure.

12.3.9 Questions for Consideration

1. Which changes would you make in running OrgCon for 1990?

2. What were the major changes?

12.4 Case 3: What Happens if the Spaghetti Gets Cold? 419

3. What is the relationship between the physical structure and organiza-

tional structure?

4. Would you like to work for Oticon? Why / why not?

5. What is the relationship between the organizational/management

changes and product innovation?

12.4 Case 3: What Happens if the Spa-

ghetti Gets Cold?
In September 1996, Oticon, the Danish hearing aid manufacturing
company, launched something which was considered no less than a
sensation in the world of hearing aids. Oticon had produced the world's
first fully automatic hearing instrument, based on a revolutionary new
microchip. This microchip had the same processing power as a 486 PC,
weighed only 4 grams, was smaller than a little finger nail and used only
five thousandths of the energy used by a conventional hearing aid chip.
Oticon had built the microchip into Digifocus, which the company claimed
to be the world's smallest hearing aid, with a sound reproduction quality
so far unprecedented (Berlingske Tidende, September 13, 1995).
The Digifocus was not the first thing about Oticon to catch the public's
eyes. Ever since 1991, when Oticon restructured its head office into what
the company's CEO, Lars Kolind calls a spaghetti organization, Oticon has
received world wide attention because of its untraditional organizational
structure and management style.

12.4.1 Kolind's Management Style

"Obviously Lars' management style has been very visible. Lars'strength is
that he is dazzling at..., he has been dazzlingly good at selling the company.
And that is one of the reasons that he has so much penetration power in the
organization, because everybody wants to be where he ...you hear a lot about
how interesting Oticon is, right..., and that is why everybody wants to be
where he is, also internally I mean. There is no doubt his personality means
a lot to Oticon, it definitely does. " (Torben Petersen, Oticon, May 12, 1997)
Lars Kolind managed Oticon in a somewhat untraditional way. This
was accentuated in December 1995. Just after the introduction of
420 Chapter 12. Oticon

Digifocus, Lars Kolind did something that, seen from an outsider's

perspective, was rather unusual. On the other hand, what he did only went
to show that Oticon's untraditional management style was not all talk.
Kolind relocated all the former Digifocus project groups, based on the time
horizon of their new projects: "I exploded the organization" "It was total
chaos-within three hours, over a hundred people had moved. To keep a
company alive, one ofthe jobs oftop management is to keep it disorganized. "
(Kolind, Polly LaBarre,Fast Comp., 1996)
Later he explained his approach: "I have to admit that I think that in
many companies, you can say that the management doesn't really attribute
anything positive at all. Because the management is demotivating. But we
do have a management at Oticon ... very much so.. and that management,
including myself, does make many decisions. But the difference is that there
is not a whole hierarchy of managers who spend inappropriately amounts
of time to let the hierarchy stay alive. The whole aspect of hierarchy, we
have avoided that at Oticon to a large extent. " (Kolind, Dr~mmejob, April
23, 1997).
"When people arrived at their new desks their business cards were title
less. And that has been, and still is by the way, a problem. It's such an
important part ofyour identity, the title you have that it can be really hard
on someone not to have it. " (Kolind, Dr~mmejob, April 24, 1997)
"Management must be based on shared values rather than formal
structures" "The better we understand how our job fits into Oticon S goals
and strategy, the less we need traditional management and control. And the
less time we need to spend on traditional management and control, the more
time we have to spend on things that our customers can appreciate."
(Kolind, Dr~mmejob, April 23,1997)

12.4.2 How to Get Value for Values

Top management's wish to communicate its values to employees has
resulted in a special "employee vision" which has been formally docu-
mented in the leaflet: "Oticon's Employee Vision-Vision and Reality."
Among other things it states:

• We regard all employees as mature and responsible people

• We assume that Oticon S employees wish to take on responsibility if they
get the possibility to do so
• We assume that Oticon S employees are more interested in exiting and
challenging tasks than in formal status and titles
(Oticon s Employee Vision -Vision and Reality)
12.4 Case 3: What Happens if the Spaghetti Gets Cold? 421

Management's basic belief about human nature is that people like changes.
(Torben Petersen, Oticon, May 12, 1997) and that employees will always
welcome changes as long as they are let in on what they encompass and as
long as they have a chance to influence their implementation.

12.4.3 Interaction and Integration

All work at Oticon is completed in interdisciplinary groups with very
diverse theoretical and practical backgrounds. There are two types of
projects: Technology development projects and hearing aid construction
projects.
Technology development projects often proceed hearing aid construction
projects by five to ten years. Technology development proje'cts deal with the
fundamental audiological research. The goal is to dete~mine the basic
technological techniques required when creating the actual hearing aid.
Technology development projects usually take place at Eriksholm, Oticon' s
center for psycho acoustic research.
In the case of Digifocus, the technology development project objective
was to find a way to use the new, digital technology in a hearing aid. This
led to the development and design of an integrated circuit that could be
used for hearing tests. On the basis of these hearing tests, it was decided
that digital technology was suitable for use in hearing aids.
During technology development projects at Eriksholm, Oticon cooper-
ated with international audiologists and hearing aid specialists. They were
connected to via Oticon's foreign subsidiaries and through Bernafon, a
Swiss hearing aid manufacturing company that Oticon took over in 1995.
When technology development projects turn out successfully, which was
the case with Digifocus, they are followed by hearing aid construction
projects.
Hearing aid construction projects are concerned with the building,
producing and marketing of the actual hearing aid. With Digifocus, the
hearing aid construction project was to find a way to use the new,
integrated circuit in a way that could help hearing impaired. This involved
market research on the focus of hearing improvements. It also involved
designing and building the actual plastic shell containing the aid, and
later, the production and marketing of Digifocus. Hearing aid construction
projects usually take place at Oticon's head office.
During hearing aid construction projects, employees at Oticon's head
office collaborate with the company's foreign subsidiaries, with interna-
422 Chapter 12. Oticon

tional audiologists and with the company's production factory in the

smaller, provincial town, Thisted, in Northern Jutland.

12.4.4 Thisted's Role in the Spaghetti

The factory in Thisted is organized into: A production department and a
department for production techniques.
The production department is where the actual production takes place,
and it mostly employs unskilled workers. In 1990, Thisted started making
smaller organizational restructurings which, among others, included the
distribution of more assignments and responsibility to employees.
Additionally, holes were made in the factory walls, separating the different
production chains. This was done to create a more open working environ-
ment. Foremen manage up to 40 workers. The foremen are responsible to
the factory manager, Lars Kirk, who is also the top manager for the
department of production techniques.
The department for production techniques employs salaried workers,
typically engineers and technicians. Work in this department is completed
in project groups, the office landscape is open and transparent, and there
are no job titles. Salaried workers are all responsible to the technical
manager, who, in turn, is responsible to the factory's top manager, Lars
Kirk.
Thisted has been connected to the head office's computer network since
1991. Employees at the head office and at the factory in Thisted communi-
cate on a regular basis-either through the computer system, e-mail, phone,
and fax. Additionally, video conferences are held almost every day.
Personal contact takes place when employees fly over and visit. Oticon is
a frequent flyer on the route: Thisted-Copenhagen, even though the
frequency is a little imbalanced in favor of Copenhagen.
Collaboration with Thisted is particularly intense when the project
reaches the stage of designing and developing the actual hearing aid. At
this stage, employees from Thisted participate in the actual projects as
experts on production techniques. Their role is to judge whether it is
technically feasible to produce the hearing aids that researchers have
developed, or whether adjustments are needed. As it is their role to get the
actual production running, they will typically participate towards the end
of the overlying project.
Oticon owns its own component factory at Hejrevejen in Copenhagen.
To increase the mutual connection between the two factories and to
increase their understanding of each other's work, an exchange program
12.4 Case 8: What Happens if the Spaghetti Gets Cold? 423

between the two factories has been established, where four employees at
a time change jobs for three days. (Thygesen Poulsen, 1993).
The component factory at Hejrevejen has introduced a partial self-
governing system. By openly showing lists with the company's goals, where
employees write down to what extent they are attaining the goals,
everyone can clearly see which direction they are heading.

12.4.5 Spaghetti for All Nationalities

Oticon's subsidiaries also contribute with knowledge to projects.
Oticon sells its hearing aids in more than 100 countries, and about 93%
of the company's sales are outside Denmark (appendix 2). Not counting
Bernafon, Oticon has 15 foreign subsidiaries and 80 independent distribu-
tors. The Bernafon distribution network consists of 6 sales companies and
about 40 independent distributors throughout the world (Oticon, company
leaflet, 1997). Oticon and Bernafon function as two parallel and dependent
hearing aid companies, they compete on the same markets, but exploit a
common capacity within production and logistics.
Most subsidiaries have less than 20 employees, and their primary task
is selling. Each subsidiary has its own manager. Several of the subsidiar-
ies, e.g. subsidiaries in the USA, New Zealand, Australia, Norway, Sweden,
Holland, Spain, Switzerland and within next year Germany, have already,
either moved into new facilities, or restructured the company's offices in
order to facilitate dialogue. The idea is to gradually connect Oticon's
foreign subsidiaries to the head office's computer network.
Oticon keeps itself abreast of changes in foreign customer demands and
environmental trends such as law changes, by staying in regular contact
with subsidiaries. Also, different countries like different products. As an
example Germans prefer Behind The Ear aids and Italians In The Ear
aids. Keeping in mind that 93% of Oticon's sales are outside Denmark, the
importance of the regular contact with subsidiaries cannot be neglected.
This happens bye-mail, phone or fax. Employees from the head office visit
subsidiaries whenever it is relevant to their project. When going abroad,
employees also visit audiologists. Contact with audiologists is important as
they are the ones who are in direct contact with the end users. Because it
is the audiologists who help hearing disabled patients ease their problems,
it is important that they know how to use Oticon's products.
While visits abroad are regular, staff transfer and stationings lasting
more than a few months are rare: "We don ~ have a lot of that, as a matter
of fact that is probably one of the things we ought to do something about. "
(Torben Petersen, May 12, 1997).
424 Chapter 12. Oticon

Contact with subsidiaries and audiologists is an ongoing process that

happens all along the projects. In the Digifocus production development
project, the chip was developed in collaboration with international
researchers. Subsidiaries contributed with knowledge on customers, and
estimated the market potential for digital hearing aids in the individual
countries. They also regularly kept Oticon's researchers and designers up
to date with new laws and product trends so that these could be integrated
into Digifocus. Contact between subsidiaries and project groups happens
on all levels. No one person or role within Oticon, except the subsidiary
managers and Oticon's CEO's, has the overall responsibility to assure that
co-operation and co-ordination are sufficient.

12.4.6 How to Become Part of the Spaghetti

Employees at Oticon all have specific functions that they perform.
Otherwise, they are free to participate in as many projects they like.
Projects can be initiated in two ways. Either top management initiates
a new project and appoints a project owner, or the employees themselves
get an idea, get approval from the management group, appoint themselves
project owners, find a project leader, and put together a group of employ-
ees suitable for the project. The latter situation is rare. Project initiatives
are usually made by top management or the management group.
Projects all have a project owner and a project leader. The role of the
project owner is to provide the funding for the project, and to outline work
plans and time schedules. The project owner acts as a spokesman of the
project towards the management group.
The role of the project leader is to carry the project from start to finish.
Project leaders constitute the group, integrate the different processes, and
manage conflicts occurring within the projects. Even though the project
owners are the formal "spokesmen" of the projects, project leaders are
informally responsible to the management group when problems occur.
The actual project recruiting takes place in an informal way. The
transparency of Oticon enables everybody to keep an eye on what projects
are coming up. Employees who are interested in a new project will
normally contact the project leader directly. The project leader needs
acceptance from the management group before hislher group can start
working. This is needed to assure that all groups are constituted of people
with the right mixture of skills.
When asking for acceptance for a group's constitution in the manage-
ment group, this will usually be granted by Oticon's project coordinator.
The role of the project coordinator is to assure that the organization fits
12.4 Case 3: What Happens if the Spaghetti Gets Cold? 425

together. He or she acts as a controller on knowledge and resources. The

project coordinator is a member of the management group, and discusses
problems with other members, including the CEOs, before making
decisions. The project coordinator has to make sure that all projects make
use of the company's existing knowledge. Likewise, she has to keep an
overview on whether important projects get the resources they need and
also that resources are divided between projects in an equitable manner.
The role of the coordinator is thus to make sure that all projects proceed,
and specifically smaller projects, which may eventually become important
over time, are not neglected while focusing on bigger projects. "Basically
budgets are not interesting. It s the project that counts, the money spent on
a project is less important than to get the project finished and the products
launched-that we attain the results we have to. Otherwise you risk having
a lot ofprojects that do not progress. We live on the product. Budget control
is a secondary issue. " (Torben Petersen, Oticon, May 12, 1997).

12.4.7 More Control than Meets the Eye

Oticon has managed to get an ISO 9000 certification by setting up
guidelines for the course of projects. Each project is divided into four
stages:

1) The idea stage,

2) The feasibility stage,
3) The introductory stage,
4) The realization stage.

For the achievement of each stage, certain checkpoints are required.

Documentation has to be made to prove that all checkpoints have been
followed.
Oticon has three professional managers, who are not members of the
projects. Their role is to make sure that work within projects is accom-
plished in compliance with rules. These rules are formal rules however, and
they only concern the ISO standards. Projects cannot continue to the next
stage before the professional managers have given their approval on the
ISO 9000 standards. "The ISO certification is perfectly feasible-Laws of
physics have to be obeyed no matter how you organize your work. " (Torben
Petersen, May 22, 1997).
Sometimes control is made by others than the professional managers.
This happens when third persons are judged to have the longest experience
and deepest knowledge on the specific subject. Employees at Oticon are
426 Chapter 12. Oticon

said to continuously update their knowledge during projects because they

talk to each other about technological and environmental trends, internally
as well as externally. Because professional managers do not participate in
the actual projects, their ability to improve their skills has been questioned
at Oticon.
Collective self regulation is the term used at Oticon to describe the
process that professional managers are excluded from. Collective self
regulation is made possible by the cross organizational working structure.
This results in the interaction between many different specialists, each
with an extensive knowledge within their special field, but all with an
understanding of other areas and for the totality. This enables them to
adjust their own work and knowledge to that ofthe rest of the organiza-
tion.
Employees have different functions at the same time. Because
employees may very well be project leaders in one project and participate
as ordinary group members in another; this means that the same people
assumes different roles. This working structure was introduced to increase
coordination between employees. However, interdisciplinary and cross
sectional working structures are not always enough to assure co-operation.
Because the role of the management group is to function as an intermedi-
ate organ between project groups and the CEOs, crises usually arise
further down than Lars Kolind's level, and thus are normally solved
without his intervention. (Torben Petersen, Oticon, May 12, 1997) But
when problems occur that can not be solved through talking, Lars Kolind
does not hesitate to intervene. an case of crisis) "He would support us a
long way down the road. He would try to solve the problem through
discussion, but if this did not work he would take over and make a
decision. " (Torben Petersen, Oticon, May 12, 1997).

12.4.8 The Transparent Office

CEO's have no offices and they work together with all other employees.
This gives them a chance to listen to what is going on in the organization,
to see how the atmosphere is, and what conflicts are under way.
The transparency of Oticon involves all levels. The openness is
underlined by the computer network which gives everybody a chance to log
onto anybody else's computer and see how their work is proceeding. All
employees have written down their personal "success criterion" on the
network, stating three to eight elements regarding, what skills they wish
to acquire, and what goals they want to attain. "When your own results are
visible to everybody you become very motivated to attain the goals you have
12.4 Case 8: What Happens if the Spaghetti Gets Cold? 427

made for yourself. Everybody can see how his / her critical success factors
contribute to me fulfilling mine." (Lars Kolind, Berlingske Tidende, April
21,1991).

12.4.9 Authority Versus Group Mentality

Oticon's titleless structure means that authority can only be attained
through the successful solving of problems: "What it s all about is not
becoming a member of the good projects but to get the key role, that s what
it s all about. I can ~ tell you how this is experienced psychologically, but of
course it gives you status to become a project leader-we don ~ mind
status-and it gives you a professional status to be the one who is always
offeredjobs." (Torben Petersen, Oticon, May 12,1997).
Employees are dependent on recognition to be appointed to new
projects. Project members with several project successes are popular when
new groups are formed. So communication is not only about finding
information, but also about establishing good connections to project
leaders. Good social relations are not equally obtained by all employees.
Nevertheless, Oticon' s working structures are said to create a certain level
of fairness: "When we work in projects we immediately notice who is
contributing something. Especially in groups "star people" can not promote
themselves as they can at parties. " "You do not have to be eloquent on that
exact day when you have an appointment with the CEO, because he is there
all the time. SO it is easy to see who is working hard and who is not. Also,
in small groups people don f have to be shy about saying what they mean. "
(Torben Petersen, Oticon, May 12, 1997).
To utilize all employee knowledge is made use of, employees are all
assigned to a personnel manager. Personnel managers are each responsible
for 10-15 employees. Employees are free to find themselves another
personnel manager if they do not get along with the one that has been
assigned to them. Part of the role of the personnel manager is to advise and
support individuals, take initiative for job changes, and enable further
education. Personnel managers carry out salary negotiations with each
employee.

12.4.10 Enthusiastic Salaries

Employees are paid on the basis of a negotiated fixed salary based on
employee's educational and professional backgrounds. Pay rises are based
428 Chapter 12. Oticon

on work results, enthusiasm and motivation. (Lars Kolind, Stiftstidende,

April 27, 1997).
Salary negotiations take place once a year. Before the actual negotia-
tions, personnel managers talk to the individual employee's colleagues and
project leaders about how she has performed. During the salary negotia-
tions, personnel managers discuss with the employee what she has done
well or should work harder on. On that basis, it is decided whether the
employee will be recommended for a salary raise to the management group.
There are three types of possible pay rises: 1) Ordinary pay rises given
every year, 2) Pay rises given to people who have obtained good work
results, been enthusiastic and made great efforts and 3) Bonuses given as
an acknowledgment to people who have made extraordinary work efforts.
General pay rises are not considered and gratuities are rare (Torben
Petersen, May 12, 1997). "I think employees regard payments as equitable.
But first of all you need to remember that they (pay levels) are not public.
But generally I think we hit the mark all right. And that is also what salary
negotiations are all about-to discuss with employees what they are doing
well and what they should be working harder on." (Torben Petersen,
Oticon, May 12, 1997).
Oticon has flexible working hours and employees do not get paid
overtime hours. But you can still find employees working on Sundays:
"Sometimes I find people s working morale frightening! It sometimes tends
to be a little extreme (how much they work). I sometimes worry that all that
work will wear them out. " (Torben Petersen, Oticon, May 12, 1997).

12.4.11 It Takes More than Success to Prevent

Failure
Oticon has received a lot of attention from management experts and
journalists from all over the world. The focus has been: Oticon's economic
success, Lars Kolind's untraditional management style and the very
concept of being a knowledge based employee. ''And of course there is
nothing as wonderful than to read in the paper that you are a part of a
success-it releases a lot of positive energy." (Torben Petersen, Oticon, May
12,1997).
Oticon's economic results and the number of new product launches have
been remarkable. But according to Torben Petersen, the energy would not
disappear with one or two major product failures. Oticon has already had
failures, but that did not prevent the company from continuing its way of
working: "We have had both success and failure. We have had failures too,
we just don't talk that much about them, and we normally manage to stop
12.4 Case 8: What Happens if the Spaghetti Gets Cold? 429

them in time. That is, when we have noticed that there was something that
didn't ~ work we have stopped it. And that is one of the most important
things: The ability to experiment, to take the risk. The ability to dare try
something new. That is probably one of the major differences-that we have
room for that. But of course, our figures are positive now which makes it a
little easier. .. " (Torben Petersen, Oticon, May 12, 1997) .
In April 1997, Lars Kolind announced a slight change of strategy as
Oticon had decided to start concentrating on making company acquisitions.
It was important that making these had to be "fun."
Oticon's management and board of directors believe that the company
will need to invest money more actively in order to keep up the company's
previous economic results. (Oticon, annual report 1997) This is not only
necessary to keep investors happy but also: "If larger expansions don ~
occur, the unique type of organization and anarchy will die. " "Once you set
free the anarchy, you have to continue. These many qualified people need
funny and creative assignments. Otherwise they disappear." (Kolind,
B~rsen, April 9, 1997) .
To prevent employees from leaving the company, Oticon may have
promised itselfto the spaghetti structure. Employees at Oticon seem to like
working in an ever changing company. Some would even consider leaving
the company if, over time, it started getting more risk averse and
traditional in its organization structure. "If it started changing towards a
traditional organizational structure, then some of us would probably try
find somewhere else to work. One of the interesting things about this is that
it changes all the time, that it is not static. " (Torben Petersen, Oticon, May
12,1997).
The question then is, how Oticon would handle the situation if a new,
more tradition bound CEO took over, or if environmental trends suggested
a more traditional company structure. Would Oticon have what it takes to
stay flexible and keep the spaghetti from getting cold and would Oticon's
management have the necessary overview to prevent Oticon from turning
back into a traditional company?

12.4.12 Questions for Consideration

1. Oticon wants to avoid being hierarchical. What actions did they take? Is
it a good idea?

2. Describe Oticon's project management style.

430 Chapter 12. Omon

3. Discuss Oticon's incentive scheme. How does this fit with organizational
structure?

12.5 Case 4: Scouts are More Attractive

Lars Kolind, was originally employed at Oticon to make a traditional cold,
turnaround: an axeman responsible for trimming down the organization,
but also saving Oticon. Lars Kolind managed the turnaround in the first
year. Supported by very positive feedback from colleagues and employees,
who acknowledged his first year result-he decided to completely reorganize
the whole company. The result is well known: Oticon shook the world by
introducing the spaghetti organization.
Ten years later, Lars Kolind still has surprises: On December 3, 1997
Lars Kolind decided to retire from business as he wished to devote more
time and energy to his private life. Lars Kolind's decision to leave Oticon
was purely out of personal reasons; he leaves no sinking ship behind. Based
on a forecast of a DKK 1.4 billion revenue, a profit ofDKK 205-210 million
before tax is expected for 1997.
Lars Kolind had started playing with the idea of leaving Oticon in
October 1997, after a long 27 day trip abroad, in countries such as New
Zealand, North and South America and Japan. He thought carefully about
the idea, and finally made his decision in late November 1997: "During the
last ten years I have spend so incredibly many hours working for Oticon
that I think that I have worked for almost twenty years. I have also traveled
almost 100 days a year. That is difficult to combine with a wife and four
children between eight and eighteen. I think the price has been very
high-but it has been worth paying. But I must say that it will not be worth
paying for the next ten years. " (Berlingske Tidende, December 4, 1997).
The stock market's response to Kolind's decision was unambiguous: a
vertical dive in the price of shares from 340 to DKK 305, meaning a total
drop in the company's market value ofDKK 571 million.
Whereas most of Lars Kolind's colleagues have shown great sympathy
with his decision, a great number of others find his withdrawal too sudden.
But, Lars Kolind argues: "There are lots ofexamples ofdirectors who reduce
working hours and start playing more golf during the last ten years of their
CEO tenure... But I don't think that is good enough for Oticon. A CEO of
such a dynamic firm must drive with a 120 km an hour-also around the
corners. " (Berlingske Tidende, December 7, 1997).
Lars Kolind became a multimillionaire in 1995, when Oticon was
introduced at the Copenhagen Stock Exchange. Lars Kolind is said to have
12.5 Case 4: Scouts are More Attractive 431

more than DKK 100 million-and is thereby in a very fortunate position to

leave his company and do as he pleases.
Besides spending more time with his family, Lars Kolind wishes to be
more engaged with things that he believes are "worth fighting for and that
are essential." He will devote more time and energy to the scout movement
and to societal and church matters-particularly societal matters. He also
wishes to become a better chairman on the various boards on which he
serves.
Lars Kolind's resignation was effective from May 12, 1998 when the
hearing aid manufacturing company holds its regular stockholders
meeting. His successor has already been appointed. Niels Jacobsen who
has worked as Executive Vice President at Oticon since 1992 will be the
new CEO. He has been Lars Kolin's racing partner for six years. Niels
Jacobsen, who holds a M.Sc. in economics, is ten years younger than
Kolind. Although Niels Jacobsen will not be solely responsible for Oticon,
as a second director will be appointed also, it is still interesting to see that
Oticon has chosen a director with a personality which is clearly very
different from Lars Kolind's.
Niels Jacobsen has been described as something as rare as an "intellec-
tual economist, with more than just numbers on his brains" (Bf/Jrsen,
December 4, 1997). Probably the largest difference however is that Niels
Jacobsen is much less of an extrovert than Lars Kolind. Niels Jacobsen has
felt comfortable in Lars Kolind's' shadow, and has no intension of changing
that attitude: "Lars Kolind has on several occasions held a high profile in
the press, but I will only perform publicly to the amount that is good and
appropriate for the business we are driving" (Politiken, December 4, 1997)
and "I will be keeping a lower profile than Lars Kolind, but the strategy
remains unchanged. " (Jyllands Posten, December 4, 1997).
Lars Kolind in turn describes his successor with the following: "He is
very goal-oriented and he knows what he wants. But that also means that
he is predictable so that you know exactly where he is heading. That is both
positive but also critical in that he is very hard to budge." (Berlingske
Tidende, December 10, 1997).
It will be interstingto see whether Niels Jacobsen's perception of Oticon
is in line with that of the board of directors and that of the employees, or
whether they will all have to work haI'd in order to budge their new CEO
to compromise his and their ideas of what Oticon should be.
Either way, Niels Jacobsen does not plan to make any radical changes
in the company. Niels Jacobsen describes his own management style: "I
have a very straight forward and direct leadership style with a consistent
message. I find it important that the direction and the values are not
changing or varying, but that especially my closest employees know my
attitudes towards most problems in the company. I find it important to say
432 Chapter 12. Oticon

things as they are. It is so much easier, and problems can be solved

immediately in that way. "
At this point however, only time will tell whether this is a management
style that matches Oticon's corporate strategy in the future.

12.5.1 Questions for Consideration

1. How will the change in management change Oticon?

2. Discuss situational misfits.

3 Is the stock market reaction appropriate?

4. Which management style would be most inappropriate?

5. Discuss the development of Oticon in the context ofa company's life cycle
Appendix 1 433

Appendix 1
Table 12.1. Oticon's Subsidiary Companies

Name Country Date of establishment

OticonAB Sweden October 2,1944
Oticon Netherlands RY. Holland May 29, 1964
Oticon Inc. USA September 30, 1965
Oticon S.A. Switzerland July 12, 1968
OticonNS Norway October 1, 1968
OticonGmbH Germany July 1,1969
Oticon KK Japan March 7, 1973
Oticon ItaIia s.r.l. Italy June 19, 1973
Oticon New Zealand Ltd. New Zealand August 1, 1974
Oticon France S.A. France October 1, 1974
Oticon Limited Scotland February 12, 1975
Oticon Export NS Denmark April 2, 1979
Oticon Espana, S.A. Spain December 20, 1985
CLUB Hearing Instruments
NS Denmark January 1, 1992
BernafonAG Switzerland Taken over January 30,
1995

Source: Morsing (1995:236), Oticon's annual report (1996).

434 Chapter 12. Oticon

Appendix 2
Table 12.2. Oticon Holding Group -Financial ratios, key financial figures
1990-1996

MillionDKK 1996 1995 1994 1993 1992 1991 1990

Net turnover 1.087.266 940.2 750.29 661.28 538.82 476.53 455.93
R& D costs 98.991 78.209 49.415 43.022 39.528 30.422 15.822
Profit before tax 160.252 136.46 134.46 83.738 18.154 5.465 13.127
Net profit 120.286 99.184 88.316 62.305 8.822 _287 40.425
Return on sales 12.8% 12,4% 3,7% 13,1% 5,8% 1,8% 3,7%
Net capital 539.793 448.87 273.43 191.49 146.02 143.5 160.33

Source: Oticon's Annual Report (1994, 1995, 1996), B0rsprojekt (1995).

Table 12.3. Oticon's net turnover (1996) distributed on regions

Region % of net turnover

Scandinavia 11%
The rest of Westem Europe 37%
North America 21%
Asia 10%
Others 7%

Source: Oticon's annual report (1996)

Appendix 3 435

Appendix 3
Table 12.4. Financial figures and key financial figures, GN Danavox,
Oticon& Widex

1988 1989 1990 1991 1992 1993 1994 1995

Turnover
-GNDanavox 125999 130107 115353 150226 163759 180932 187986 191319
-Oticon 432756 449601 455391 476531 538816 661284 750289 940198
- Widex - - - - - 293625 259494 252703
Profit before tax
-GNDanavox 5803 10819 2142 7512 24299 16445 17838 19961
-Oticon -6208 16425 19670 2572 17684 83738 134456 136454
- Widex - - - - - 276007 243924 13394
Net profit
-GNDanavox 4466 9726 378 8767 22948 14329 15207 17422
- Oticon -9040 13719 14724 -1762 10058 62565 88452 99351
- Widex - - - - - 22800 12688 8438
Return on net
capital -
-GNDanavox 0 22 13 19 44 20 18 11
- Oticon - 13 - - 7 45 46 46
- Widex - - - - - 12 12
Share of net
capital 26 35 28 34 51 54 56 52
-GNDanavox 33 34 45 36 38 39 49 41
- Oticon - - - - - 44 47 44
- Widex
Return on sales*
-GNDanavox 1 5 0 2 7 5 - 10
- Oticon - 5 6 1 5 17 - 15
- Widex - - - - - - - 13
Return on assets*
-GNDanavox 1 8 0 4 13 8 - 12
-Oticon - 5 6 1 5 17 - 21
- Widex - - - - - - - 16
Rate of turnover,
assets· - - - 103 116 119 114 112
-GNDanavox - - - 126 145 135 135 112
- Oticon - - - - - - 133 152
- Widex

*Own calculations according to Greens

Source: Greens (1994, 1996, 1997).

436 Chapter 12. Oticon

References
Bach, K. 1995. "En chip bag f6ret." Berlingske Tidende. 13/09.
Boldt, A. 1992. "Lodrette fyringer i flad struktur." Politiken, 29/8.
Boye, S. 1994. "En fomuftig fyring." Det Fri Aktuelt. 20/12.
Bruun, F. 1991. "Oticons nye ideer mf6dt med skepsis." Erhvervs-Bladet. 25/10.
Frank, S. 1997. "Nreste udfordring: Livet." Berlingske Tidende, Magasin. 7/12, 6.
Greens. 1994. B(6rsens handbog om dansk erhvervsliv. Copenhagen: Bf6rsens Forlag.
Greens. 1996. B(6rsens handbog om dansk erhvervsliv. Copenhagen: Bf6rsens Forlag.
Greens. 1997. B(6rsens handbog om dansk erhvervsliv. Copenhagen: Bf6rsens Forlag.
Hemmer, L. 1997. "Spejderarbejdet blev en skole for livet." Stiftstidende. 2714.
Holmstad, D. 1997. "Chefskifte hos Oticon chokerede aktiemarkedet." Jyllands
Posten. 4112.
Hf6jbo, F. 1993. "Gode f6rer er guld vrerd, Politiken." 10/2.
Hf6yer, P. 1995. "Lyden har filet en ny dimension." Politiken. 1219.
Kolind, L. 1993. Civil(6konomen. 8, 16-21.
Kongstad, J. 1997. "Lyt eller lad vrer'." Berlingske Tidende, Erhverv, 10/12,30.
Kragballe, S. 1997. "Spaghetti-organisation med kf6d pa." B(6rsen, 9/4, 43.
Kryger, M. 1997. "Giver konkurrenteme kamp til sidste mikrovolt." B(6rsen, 215.
Kryger, M. 1993. "Oticon direktf6r som storaktionrer." B(6rsen. 29/11.
Jensen, T.K. 1997. "Topchefens farvel udlf6ste bf6rs-nedtur." Politiken. 4112,17.
Lehmann, A. 1997. "Jeg vii lave noget, som er vresentligt." Berlingske Tidende,
Erhverv.4112, 1 and 6.
LaBarre, P. 1994. "The Dis-Organization of Oticon." Industry Week, 14,22-28.
Lange, M. 1994. "Det er ikke nok at kunne bruge en PC." Information. 6/11.
Morsing, M. 1995. Omstigning til Paradis?, Copenhagen: Handelshl/ljskolens Forlag.
Mynderup, J. 1992. "HI/lreapparatindustrien runder milliard i omsretning." B(6rsen.
30n.
Oticon. 1993. Annual Report. Hellerup: Oticon Holding AlS.
Oticon. 1994. Annual Report. Hellerup: Oticon Holding AlS.
Oticon. 1994. Otieon's Employee Vision: Cogitate Incognita-Vision and Reality.
Hellerup: Oticon Holding AlS.
Oticon. 1995. Annual Report. Hellerup: Oticon Holding AlS.
Oticon. 1995. BI/lrsprospekt. Hellerup: Oticon Holding AlS.
Oticon. 1996. Annual Report. Hellerup: Oticon Holding AlS.
Oticon. 1997. Otieon (Company Introduction Leaflet). Hellerup: Oticon Holding AlS.
Panduro; A. 1994. "En I/lredl/lvende succes gar pa bl/lrsen." Jyllands-Posten 21111.
Poulsen, P.T. 1991. "Satser pa det danske srerprreg." Berlingske Tidende. 2114.
Poulsen, P.T. 1993. Tamk det Utrenkeligt-Revolutionen i Otieon. J.H. Schultz AlS.
Roelsgaard, J.M. 1988. "Hele topledelsen forlader Oticon." B(6rsen. 20/5.
Rundle, P. 1995. "Revolutionen fra Oticon." B(6rsen. 13/9.
Sandl/le, N. 1989. "Oticon Gruppen pa vej mod bl/lrsnotering." B(6rsen. 15/9.
Sandl/le, N. 1992. "Magten til de underordnede." B(6rsen. 27/11.
References 437

Interviews:

May 12, 1997 at Oticon: Personal interview with Torben Petersen, IT-responsible
at Oticon.
April 24, 1997 at Channel 2 "Drlllmmejob": Interview with Lars Kolind
CHAPTER 13

Sample Cases and Exercises

13.1 Introduction
This chapter provides sample cases written by executive MBA students of
the Fuqua School of Business at Duke University and by Civil Engineering
students at Stanford University. These cases illustrate how to use the
Organizational Consultant program to analyze organizations as well as
how to use the Organizational Consultant's output to create written
reports. Most of the reports were written for a course professor, however,
a few of the results were later presented for the student's corporate
sponsors. (Some of the corporations actually made the recommended
changes.)
The reports illustrate not only a wide range of writing and argument
styles, but also a wide variety of organizations that have used Organiza-
tional Consultant. It is hoped that at least one ofthese written reports-of
manufacturing industries, nonprofit agencies, government agencies, service
industries and academic institutions-will help provide insights as to how
Organizational Consultant can help you make positive changes in your
organization.
For ofthe cases, we have included the authors' choices of input data in
the input-directory. We recommend that the user try to input data based
on the information provided here; however it might be an interesting task
to compare your input with the authors'. If input files are available on the
program disk, we have noted the name of the file(s) as we discuss each
case.
At the end of each case, there are questions to help focus your analysis.
Some questions act as "exercises" to help you learn certain principles or
learn special features of Organizational Consultant. We recommend that
you try to input the data as you go through each case and then consider the
questions for consideration. The cases begin by asking you to perform basic
analyses and progress to questions which require more sophisticated
sensitivity (delta) analyses.
440 Chapter 13. Sample Cases and Exercises

13.2 ABB Electromechanical Meters1

ABB is a multinational corporation with a large number of quite autono-
mous divisions or product groups. ABB in Raleigh, NC has two divisions:
the electronics group and the ABB Electromechanical Meter division. This
case focuses on the latter, i.e., a division within a local organization and
within an even larger corporation.
The focus for using Organizational Consultant need not be the entire
corporation, but can be any relatively independent unit within the corpora-
tion. ABB Electromechanical Meters is not truly independent as it obvious-
ly must meet corporate goals given to it, i.e, profits. Further, it has a
special relationship with its sister division, electronics. Nonetheless, we
can consider ABB Electromechanical Meter as the unit offocus and analy-
sis. Thus, all questions should be answered with respect to it, not ABB
overall.
This case does not have the inputs well defined and you will have to
make some reasonable guesses. If there is an input fact which is very
problematic for you, examine how sensitive the recommendations are to
your input. If the recommendations are not very sensitive, then your best
guess is fine. If the recommendations are very sensitive to your best guess,
you may want to determine what the input should be. (HINT: this analysis
is very similar to the delta analysis, but for a different purpose, namely, to
be how accurate the input answers should be.?

13.2.1 Background
The Asea Brown Boveri (ABB) Meter Plant in Raleigh, NC began as the
Westinghouse Meter Plant in 1953. The meter plant was a full service,
fully integrated operation which relied upon no outside resources, except
raw materials and a few components bought outside the company.
In February, 1989, ABB entered into ajoint venture with Westinghouse
which culminated in the sale of the meter operation to ABB in December,
1989. This led to the revision ofthe Meter plant's organizational structure
to its current form.
ABB has long held the opinion that individual product lines should live
or die on their own merit. As a result, all of their operations are organized
to maximize the independence and autonomy of product lines. This
organizational philosophy was transferred to the meter plant when ABB
assumed control of the operation. The meter plant has two dominant
product groups, electromechanical and electronic. Thus, the operation was
13.2 ABB Electromechanical Meters 441

divided into two basically independent operational groups, split along these
general boundaries. All product specific functions followed this split, with
each group having its own Product Design, Manufacturing, and Marketing.
Accounting, Human Resources, International Marketing, Purchasing,
Receiving and Shipping, Production Planning and Management Informa-
tion Systems remained independent of the product groups, with their
respective managers reporting directly to the Site General Manager, Chuck
Hodges. The product groups pay for these "general support" functions
through an overhead charge based upon some measure of their relative
utilizations of these services.
Both product groups are dependent upon each other to some extent for
services which they don't have themselves. The EM group has all of the
basic manufacturing capability (screw machines, punch presses, diecasters,
molding presses, etc.) along with the support for those operations such as
the tool room, plant services (maintenance), etc. While not bound to
purchase the services of the electromechanical group, the electronics group
typically finds it advantageous to utilize the on-site productive capacity of
the electromechanical group (with all the inherent conveniences) rather
than go to another source for many of its components. For the same reason,
the electromechanical group utilizes the electronics group for the design
and assembly of certain components. These components, though electronic
in nature, are an evolution of products which have no real relationship with
the electronics product line and hence remained with the EM group at the
split.

13.2.2 The Electromechanical Meter

Organization
The Electromechanical Meter (EM) group is headed by Enrique Santacana
and consists of six major functional groups (Manufacturing, Manufacturing
Services, Plant Services, Design Engineering, Quality Assurance, and
Marketing) under three managers. Enrique has direct profit and loss
responsibility for the EM product line, which consists primarily of single
phase and polyphase meters and mechanical demand registers.
The EM group structure is functional and fairly centralized. Its basic
goal is to design, manufacture, and sell high performance, cost effective
products to utilities and electrical cooperatives across the United States
while earning an acceptable return on investment.
442 Chapter 18. Sample Cases and Exercises

13.2.3 Manufacturing
The Manager of Manufacturing, Charles Darden, is responsible for
production, manufacturing services, and plant services. Three managers
report directly to him: Jerry Williams, Production Manager; Dave
Daugherty, Manufacturing Services Manager; and W. A. Peoples, Plant
Services Manager. The manufacturing organization has eleven production
sections utilizing over thirty individual manufacturing technologies. In
addition, manufacturing services maintains virtually all of the tooling in
the entire plant, including the electronics production area. Plant services
maintains all plant equipment and the facility itself.
Jerry Williams has eight supervisors and two manufacturing techni-
cians reporting directly to him. There are approximately 400 production
workers in the electromechanical production operation. Production
management makes decisions affecting such things as short-time produc-
tion schedules, placement of production manpower, and short-time
production overtime. Typical actions taken include employee discipline,
coordination of routine machine maintenance with production require-
ments, and coordination of production runs between sections.
Dave Daugherty has one supervisor, six manufacturing engineers, and
six industrial engineers reporting directly to him. There are three tool
designers and about twenty-five tool makers which report to the toolroom
supervisor. Manufacturing services makes decisions such as whether to
make or buy production components, when to replace expendable tooling,
how to route work through the plant, and when and how to rework worn
tooling. Typical actions include the specification of process parameters,
specifications and purchase of tooling components, and repair of wornl
broken tooling.
W. A. Peoples has one supervisor and one technician reporting directly
to him. In addition, he takes an active role in the day to day functioning of
the plant services operation. There are about fifteen maintenance
personnel who report to the maintenance supervisor. Plant services makes
decisions about how to repair damaged equipment, how to best install new
equipment, and when to perform routine (or preventative) maintenance.
Typical actions include building repair, equipment repair, and the
specification of repair parts and materials.
13.2 ABB Electromechanical Meters 443

13.2.4 Design Engineering

The Manager of Design Engineering and Quality Assurance, Rich Greb, is
responsible for the product design function, the engineering lab, the model
shop, and quality assurance. Five design engineers, three model makers,
and two quality assurance supervisors report directly to him. Four
engineering lab technicians receive their work instruction through the
design engineers. The quality assurance supervisors have about 30 people
serving under them in various capacities.
The design group makes decisions such as when to modify an existing
component or product, what the best design is for a new product, and what
critical parameters need to be maintained in the production ofthe product.
Typical actions include specifying and authorizing changes to component
design, testing product performance, and developing new product concepts.
The quality assurance group makes decisions such as when a part or
process no longer meets specifications, whether or not a new tool or
machine meets all functional requirements, and when a noncritical
parameter can deviate from design specifications. Typical actions include
rejecting a new tool because of a manufacturing defect, coordinating the
establishment of process control guidelines, and shutting down processes
and equipment which fail to meet specifications.

13.2.5 Marketing
Enrique is directly in charge of the marketing operation. There are
currently about five marketing representatives which serve various
sections of the country. Currently, the marketing reps report to both
Enrique and the Electronics product manager, though this may change in
the future. Decisions made in marketing include such things as new
product pricing, how to best establish market position, and when to
discontinue product lines. Typical actions include issuing quotes, entertain-
ing customers, and requesting/specifying product revisions.

13.2.6 Intergroup Communications

There are several dominant lines of communication within the EM group,
most of which deal with the continued effective production of the EM
product line. The dominant communication channels and the information
typically communicated are as follows:
444 Chapter 18. Sample Cases and Exercises

• Manufacturing services to plant services, equipment problems or layout

changes.
• Manufacturing services to design engineering, product design changes
and new product information.
• Plant services to manufacturing, equipment problems.
• Manufacturing services to manufacturing, major equipment problems
and new equipment startup.
• Design engineering to marketing, product design changes and new
product initiation.

13.2.7 Managerial Responsibilities

Managers within each of the functional groups are given some leeway in
planning work, evaluating subordinates, setting goals, and allocating
budgets. However, there are limits to all of the above, basically surround-
ing the need to work to a uniform business plan and operate within a fixed
set of financial constraints. Therefore, while these managers can generally
determine what to do with the people and resources at their disposal, these
resources are typically rationed at a higher level.

13.2.8 Questions for Consideration

1. Are there any situational misfits? If so, are they serious issues that
should concern management or should they simply be noted?

2. Examine your inputs on the environment. What would be the implica-

tions ofan environment which is more uncertain and more complex? Do
you think the environment will become more uncertain and more
complex in the next five years? What advice would you give to the Plant
Manager to prepare for the next five years?

3. Examine the technology inputs for the Electromechanical Meter division.

Do you think it will change in the next five years? If so, what are the
implications?
18.8 Duke University Press 445

13.3 Duke University Press3

Duke University Press (DUP) publishes journals and books, both of very
high academic quality. It is a service organization in many ways, but its
product is realized in book and journal form. AP, such, the organization is
an interesting hybrid. You will need to guess on some of the inputs.
Examine your answers carefully to see that you are comfortable with your
responses.

13.3.1 Organizational Description

Duke University Press is a not-for-profit publisher dedicated to the
publication of peer-reviewed scholarly books and journals. It published its
first book in 1922. The Press is a semiautonomous unit of Duke University;
the Press' director answers directly to the university provost, but is not
connected organizationally to any other unit of the university. However, it
is permitted to use the services of general university support departments
(e.g., payroll, purchasing, employment).
The Press is unusual among university presses in that it is largely self-
supporting. Most university presses receive approximately 10-15 percent
of their operating funds as subsidies from their universities. Duke Press
receives only 1-2 percent of its operating funds from subsidies provided by
Duke University; the remainder comes from revenues generated from the
sales of books and journals. For FY 1989-1990, the Press expenses will
total approximately $2.8 million.
The Press' currently publishes eighteen journals. Journals represent a
predictable sour~e of revenues, since subscription levels remain stable or
increase slightly over time. Journals are easy to price, because the price
elasticity of demand in each market has generally already been deter-
mined, especially for those journals that have been in existence for a
number of years. There is very little uncertainty in terms of outside compe-
tition, since most of the Press' journals are aimed at niche markets that
may be served by one or two other journals at most. Since the market each
journal covers is limited, marketing can be accomplished fairly efficiently
through journal-specific direct mail campaigns. The "selling life" of a
journal is indefinite because ofthe potential for renewals, and because each
journal is produced continuously for its respective market, it can be con-
stantly developed and refined to better meet the needs of its subscribers.
Books are more problematic. The Press publishes approximately 70 new
books per year, and continues to market books produced in previous years
446 Chapter 18. Sample Cases and Exercises

if they are still deemed salable. It is extremely difficult to predict how well
a new book will sell; this presents problems in pricing because books are
priced only to cover fixed and variable costs, not to make a profit. If a book
is predicted to sell fewer copies, it will be priced higher than a book that is
predicted to sell a greater number of copies. Because predictions cannot
always be accurate, it can be easy to lose money on any given book project.
Each book produced can have its own distinct niche (e.g., economists,
musicologists, etc.) among the general market for academic books, which
can make marketing difficult, (typically, university press books are
marketed through catalogs which contain all of a press' products). In
addition, books have variable selling lives. Some address current public
policy issues and have short (two-year) selling lives; others that are written
on more timeless topics (e.g., literary criticism) may have selling lives as
long as twenty years. There is also a great deal of competition among
university presses for the acquisition of book manuscripts.
These two distinct product types (books andjournals) have affected the
organization of the Press. Originally, the Press was set up as a functional
organization, with departments dedicated to editorial, production,
marketing, and "business," which includes finance and fulfillment (i.e.,
shipping and customer service). ("Production" refers to design, typesetting,
and printing. All typesetting and printing is performed by subcontractors,
with Press personnel gathering bids for each project and coordinating the
traffic of materials between editorial workers and the subcontractors.) A
functional structure is typical of small publishing companies. Within each
functional department at Duke Press, some personnel were dedicated to
journals and some to books, while some could float between the two. All
personnel in a particular department reported to their functional manager,
who in turn reported to the director.
Over time, due in part to the different nature of the market for books
and the market for journals, and due in part to the rapid growth of the
number of journals produced by the Press (from nine in 1985 to eighteen
in 1990) and the proportion of revenue represented by journals (approxi-
mately 50 percent of total Press revenues in FY 1989-1990), the Press
began to evolve toward a hybrid divisional/functional structure. Now
journals are organized separately from books, with a journals manager
overseeing the work of editorial, production, marketing, and customer
service personnel that work exclusively on journals. The journals manager
reports to the director. Mid-level books managers in the functional areas
of editorial, production, and marketing report both to the director and to
an associate director/editorial director. "Business" is still a separate
functional department serving both divisions, although it continues to
encompass some dedicated book personnel.
Although the Press has begun to move toward a more divisional
structure, decision making is still moderately centralized in the director.
13.3 Duke University Press 447

Middle managers must have budgets, new projects, major purchases, and
most major personnel decisions (such as firing) approved by the director.
The books division is further constrained in that it must have new books
approved by the Press' editorial board (similar in function to the board of
trustees of a corporation). However, all middle managers have considerable
latitude in hiring and firing personnel, determining employee rewards, and
handling work exceptions. The journals manager is also attempting to gain
more power in setting his division's budget, because the Press often
borrows funds from the journals division (which is usually in the black) to
support the book division (which is often in the red).
Although major decision making responsibility is centralized in the
director, he is rarely involved in gathering information for decision making
purposes or in executing decisions himsel£ This is because the middle
managers are attempting to reinforce the newly emerging structure by
minimizing contact between the personnel beneath them and upper
management. Only three levels separate the lowest-level employee from
the director, so on occasion the director will directly ask a lower-level
employee to perform a task. The middle managers are beginning to insist
that he approach them first so they can have control over who is given the
assignment and how they are evaluated.
Because the Press is part of Duke University and thus must comply
with the university's personnel policies, all Press personnel (including
upper management) have written job descriptions. However, those
descriptions are largely ignored, and personnel are free to develop in other
directions. The journals manager is particularly lenient in allowing his
subordinates to pursue unusual functional combinations-for example, one
journal editor is also allowed to develop marketing campaigns for her
journal because she has several years of marketing experience. This
latitude is possible because Press specific rules and procedures are
generally nonexistent. As long as managers and personnel comply with
general university requirements (i.e., regulations for firing personnel), they
are allowed considerable individual latitude.
Most of the technology involved in publishing can be characterized as
engineering technology-fairly routine, but with a large number of
analyzable exceptions. For example, although all editors basically follow
the grammatical rules of the Chicago Manual of Style, they may apply the
rules slightly differently to each article or book because of the subject
matter, the writing ability of the author, and the particular market they
are editing for. In marketing, although there are generally accepted
methods of promoting books andjournals, each campaign and each catalog
description must be written for a particular market. In design, the only
rules are that the product be in book or journal format (e.g., with a cover,
table of contents, text pages, etc.) and that it be readable. Customer service
448 Chapter 13. Sample Cases and Exercises

is slightly more rigid, but only because most orders for a particular journal
or book are entered into the computer system the same way (the computer
prompts the data entry clerk for information). But even in customer service
there is some latitude in deciding, for example, whether to allow a
customer to pay for a journal subscription at a "personal" rate when they
are actually paying with an institutional check. It should be noted that
there is no real difference in technology between producing books and
producing journals.
Duke Press employs approximately 50 full-time equivalent employees,
which classifies it as a medium-size university press. Most of the Press'
employees have bachelor's degrees, and approximately 15 percent also have
advanced degrees. This is typical of the publishing industry in general,
which can be very selective in hiring personnel because of the large pool of
qualified applicants that want to work in the industry. For example, a
recent job search by the Press for an editorial assistant (a low-paying
position) yielded over sixty applicants. The candidate that was offered the
job had a Ph.D. in English. (This overabundance of qualified applicants
also explains why publishing salaries tend to be low, although the
educational level of publishing personnel tends to be high.) Because the
Press has a small number of personnel producing a large number of
products, there are almost as many job titles as there are employees.
Most ofthe Press employees work at a central site located on the East
Campus of Duke University. Because of its expanded staff size, most ofthe
journals department has moved to a separate office on Broad Street. (The
journals manager maintains offices both at the central office and at the
Broad Street site.) Three employees work in the Press's warehouse, which
is located on Pettigrew Street. A few editorial staffwho are dedicated to a
particular journal work in academic departments in the university; usually
these are the departments that began publishing the journal before
handing control over to the Press. Although many members of the Press
are scattered, they are all located within a two-mile radius. Communica-
tion is easily achieved via telephone, interoffice mail, or personal visits.
Although there can be considerable environmental uncertainty in terms
of customer demand and competitor actions, the Press faces little uncer-
tainty from suppliers. For most books and journals, prices from printers,
typesetters, and other subcontractors are negotiated in advance. (Printers
are responsible for obtaining paper and other supplies, and must absorb
the cost of fluctuations in those markets.) The Press does not have to
contend with labor unions, and it faces only those government regulations
that apply to Duke University as a whole (e.g., regarding hiring and firing).
13.4 GTE Government Systems Mobile Subscriber Equipment Division 449

13.3.2 Questions for Consideration

1. Are there any situational misfits? How serious are they and do they
require management to re-think the situation before considering the best
organization for Duke University Press (DUP)?

2. The case discusses both functional and divisional aspects ofpublishing.

Intuitively, do you think that a divisional configuration ofJournals and
books would be appropriate? Did Organizational Consultant give you
that recommendation? Ifnot, you may want to do a delta analysis on the
size of DUP. Should it operate as if it were a larger organization, how
would you support such a recommendation?

3. There are numerous functions in publishing. Do you think a functional

configuration would be appropriate?

4. There is some speculation that the publishing world will become

electronic and paperless. How should DUP prepare itselffor this NEW
world?

13.4 GTE Government Systems Mobile

Subscriber Equipment Division4
The government systems division is going through a good deal of change as
the government changes the way it does its business and is cutting its
expenditures. There are many questions around the issue of what the
organization should do and how it should structure itselffor the future.

13.4.1 Background and Issues

GTE is a multinational corporation renowned for leadership and innovation
in its three core businesses: telecommunications, lighting, and precision
materials. Boasting in excess of $17 billion in annual sales, GTE employs
more than 158,000 employees world-wide.
Under this corporate umbrella stands GTE Government Systems-a
pioneer in the advancement of Command, Control, Communications,
Computers and Intelligence in the Federal Government. From its earliest
work with radar and countermeasures to its contribution to today's tactical
450 Chapter 13. Sample Cases and Exercises

telephone and satellite communications, GTE Government Systems has

always been an innovator. GTE Government Systems is divided into three
sectors, representing ten divisions. Each division represents a major
product line, or a collection of related products, and the expertise to develop
these products. The Mobile Subscriber Equipment Division (MSED) was
formed in 1985 when GTE Government Systems won the largest U.S. Army
contract ever awarded: development of a multi-billion dollar mobile
telecommunications system. This system operates much like today's cellu-
lar telephone system, with the exception that there are no fixed antennae
sites nor permanent buildings to host the cellular switching centers.
Although the division, based in Taunton, Massachusetts, is fairly young,
most of its personnel and all of its policies and procedures were taken from
the nearby, thirty-three year old Communications Systems Division. Thus,
the division could take advantage of a mature, well-exercised development
process, and an experienced management. When the division expanded in
1987 by opening remote facilities in Research Triangle Park, North
Carolina, New Jersey, and Alabama, experienced management from within
the division was transferred to these newly established locations.
The contract which established MSED was won during the Reagan era,
when the Department of Defence (DOD) budgets were flush with money
and contract award amounts staggered the imagination. Also, there did not
seem to be an end to this era, and thus future contracts would be assured,
as long as adequate performance was performed on current projects. It was
with this spirit that MSED began its scheduled ten year development
program. During the last few years, several smaller related programs were
spun off from the parent contract. These related programs were based on
technology discovered as a by-product of developing the hardware and
software. Examples of these products include tactical facsimile terminals
for sending maps and battle plans, and quick-erecting antennae mast
systems. However, with minor exception, no additional dollars have been
spent on research and development of new products, partially because no
additional staff was available to perform this research, and partially
because the business plan for this division, clearly specified in a 150 page
bound document, is to support the government contract for which it is
named, as opposed to furthering technology or chasing after new contracts.
The environment in which MSED now finds itself is vastly different
than what it once was. As the division winds down from its one major ten-
year contract, it struggles to defend its turf of products related to military
tactical communications. With no new development close to maturity and
no new products to offer, marketing people find it difficult to sell the
division's capabilities. Just a few years ago, a potential for another major
contract to take up the slack was almost assured. Today's DOD budget has
18.4 GTE Government Systems Mobile Subscriber Equipment Division 451

caused every defense contractor to aggressively seek what is left of the

remaining contract allocations.
The general manager of the division, Bernie Resnick, appreciates the
predicament which the division is currently in and has started making
some changes to the functional structure to accommodate the uncertain
future. One major change is the addition of a function-Business Acquisi-
tion and Development. Fred Susi, a GTE employee of over 20 years, was
given the responsibility of heading this function to develop new business
leads. Once a lead was established, as long as the prospect fits within the
business plan, the Special Projects Office would be responsible for
researching the technology, providing subject matter expertise to the
marketing staff under Mr. Susi, and developing hardware and software
prototypes. If a business lead turned into a contract, then the engineering
function, in which the majority of the over 1000 employees in the division
work, would take over the project, freeing the marketing staff to further
pursue new business.
To date, this strategy has not generated any new leads. Although the
marketing staff was prepared for a tough market, they are still thinking
too big for today's shrinking military. Talk of cutbacks and layoffs are
rampant throughout the organization, and threats of closing down the
remote facilities are becoming more frequent.

13.4.2 Analysis
One of the reasons for the rumors is that employees do not see a well-
defined future. They do not see new products being developed. They do not
see the state of the art in developing military software systems being
advanced. And they are correct in their perception. The division's current
1992 independent research budget is $5.4 million, or 2 percent of the
division's yearly sales income. This is low even by defense contractor
standards. The marketing budget, in contrast, is approximately 30 percent.
This is indicative of an organization spending great sums of money
defending its current position, and ignoring new product development. This
will cause problems not just in 1992, but in 1993 and beyond when the
marketing staff still have no new products to sell, and no new technology
to promote.
Clearly the division needs to break out of this defensive strategy if it
wishes to have any future business. However, it should not, and it can not,
give up its core business. An analyzer strategy is well-suited to provide the
flexibility necessary to open up new markets, while still maintaining
stability in its main product. To do this, the division will need to undergo
some changes.
452 Chapter 18. Sample Cases and Exercises

Currently, the division is dedicated to the development of a single

product line. MSED is primarily an engineering organization. A large
majority ofthe 1081 employees are professionals, with several engineers
holding post graduate degrees. Frequent management reviews, from
personal performance reviews, to department reviews and project reviews
are ingrained into every engineers work habits. Functional managers are
actively involved in all decision making, including strategic and long-range
decisions, as well as operating procedures on a month-to-month basis.
Decisions regarding personnel, especially in terms of promotions, and
assignment to projects, are made by line supervisors, with final approval
resting with upper management.
This part of the organization, with its medium levels of centralization,
formalization, and complexity should be maintained. An adaptive, less
formalized, less centralized organization should be developed within the
division to provide the flexibility which stimulates innovation.

13.4.3 Questions for Consideration

1. Are there any situational misfits? What should management do about
them, if anything?

2. Do a delta analysis where Government Systems has a prospector

strategy? What are the implications for change? Should Government
Systems undertake a prospector strategy?

3. Examine carefully the current formalization and the recommended

formalization. Is there a difference? Explain how Government Systems
evolved to its current level of formalization.

4. Examine the centralization for Government Systems.

5. Prepare a managerial plan for Government Systems

13.5 Bluestone Group, Inc. 5

The case was written as part of a Stanford University Civil Engineering

course entitled "Organization and Management of Human Resources." The
primary purpose of the case (which was a course term project) is to provide
18.5 Bluestone Group, Inc. 453

the students with an opportunity to apply the theories and design tools
presented in the organizational theory class by observing and analyzing
how a real project organization in the construction/engineering industry is
set up and managed.
The case can be divided into the following two parts: Part I is a survey
of the Bluestone organization overall. Information about the company was
obtained from published sources as well as several interviews with
Bluestone personnel. These data were then entered into Organizational
Consultant to generate a recommended organizational structure for
Bluestone.
Part II deals with the structure of a smaller organization within
Bluestone, and the students chose to study the field engineering unit in
Bluestone's Alkylation Plant Modernization Project at Oilco Baylands
Refinery. The configuration of the unit is analyzed as well the inter-
relationships both within and outside the unit of its members using the
definitions laid down in this book. From this analysis the students were
able to suggest the optimal organizational structure for this unit.
Course Professor Raymond E. Levitt comments that "for Part II of this
case, the students did not apply Organizational Consultant directly,
although you can clearly see the influence of Organizational Consultant in
their analysis. Although students may not routinely use it to solve cases,
they can assimilate many of the important ideas and approaches inherent
in the Organizational Consultant knowledge base by going through the
process with Organizational Consultant only once."
The case is very comprehensive in its analysis and application of
Organizational Consultant.

13.5.1 Background
Bluestone is a Los Angeles-based professional engineering and construction
organization whose stated purpose is to provide premier technical
management and directly related services to develop, manage, engineer,
build, and operate installations for our customers worldwide. It was
founded in 1898 as a builder of railroads. Since then its services have
grown to encompass diverse technical fields in many geographical
locations. Bluestone currently consists of companies organized along three
different fronts with business lines (e.g. petrochemical and power),
functional organizations (e.g. construction and financing), and geographical
offices (e.g. Los Angeles and Houston) (Figure 13.1). All ofthese privately
held companies are brought under the aegis of Bluestone Group, Inc., with
Richard Bluestone, a fourth-generation representative of the founding
454 Chapter 13. Sample Cases and Exercises

family, at the helm. In 1993 Bluestone worked on 173 major projects in

twenty-one countries throughout the world

Bluestone Corporation

Functional
OrganrliOns
Geographical Business Lines
Offices

I
:=====~ ---T- - Bluestone Petroleum
l- I Bluestone Construction
l-
H
Chemical & Industrial

I U
.... -Fl I
Bluestone Power Engineering &
Reid Supervision I
iI ~========~
-.J
r- '-1 I
Blueslone Civil

~======::::::;
H Safety
I
I
1 ---1
r- =,
BlUestone National
I H Labor Relalion
I
Iu-d--'----;-----'
~======~
Dallas
I
~ _9 Mining & Melals
I Y Procurement
I
H
I
o L__~ Bluestone Environmental J Purplestone Construction
I
H Bluestone Construction
International I
Formal Structure
H Bluestone Enterprises
I
Working Setup
I y Research &
Development I
Matrix Relation
I OileD 8aylands
Refinery Project

Figure 13.1. Bluestone Group, Inc. Organization chart

For this report members of Bluestone's Alkylation Plant Modernization

Project at the Oilco Baylands Refinery were interviewed. They are all
Bluestone Construction Company personnel on assignment to this project.
They were asked to give us their perspectives on not only the project itself
but the Bluestone organization overall as well.

Those interviewed were:

• Site Manager
• Project Field Engineer
• ALKY Area Engineer
• Piping Engineer.
18.5 Bluestone Group, Inc. 455

13.5.2 Part I
Several methods were used to collect input data for Organizational
Consultant. Company brochures and informal conversations with people
both inside and outside of Bluestone provided general information about
the organization. The actual Organizational Consultant questions were
also incorporated into a questionnaire and distributed to the interview-
ees-trusting that the employees themselves can offer better insights in
areas such as operating procedures and management style. Interestingly,
while these questionnaires provide invaluable information regarding
Bluestone, there were some definite inconsistencies which seemed to reflect
the respondents' perspective their niche in the organization. For example,
the project engineers tended to believe more that they had clearly defined
tasks to follow than the managers did. Also, although there are indications
that some of Bluestone's business lines may actually embrace a rather
functionally-oriented philosophy, most of the interviewees believed
Bluestone overall to be a matrix structure. It is a case of projecting their
immediate environment in Bluestone Construction and the Oilco site onto
the larger Bluestone whole. The input data then are based firstly on the
questionnaire results. Should there be glaring inconsistencies among the
answers, the most likely answer was selected by evaluating other sources
of information, and the certainty factors were adjusted accordingly. In
some cases, such as the size of the company, a published value was used in
favor of the questionnaire results.

13.5.2.1 Current Organization

Bluestone is a large organization which has a high level of education

among its 20,000 employees. Its management has a medium preference for
microinvolvement. This is demonstrated by the fact that the top manage-
ment:

• Prefers to let some decisions be made by other managers,

• Has both a short-time and long-term horizon when making decisions,
• Is risk neutral,
• Prefers to use both motivation and control to coordinate the activities.

The organization currently executes an analyzer with innovation strategy.

This strategy of seeking new opportunities but also maintaining its
profitable position is appropriate when there is:
456 Chapter 13. Sample Cases and Exercises

• Medium requirement for product innovation,

• Medium-routine technology-with some flexibility allowed,
• Product diversity,
• Medium capital investment-constraining some capabilities but leaving
others adjustable,
• Strong concern for high quality.

Bluestone meets all of the above criteria. Further, the top management
shows a preference for a medium level of microinvolvement. This can be
obtained via control over current operations. Product innovation should be
less controlled, however. This again calls for an analyzer with innovation
strategy. The organization's characteristics are determined to be as follows:

• Organizational complexity is high,

• Horizontal differentiation is high,
• Vertical differentiation is high,
• Spatial differentiation is high,
• Centralization is medium,
• Formalization is medium.

Based on the above analysis, an Organizational Consultant analysis can

be performed.

13.5.3 Part II
The Bluestone Alkylation Plant Modernization Project was observed for the
second part of this report. The goal of this project is to construct a new
alkylation plant in the existing Oilco Baylands Refinery to help produce
cleaner automobile fuel which comply with the new federal regulations to
be in effect by 1996. Bluestone is the prime contractor in this project; Oilco,
the owner; and Peterson, the designer. The project is executed under a cost
reimbursable contract valued at $253 million. It started on October 4,
1993, and is scheduled to be completed by September 21, 1995.
The Bluestone project staffis directed by a two-boss site manager who
reports to both Bluestone Petrochemical and Bluestone Construction
Company (Figure 13.1). The organization of the project has a staff of 63.
The Site Manager supervises the field division and the service division
through an Assistant Site Manager and a Project Service Superintendent
respectively. He supervises the rest of the staff directly. Under the
18.5 Bluestone Group, Inc. 457

assistant site manager is a Project Field Engineer who supervises the field
engineering unit.
The specific unit whose structure and coordination we chose to study
was this field engineering unit and its nineteen members.

13.5.3.1 Goals and Objectives of the Dominant Coalition

The overall goal ofthe Alkylation Plant Modernization Project is to achieve

the points defined in the cost reimbursable contract for a successful project,
such as completing the project 15 percent under the budget in order to
receive a one million dollar incentive. To achieve this, Bluestone tries to
build a harmonious and trusting working relationship with the owner and
the designer following the general guidelines set forth in Cost-Trust
Relationship. At the field engineering level this strategy translates into
developing a highly-qualified team to guide and direct the project to the
highest quality at the lowest cost and promoting this team spirit not only
amongst the engineers but also to the owner and the designer.

13.5.3.2. Environment

There are several organizations external to the engineering unit with

which the unit must interact (Figure 13.2). The engineers identified their
interactions with the owner, the designer and the superintendents as the
key relationships:

• Owner and designer: The engineering unit maintains very close rela-
tionship with these two groups of people. As is consistent with the team
building concept the members of all three groups share the same office
space and communicate with each other freely and constantly.
• Superintendents: This is the group of people the engineering unit come
in contact the most. The engineers rely on the superintendents to make
sure that designs are constructed out properly, and feedbacks from the
superintendents allow the engineers to act as the intermediary between
the field and the owner and designer. Problems between field supervi-
sion and engineering are discussed and resolved at weekly meetings.

The greatest source of uncertainty is the design changes requested by the

owner and designer. Since the drawings are not 100 percent completed, the
uncertainty is high. However, because those with the owner, the designer,
and the superintendents are the engineering unit's only major external
458 Chapter 13. Sample Cases and Exercises

interactions, its environmental complexity is not high and the equivocality

low.

Bluestone Oilco Site

Key Relations Mechanical, Welding, and Field Engineers

legend:
Relations are shared by the wto areas.

Figure 13.2. Environment which surrounds the Engineering Unit

13.5.3.3 Work-flow and Technologies Used

The most important job for the engineering unit is to carry forward the
installation on schedule while minimizing rework. The unit holds a weekly
meeting to check on the progress in all engineering concerns and to plan
out in detail the work to be done in the next three weeks. Each individual
has his own field oftechnical specialty (electricity, piping, civil, mechanical,
etc.) for which he is responsible. The interdependence in the unit among
different specialties is not overly strong. Rather, contacts between an
engineer and an outside specialist in his field occur more frequently.
However, an engineer who needs minor assistance from any other members
of the unit can simply ask for it directly. When more substantial assistance
is required, the two leads in the unit coordinate this matter (Figure 13.2).
Bluestone uses a formal service grade system as well. An entry-level
engineer usually starts from Level 22 or 23, and his or her grade advances
based on seniority and performance. Bluestone also encourages its
employees to seek out further training and education if they so desire.
Automation is not prevalent inside the engineering unit. E-mail is
implemented widely, and there is a project-wide Procurement Tracking
System (PTS) which the engineers use to place purchase orders. However,
because of the nature of the project, face to face communication is still the
dominant means of information flow. The engineers also indicated that
paper rather than electronic media is the preferred method of documenta-
18.5 Bluestone Group, Inc. 459

tion from both the legal and the ease of use points of view. The piping
engineers do have a computer database called Piping which details the
layout, length, size, and materials of all the piping connections. The
electrical engineers use a similar system called Setroute.

13.5.3.4 Existing Organizational Structure

The engineering unit is headed by a project field engineer, who is delegated

much authority by his direct superior, the assistant site manager. The
project field engineer can make purchasing decisions up to about $500,000
without his superior's prior approval. He in turn also prefers to delegate his
authority to his subordinates, indicating a low preference for micro-
involvement. His main task is planning and meeting a long-term schedule
along with the owner and the designer.
Underneath him there is no distinct official hierarchical structure.
Instead of having formal managers between the Project Engineer and his
engineers, the unit has two senior engineers acting as area leads during
the bulk installation stage (Figure 13.2). The situation-specific nature of
this arrangement exhibits the trait of an ad hoc configuration. However, it
is also a divisional configuration since the engineering responsibility is
divided into the ALKY and TAME Areas at this stage. Each area has a full
complement of civil, piping, and electrical engineers. The line between
areas is not clearly drawn, however. Each lead is free to assist or borrow
personnel from the other area. Each area lead can also make decisions
affecting up to about $5,000 based on only his judgment. This amount is
not stated in the regulations but rather comes about as a tacit understand-
ing among the project members. Another important role of the leads is to
instill the company culture in younger engineers.
Clearly a large portion of the project field engineer's authority is trans-
ferred to his subordinates. Therefore, we can conclude that the engineering
unit is very decentralized. However, the project field engineer and the area
leads still prefer to be informed the decisions that their subordinates or
unit members make, and bucking the chain of command is generally
frowned upon.
Currently the project is in the transition period of swinging, from bulk
installation to turnover (start-up) mode. Consequently the unit's configura-
tion is now changing from the area engineering configuration to a func-
tional one. The two area leads become turnover coordinators at this stage.
Some engineers may be experiencing slight confusion regarding their
responsibility as a result of the change, but overall the transition can be
considered as smooth. The structure of and the atmosphere in the
460 Chapter 18. Sample Cases and Exercises

engineering unit remain quite informal. There are comparatively few

written rules. Instead, the staff have a tacit understanding of how some
things should be done which comes about as part of the company culture.
Incentives are based on Bluestone's overall performance as stipulated
in the contract and can broadly categorized into three award types based
on safety, quality, and on-time and on-budget delivery. Any reward is
distributed to all project staff members.
We may now summarize and conclude that the engineering unit
possesses the following characteristics:

• Low management preference for microinvolvement-Iow centralization,

• Proactive decision making-medium formalization,
• Two informal groups with two leads in the unit allocated to two
areas-ad hoc and divisional configuration,
• Medium media richness,
• Staff of three to five under each lead (total of eighteen underneath the
manager) and two informal vertical layers-short span of control and
small number of vertical levels,
• Tacit understanding rather than written rules and standards,
• Many face to face meetings and interactions among individuals,
• Informal and free communication,
• Freedom to act first based on one's own judgment and report later,
• Reward distribution based on overall organization's performance.

13.5.4 Questions for Consideration

1. Are there situational misfits?

2. Which organizational structure would be appropriate for Part I?

3. Which organizational structure would be appropriate for Part II?

4. Compare the current and recommended structure.

References 461

Notes
1. Thanks go to David R.Thomason, who wrote this case.

2. Input decisions of the authors are provided in the file named "Ahh.ocd"
which you may retrieve in the Organizational Consultant program.

3. Thanks go to Kathy Brodeur, who wrote this case. To view author's

input, access the file named "DuPress.ocd" from the Organizational Consul-
tant program disk.

4. Thanks go to Louis Grilli, who wrote this case. To examine the author's
input file, access the "GteGovsy.ocd".

5. Thanks go to Stanford University Civil Engineering students Hyun-Chan

Cho, Hajime Kitamura and Ting-Hsuan Wu, who wrote this case.

References
Bluestone Group, Inc. 1993. Bluestone Annual Report and other company
brochures. Bluestone: 1992 and 1993.
Construction Industry Institute. 1993. Cost-Trust Relationship. Austin,
Texas:The University of Texas at Austin.
Subject Index

ABB ..................................................... 63, 68, 440

Ad hoc .......... 30,42,45,64,65,70,71,73,74,80,102,108,150,155,159,
190,192,193,196,197,199,202,207,230,235,238,257,
259,260,287,292,293,323-325,334,338,390,459
Adaptability ........................................... 80, 89, 308, 318
Adaptable ............................................ 89, 130, 313, 412
Adaptive ........................................ 239, 314, 316, 334, 452
Adhocracy ........................... 55,125-127,130,389,390,392,393
Administration ............. 50,51,57,162,163,209,344,366,401,402,409
Administrative .............. 38-40,45, 109, 145-147, 162-164,206,208-210,
212,239,256,274,277,344,380
Agency ............................................... 51,55,145,400
Agents ......................................................... 172
Alcazar ......................................................... 274
Alignment ................................... 17,138,245,258,347,351
Analyzer ........ " 25,35,242,248-256,258,260-266,270,275,276,287-292,
296,298,300,301,304-306,310,311,314,315,317,318,
324,325,330,334,340-342,388,451,455,456
Ashby's Law ..................................................... 207
Aston ....................................................... 38, 162
Authority ......... 7,45,64-66,75,78,91,95,96,104, 107, 108, 121, 139, 173,
188,246,247,287,294,303,408,409,412,416,427,459
Autocratic ................................................. 89,92-95
Autonomy .......................................... 118, 119, 122, 440
Bargaining .................................................. 179,244
Bilka ........................................................... 222
Bluestone ........................................... 355, 452-458, 461
BonGout ........... 165, 166, 171, 172, 175, 176, 180, 182-187, 189,202,203,
241,242,271,284
Boundary ....................................... 3, 10, 13, 246, 247, 337
Budgeting ....................................................... 87
Budgets ........................................ 350,425, 444, 447, 450
Buffer ........................................................ 6, 173
Buffering ................................................... 203, 204
Bureaucracy ........... 45,65,66,73,74,80, 101-103, 127, 132-134, 159,214,
230,260,294,295,334,335,389,390
464 Subject Index

Capital requirement .......... 255-257,260,261,263,264,266,268,271,310,

311,316,366
Carlzon, Jan ................................................ 109, 277
Centralization ......... 13,17,18,25,26,29,30,41,75,76,78,80,89,90,99-
103, 107-109, 116, 124, 125, 127-137, 141, 154-157, 161, 162,
173,181-188,192,194-199,202,207,212-216,224-227,231,
232,238,239,243,244,250,255,259,260,262,263,265,266,
268,275,276,282-285,303-307,320,321,323-325,329,330,
334,335,338-341,360,364,385,388,389,391,409,452,456,
460
Certainty factor .......... 18, 24-30, 98, 155, 288, 319-321, 323, 349, 353, 361,
362,371,386,389
Chain-of-command hierarchy ....................................... 111
Climate ......... 1,9,13,15,17,85,86,111-120,122-138,140-143,284,286,
288-302,304-306,311,312,317-319,321,324,325,328-330,
338-341,354,360,364,375,386,387,389-391,393
Climate measure ......................................... 112, 119, 286
Coalignment .................................................... 279
Coalition ....................................................... 457
Collectivity stage ............................................. 103, 329
Communication ........... 8,69,73,80,87,104,143,149,150,175,215,232,
400,405,410,415,417,418,427,443,448,458,460
Compensation ............................................ 13, 376, 400
Competing values approach ............. 89, 110, 112, 114-116, 124, 141-143
Competition .............. 1,2,5,42- 44,48,49,55,77,79,81, 107, 123, 135,
136, 140, 146, 169, 171-173, 175, 177, 178, 180, 189,
197,201,202,205,206,210,235,241,256,261,267,
268,273,274,278,321,328,341,345,350,351,368,
375,395,396,404,405,407,409,411,412,445,446
Competitor .......................... 166, 250, 256, 313, 316, 349, 350, 448
Concern for quality ........... 255-257,261,264,266,267,269,369,384,388
Configuration ...... 13,25,31,41,42,45-50,52-71,73-75,80,85,99,101-103,
108, 109, 111, 124, 125, 127, 129, 130, 132-137, 141, 150,155,
158,159,162,166, 181, 189-199,207,219,220,224,229-232,238,
255,259,260,262,263,265,266,268,270,271,275,281-283,
285-294, 296, 298, 307, 321, 323-330, 334, 335, 338-341,354,
360,362,364,380,389,390,392,393,449,453,459,460
Conflict ........... 27,29,48,64,87,88, 102, 103, 116, 118, 121, 122, 124-129,
131-133,135,140,141,282,300,303,315,317,318,321,339,
351,375,386,387,395
Contingency fit .......... 15-18,23,30, 137, 199,284-286,303,307,320,321,
325,341,342
Contingency model .......... 13-15, 89, 112, 116, 117, 120, 124, 142, 143, 354
Contingency theory ........ 7, 14-17, 19,30,37-40,84, 111, 166, 167, 169-171,
210,271,325
Subject Index 465

Control ..... 4,7,10,17,41,45-48,70,71,73-76,78,80-82,86,87, 89-101,103,

104,106-108,110,115,124,125,127-137,145,152,154,157,
181-183,185,191-196,198,202,205,207,209-211,213-216,
224-228,231-234,237-240,242,243,246,247,250, 251, 255-258,
260-269,271,273,277,286,287,298-300,307,314,317,318,
323,328,330,336-341,362,368,371,379,385,387,389,391,
392,408,417,420,425,440,443,447-449,455,456,460
Control-flexibility ................................................ 115
Cooperation ................................ 3,42,43,67,83,87,121,172
Coordinate ......... 4,43,47,57,70,72,80,100,133,173,185,187,197,225,
229,230,275,383,455,458
Coordination ....... 3,5-7,9, 13,41,42,46,48,49,51-55,57-60,62-64,67,69,
70,72-81,100,101,103,104,124,125,127-137,141,145,
146,151,152,157,161,181,190-199,202,204,207,212,213,
215,216,220,224-226,229,232-234,238,239,242,268,270,
286,287,297,299,304,307,323-325,328-330,334,335,337-
340,389,391,401,410,411,426,442,457
Coordinators ................................................ 124, 459
Corporation .......... 10,50,55,57,67,68,182,211,221,240,242,253,254,
274,336,347-350,396,398,440,447,449
Costs ........... 6,42,44,53,70,72,77,79,80,105,139,140,179,205,206,
211,254,262,272,274,314-316,341,369,407-409,434,446
Craftshop ....................................................... 212
Creative ............................. 97,103,114,123,201,387,411,429
Creativity ............................................ 86, 119, 328, 412
Culture ............ 69,85,86, 104, 111-119, 131, 141-144, 164, 202, 203, 375,
377,379,405,410,459,460
Current organization ..................... 326,343,352,353,389,392,455
Customer ........... 8,12,41,45,46,50,54,55,58,59,62,63,70,71,74,79,
106,111,166,177,179,195,221-223,233,237,238,242,
250,272,313,321,341,369,371,374,375,378,382,391,
395, 400 401, 404, 411, 412, 423, 446-448
Customization ...................................... 52,53,57,357,371
Cybernetics ............................................. 208, 209, 277
Databases ...................................... 230, 232, 358, 359, 378
Decentralization ........ 9,14,17,18,27,28,30,35,44,45,63,75,76,80,89,
141,145,146,152,154,156,157,173,187,230,231,237,
250,282,283,303,305,306,317,334,336,338,353,390
Decentralized ............ 1,9,11,17,19-21,25,63,66,67,75,78,79,85,102,
106,108,140,141;156,161,183,187,196,231,
283,303,306,323,339,459
Decline stage .................................................... 330
Decomposability ..................................... 219-221, 223, 224
Decomposable ............................................... 219,224
Decouple ................................................... 173, 336
 466 Subject Index

Defender .... 25, 242, 248-256, 262, 266-270, 275, 276, 288, 289, 298, 300, 302-
304,306,309,312-315,317,318,333,334,338-340,342,388
Delegate ....................... 92,95,98,104,106,157,377,385,401,459
Delegation ................. 1,8,87,89-98,103,106-108,156,328,384,412
Delta analysis ....................................... 354, 440, 449, 452
Democratic ......................................... 89,92-96,107,204
Department .......... 10, 12,46,48,49,51,60,67,69,73,159, 182,211,236,
286,288,383,401,402,405,406,409,411,412,415,
422,446,448,450,452
Deregulation .................. 44, 81, 139, 165, 166, 172, 175, 206, 208, 278
Design parameter fit ......... 15,17,18,24,27,28,30,284,285,298,319-321,
323,340
Developmental Climate ....... 123, 128-130, 140,311, 312, 318, 321, 340, 387
Developmental Culture ........................................ 114,116
Diagnose ................................................. 11, 19, 347
Differentiation ........... 25, 27, 45, 68-73, 76, 85, 99, 100, 108, 127, 130, 131,
134, 138, 148, 150, 152, 156, 162-164, 172, 173, 179, 182,
186,203,213,214,227-229,232,238,242,243,250,296,
298,319,323,337,361,382,388,390,392,456
Digifocus ............................................... 419-421,424
Diversification ........................................... 147, 159,242
Diversified . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1, 242
Diversity ............ 68, 147,216,218-220,250,251,260,263,270,271,309,
310,360,364,373,389,456
Divisibility .................................. 220, 221, 223, 224, 341, 371
Divisional configuration .......... 53-61, 65, 66, 69, 75, 85, 135, 136, 195, 197,
199,207,219,229,238,263,265,275,290,
291,325,340,449,459,460
Divisionalization ......................................... 147, 163,270
Divisions ............ 6,7,46,54-59,62,68,69,74,75,78,146, 195,229,276,
291,325,336,363,380,440,446,450
Domestic ..................................... 67,68,166, 271, 274, 361
Downsizing ................................................. 146,160
Duke University Press .................................... 355, 445, 449
Dynamism ...................................................... 168
Ecology ............................................. 166, 171, 172, 209
Economies of scale .............................. 55, 63, 159, 179, 224, 263
Effective . . . . . . . .. 4, 5, 8, 13, 32, 39, 46, 52, 59, 64, 99, 111, 112, 114, 124, 134,
142,148,161,166,199,248,259,262,263,269,282,291,
301,332,333,344,352,376,384,431,441,443
Effective performance ............................................. 111
Effectiveness ........ 2, 4, 38, 46, 63, 64, 81, 110, 115, 142, 143, 150, 210, 255,
277,282,292,327,332,334,344
Efficiency ........ 2-5,7,8,13,21,29,32,38,46,48,51,52,59,61,63,64,70,
72-74,76,80,99,112,114,116,124,134,135,142,150,155,
SUbject Index 467

157,160,161,163,166,172,173,182,184,185,194,189,209,
211,217,219,225,229,230,237,139, 248,252,261-264,265,
267-269,272,282,289,291,292,298,299,327,332-334,344,
388,390,391,412,413
Elaboration of structure stage ...................................... 330
Elasticity ....................................................... 445
Electronic information system ...................................... 231
Electronic information technology ............................... 231,232
Empower .................................................... 76, 235
Entrepreneurial stage ..................................... 103, 328, 329
Environment ...... 1,3,5,7,10,11,13,15,17,20-23,26,29,30,38,44,46,53,
58,77,81,84,85,90,99,102,109-111,113,116-118,130,
135-140,143,165-198,200-210,217,220,231-233,236,241,
242,247,253,255,260,271,272,276,278,282-293,296-308,
312-325, 328-334, 338-342, 344, 350-352, 354, 360, 364, 365,
368,378,389-393,408,415,417,422,444,450,455,457,458
Environmental complexity ........ 19, 20, 22, 23, 168, 174, 176-178, 184, 186-
188,190,194-196,198,202,203,217,287,
305,309-311,315,349,364,458
Environmental equivocality .... 186, 188,283,284,298,308-312,349,365,390
Environmental hostility ................ 26, 171, 189, 197,301,365,390,391
Environmental uncertainty ........... 26,29,38,168,169,172-174,178,180,
185,188,197,200,208-210,283,284,288,
309-311,314,365,390,448
Equifinality ...................................... 18, 38, 277, 285, 286,
Equivocal ......... 17,108,147,169,174-188,190-200,202-204,207,217,219,
221,231,233,283,284,287-290,292,293,296-302,304-318,
321,324,325,330,338-341,349,364,389-392,458
Euroclass ................................................... 237,272
Expert systems ............................. 19,24, 31, 32, 34, 38, 39, 230
Exploitation .................................. 67,258,327,332-335,344
Exploration ...................... 62,210,317,327,332-338,341,342,353
Export .......................................... 67, 368, 395, 396, 433
e-mail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 150, 214, 230-232, 422, 423, 458
Fairness ............................................ 119, 122, 125, 427
Feasibility .............................................. 352,357,425
Five forces .............................................. 178-180,201
Flexibility .......... 10,42,45,53,59,89-95,99-101,103, 107, 111, 115, 130,
146,173,201,202,217,242,245,256,264,265,269,276,
301,302,317,318,324,325,329,332,358,388,390,
395,412,416,428,429,451,452,456
Flexible form ................................................. 42, 202
Flexible organization .......................... 42, 201, 202, 242, 276, 325
Formalization ............ 5, 8, 13, 18, 24-26, 39-42, 45, 63, 65, 66, 73-81, 85,
468 Subject l1Ulex

89,90,99-104,107-109,116,124,125,127-137,141,
146,148,152,154-163,173,181-199,202,207,212-216,
224-226,231,232,236-239,243,244,250,255,259,
260-268,275,276,282,285,294,299-303,307,320-325,
328-330,334-341,360,364,383,388-393,452,456,460
Formalization and control stage ..................................... 330
Formalized ................... 45, 74, 108, 114, 123, 140, 141, 146, 152, 158,
184,185,207,211,213,228,231,232,248,
262,284,302,307,328,329,401,452
Function .............. 3,12,24,44-46,49-53,60,62,64,68-71,85,103,127,
142,145,156,177,224,229,293,328,336,337,361,
374,375,384, 391,395, 400-402,413,416,423,
424,426,441,443,447,449,451
Functional configuration ......... 48-50, 52-54, 60, 63, 65, 68, 69, 74, 85, 132,
136,159,192,194,197,199,219,229,263,268,
270,275,285,288-290,307,338,339,449
Generalists ............................................. 172, 259, 298
Globalconfiguration ................................... 68,270,271,389
GN Danavox ............................................ 404, 405, 435
Group climate ........... 123, 125-128, 130, 140, 141,284, 293, 312, 318, 319,
338,386,390,391
Group culture ............................................... 114-116
Grouping .......................................... 45, 53, 54, 290, 382
Groups ......... 6,12,29,43,44,51,54,55,70,76,87,88,123,125,135,138,
199,217,272,319,350,376,378,412,416,417,420-422,424,
426,427,440,441,444,457,460
GTE ................................................... 355, 449-451
Hans Demant ................................................... 400
Headquarters ............... 45,51-53,56,58,66,67,165,361,373,382,383
Hearing aids ................ 400,402-405,407,410,411,418,419,421-424
Hierarchical ...... 5-7,51,52,59,64,69,150,232,296,382,401,406,429,459
Hierarchy ................. 6,45,48,59,62,64,65,69-72,74,76,78,82,111,
150,164,173,182,185,186,188,216,231,247,278,
279,291,294,296,372,382,389,420
High-technology ................................................. 344
Horizontal differentiation ........... 27, 45, 68-72, 85,100,134,156,182, 186,
227,229,238,296,298,337,388,390,456
Hostile environment ........................... 26, 165, 187, 189, 197,287
Hostility ................. 26,171,174-177,179-181,184,189,193,197,200,
201,203,287,301,305,330,364-366,390,391,393
Hybrid ............................. 66,247, 249,250,253,298,445,446
Hypercompetition ............................. 42,201,202,210,245,324
lATA ............................................... 7,43,44,204,205
IBM ........................................ 1, 2, 13, 160, 253, 254, 256
Subject Index 469

If-then statements .............................................. 20, 27

Incentive .......... 2, 10,13, 18,42,53, 59,64,74,76,81, 103, 104, 124, 125,
146, 181, 127-132, 134-137, 142, 181, 187, 190-200,202,
207, 208, 224, 232-234, 238, 285, 303, 308, 320,323-325
329,334-336,338-340,376,389,391,460,430,457,460
Incorporated ............. 9,25,30,95, 142,202,225,232,283,355,380,455
Individual ......... 3,5,7,27,29,34,39,47,58,60,63,64,70,75,76,85,86,
95,97,104,107,111,118,120,123,127,128,130-132,134,
136,137,141,150,165,184,189,190,198,199,213-215,237,
281,303,312,321,327,348,349,351,378,383,387,402,404,
405,410-412,415,424,428,440,442,447,458
Inflexible ................................................ 89, 197,339
Information 2-12, 15, 16, 19,21,26-32,34,36-42,45,47,48,50-59,62-64,
68-70,72-77,80,81,85-90,92,93,96-104, 107,108, 112,115-
117, 120, 124-142, 146, 149-152, 154, 156-163, 165-167, 169,
172-178,180-183,185-187,190-203,207,208, 212-217,220, 222,
225-234,237-240,242,246,247,259,262,263,267,270,276,
281-284,287-293,295,297,298,300,302-308,316,319,320,
325,327,329,330,334,336,337,339,347-353,357-361,364,
365,372,378,380-385,387,390,391,396,397,401,404-407,
413-416,427,436,439,441,443,444,447,448,453,455,458
Information overload ....................... 9, 185, 186, 195, 291, 298, 316
Information richness .......................................... 190,372
Innovation ......... 53, 80, 81, 119, 122, 123, 159, 208, 234, 235, 243, 248-271,
275,276,278,288,289,291,292,298,300,302,304,306,
309-311,313-315,317,318,330,336,340-342,344,367-369,
387,388,396,419,449,452,455,456
Innovations .................... 5,49,249,252,259,313,336,388,391,398
Innovative ....... 91, 104, 166, 202, 288, 312, 314-316, 330, 368, 371, 378, 404
Innovator ........................................... 257,302,367,450
Integrator ...................................................... 192
Interaction ................. 26,27,85, 109, 118, 120,249,307,319,421,426
Intercontinental ................................................. 273
Interdivisional ................................................ 59, 238
Internal process climate ............... 123, 131-136, 140, 288, 295, 311, 317
Internal process culture ....................................... 114, 116
International dimensions ................................... 82, 270, 278
Interpersonal ..................................................... 87
JIT .......................................................... 6,369
KLM ........................................................... 274
Knowledge-base ................................................... 27
Lars Kolind .......................... 399,406-414,417-420,426-431,437
Leader ............... 47,53,86-92,94-97,101,102,107,116,121-126,128,
129,132,133,135,136,141,274,317-319,362,368,379,
386,387,399,406,413,424
 470 Subject Index

Leader credibility ............ 116, 121, 122, 124-126, 128, 129, 132, 133, 135,
136,141,317,318,386,387
Leadership. . . . . . . . . .. 1,30,83-87,89-95,97,99, 103, 108-110, 114, 116, 121,
123,143,242,308,328,350,363,368,376,431,449
Leadership roles ............................................... 87, 94
Leadership style .............. 83,84,86,87,89-92,95,97,99, 103, 108,431
Leadership typologies .............................................. 89
Learrring ................ 33,112,327,335-338,342,344,345,347,350,351
LEGO ............................................ 50,77,224,234,267
Lexis-Nexis ................................................. 358, 359
Liaison ................ 6,7,13,70,87,91, 101, 125, 127, 150, 190, 191, 193,
195,285,288,307,334,337
Lifecycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 8, 103, 326-330, 332-334, 341, 350
Linjeflyg ..................................................... 44, 105
Locus of control .................................. 89-92, 94, 95, 107, 110
Long-tenn ...................................................... 123
M-fonn ......................................................... 142
Mack Trucks ...................................... 49,51,69,71-75,81
Malevolent .................................................. 175, 176
Management ..... 1,5,9,13,15,17,18,26,31,38,39,41-44,46,49,52-55,57,
60-62,65,69,71,72,75,78-87,89,92-95,97-99,101-110, 113,
115,121,137-143,145,156,157,160,162-164,169,172,181,
183,185,189,196,199,202,206,208-211,214,216,217,223,
232,234,235,239,240,246,247,252,256,258,263,264,272,
276-279,281,283,284,286,287,291-318,320,321,323-334,
337,339-342,344,349-354,360,363,364,371-373,375-379,
381,383-387,389,391,393,395-397,399,401,402,406,408,
413,414,416-420,424-429,431,432,441,442,444,447-56,460
Managernentpreferences ................ 15,95,103,354,360,364,377,393
Manager ........... 17,46-48,53,60,62,63,69,71,72,75,86-90,92,94-104,
107,111,145,156,185,200,206,213,214,226,228,256,258,
279,286-288,294,295,298,299,303,306,332,349-351,368,
383,387,408,412,414,422,423,427,441-444,446-448,
451,454,456,457,459,460
Managerial ........ 7,22,62,64,85,95, 104, 109,110, 117,141-143, 151, 195,
198,199,201,202,209,230,246,247,277,278,291,303,
325,332,333,349,372,381,384,385,393,444,452
Manpower ...................................................... 442
Manufacturing .......... 3,48-51, 53, 56, 62, 63, 67, 73, 81, 154,220-223, 225,
226,228,239,246,256,277,278,319,362,368-374,381,
382,391,397,398,407-410,419,421,431,439,441-444
Marketing ......... 6,45,50,51,59,62,91,156,209,224,251,262,278,337,
360,369,371,372,374,378,396,400-403,407,409,
421,441,443-447,450,451
Matrix configuration ............ 59, 60, 62-64, 66, 69, 102, 108, 129, 130, 229,
259,270,285,291,292,340,380,389
Subject Index 471

Matsushita ............................................... 67,254,256

Medical Products ................................................. 362
Medtronic, Inc. . ................................................. 395
Meetings ....... 13,43,64,77,101,125,127-131,135-137, 141, 190-193, 196,
197,199,202,205,207,215,216,231-234,307,308,323-325,
329,330,334,335,339,340,376,391,401,414,415,457,460
Merger ......................................... 146, 160, 204, 341, 350
Metaanalysis .................................... 155,157,225,227,228
Metaflexibility ................................................... 202
Microinvolvement ......... 26,90-92,94-103,107-109,282-284,286-288,291-
300,302-304,306,307,311,312,316-319,323-325,
329,330,333,334,340,455,456,459
Microlevel ...................................................... 256
Microlink ....................................................... 355
Mintzberg .......... 17,40,55,81,87,110,159,164,189,191,209,222,240,
271,278,319,344
Mission ........................... 10,13,111,235,375,377,379,385,411
Missions ..................................................... 10, 156
Moderator .................................................. 152, 227
Morale ................. 116,121-129,131-133,135,137,138,141,303,317,
318,369,375,376,386,387,428
Motivate .............. 9,84,87-89,92-94,96-98,106-108,114,138,143,199,
291,303,306,379,387,396,426,428,455
Multidivisional .............................................. 159, 344
Multifocus .................................................. 417,418
Multinational ................................... 45,60,66-68,440,449
Mycin ............................................ 28, 98, 284, 288, 321
Negotiations ................................. 77,130,204,274, 427, 428
Negotiator ....................................................... 87
Ne~ork .................. 1,4,6,43,87,149,150,240,247,273,274,277,
278,373,382,396,412,415-417,422,423,426
Ne~ork organization ............................................. 150
Ne~orking ................................................... 87,88
Niels Jacobsen ............................................... 417,431
Nonbureaucratic ................................................. 103
Noncoordination '................................................... 6
Nondecomposable ................................................ 219
Nondivisible ..................... 229,237-239,275,290,324,338,340,341
Nonprofessional .................................................. 162
Nonroutine .......... 20,21, 23, 26, 202, 212, 214, 216, 220-223, 227-230, 233-
236,238,239,257,258,260,269,283,289,293-295,298,302,
303,307,308,317,323-325,329,333,334,390-392
Organization ....... 1-18,20-26,30,31,35,37-55,58-76,78-81,83-91,97-110,
112-121, 123-138, 140-143,145-147, 149-160, 162-164,
166-173,175-177,180-189,194-210,212,214,217,218,221,
472 Subject Index

224-242, 244-248, 251-267, 269-272, 275-278, 281-288, 290-

294,297-299,301-308,312-320,323-344,347-353,355,361,
365,371,372,375,377,380-393,396-399,404,406,408-416,
419,424,426,429,430,436,439-442,445,446,449,451-456
Organizational audit .............................................. 352
Organizational climate .... .. 9, 111-113, 116-119, 131, 140, 142, 143, 339, 386
Organizational complexity ........... 25,26,42,68-73,99, 100, 127, 130, 133,
152,154-156,172,174,180-189,192,194-199,
202,207,213,227,232,263,296-299,320,
323-325,329,330,334-340,364,388,390,456
Organizational configuration .... 45, 65, 68, 70, 80, 109, 150, 159, 162,166, 190,
192,193,196,198,199,230-232,340,341,389,390
Organizational Consultant ........ 11, 19, 24, 29-34, 36, 37, 86, 94, 95, 98,181,
200,220,221,284,321,323,341,342,347-355,
357,389,439,440,444,449,453,455,456
Organizational culture .......... 85,86,104,111-114,117,118,142-144,405
Organizational learning ........................... 335,337,338,344,345
Organizationallifecycle ........................................... 332
Organizational performance ................................. 90, 102,278
Organizational structure . 1,8-10,12-14, 18-23,28,31,44,46,67,77,84-86,89,
92,95,99,101,102,104,105,113,117,118,141,146,
148, 149, 151, 153, 154, 160, 163, 165, 172, 174, 180,
181,198,202,207,209,210,212,215,234,236,241,
242,248,254,259,267,271,282,293,296,347,393,
409-411,415,417,419,429,430,440,453,459,460
Organizations ......... 2-9,11,17,21,23,31,32,37-40,42,45,48,49,51,55,
56,59,60,62,64-66,68,71-74,76,81,83,85,86,90,91,99-
104, 110, 114, 116, 123, 143, 146, 148, 150-152, 154-158, 160-
164,172,173,177,180,183,187,197-199,203,204,208-210,
214,222,224-229,231,233,239,240,245-249,252,272,276-
278,281,294,295,299,302,303,316,319,327,328,334-337,
342,344,352,357,358,366,387,391,439,453,457
Orgcon ..... '" 325,326,329,333,347,355,357,358,360,361,364,370,386,
392,393,418
Oticon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 350, 355, 399-437
Ownership ............ 15,58,163,164,246,354,360,362,364,380,393,414
Panasonic ................................................... 242, 254
Paperless ............................................... 415, 416, 449
Participation ........................ 2,44,89,123,188,203,371,385,386
Partnerships ................................................. 66,245
Password ................................................ '.' ..... 355
Payroll ..................................................... 380,445
Performance ........ 10,17,18,38,59,81,90,91,101,102,109,111,114,116,
117,119,142,143,160,164,169,170,176,177,198,199,209,
211,216,225,230,241,258,277-279,308,324,335,350,368,
441,443,450,452,458,460
Subject Index 473

Perrow .................. 19,21-25,27,39, 214-216, 220, 225, 240, 283, 344

Personnel ......... 50,51,67,69,72,73,75,80,84,86,101,104-106,109,111,
135,140,147,148,156,206,221,228,229,231,237,238,
273,275,297,383,385,391,398,416,427,428,442,
446-448,450,452-454,459
Petrochemical ............................................... 453, 456
Pharmaceutical ......................................... 49, 56, 57, 365
Planning ......... 12, 50, 53, 57, 86-89, 91, 93-95, 116, 129-132, 135-137, 143,
173,186,190-195,197,198,207,211,234,248,307,308,325,
330,339,349,379,383,398,401,409,441,444,459
Politicking ....................................................... 87
Power .......... 43,84,85,89,102,149, 179, 180,378,403,406,419,447,453
Preintemational .................................................. 67
Pressure ......................................... 74, 119, 122, 211, 416
Private ......... 132,135,155-157,201,211,299,305,306,365,404,412,430
Proactive ............................ 91-94,96-98, 106-108,244,378,460
Procedural ........... 42,134,182,191, 192,198, 199,207,233,330,334,339
Procedure ............ 5,6,19,94,105,132,182,190,194, 195, 198,226,233,
236,257,308,323,335,338-340,350
Process innovation ... 234, 255, 256, 260, 261, 263, 264, 266-269, 271, 368, 369
Process of organizational design .................................... 326
Process production ........................... 212-214, 221, 222, 226, 370
Product innovation ........ 53, 250, 251, 255-258, 260, 261, 263-269, 309, 310,
313,314,367,419,456
Production ........... 6, 12, 19,24,34-36,45,46,48,50,52,53,55,57,60,70,
72,73,118,147,154,156,159,160,201,211-217,220-226,
229,233,234,236-238,246,249,251,252,255,257,261,269,
270,311,312,314,316,323,332,337,370-372,374,383,388,
389,398,400-405,407,410,421-424,441-443,446
Productivity .............................................. 77, 135, 163
Products ......... 1,4,12,46,51-55,57-61,63,67,70, 123, 139, 146, 147, 151,
159,165,179,194,195,212,217,221,222,224,226,233-235,
241-243,246,249-258,260,261,263,264,266-269,271,273,
290,291,312-317,329,332,361-363,365-378,381,387-391,
395,400,402-405,407-412,423,425,441,446,448,450,451
Professional bureaucracy ..... 45,66,74,101,159,230,260,294,295,389,390
Professionalization .......... 66,70,74, 152-156, 158, 162,212,225,251,294,
300,309,312
Professionals ............... 45,66,74,85, 158, 185,216,224,225,230,294,
300,362,411,452
Profitability ................................................... 53, 77
Profitable ............... 1, 10,53, 199,260,262,264,272,274,388,411,455
Project configuration ............................................... 60
Prospector ......... 22,25,27,242,248-260,262,269,270,275,276,284,287,
291-296,298,299,302,303,306,307,310-312,314-319,321,
323-325,328,329,333,334,341,388-391,393,452
474 Subject Index

Psychological climate ............................. 116-120,137, 142, 143

Public ..................... 46,49,50,60,126,155,161,187,203,336,374,
377,380,400,404,413,418,428,446
Publicly-available information ...................................... 357
Punishment ............................................... 86,87,335
Quality. . . . . . . . . .. 33, 35, 46, 49, 50, 73, 84, 113, 120, 138, 143, 145, 146, 148,
149,159,176,177,179,211,216,226,234-236,251,254-258,
260-264,266-269,271,274,299,312,321,362,368,369,371,
377,380,381,384,385,388,398,402-404,419,441,443,445,
456,457,460
Rational goal climate ................. 123,134-137,140,291,311,317,318
Rational goal culture .......................................... 114, 115
Reactive . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 91-94, 96-98, 108, 116, 378, 387
Reactor .................. 25,248-253,255,256,269,276,310,316,330,388
Recognition ............................. 119, 122, 313, 369, 371, 379, 427
Recruiting ...................................... 139,145,204,417,424
Region ............................. 59,60,63,68,341,363, 374, 391,434
Regional ................................................. 58, 359, 382
Regulated ............................................ 42, 165, 166, 204
Regulation . . . . . . . . . . . . . . . . . .. 24, 45, 165, 172, 175, 176, 201, 273, 300, 351,
426,447,448,456,459
Regulators .................................................. 366, 398
Remoteness ..................................................... 171
Reorganization ................................................... 55
Reorganize .................................................. 408, 430
Reorganized ............................................ 55, 59, 79, 165
Responsiveness .................................... 45, 59, 369, 376, 390
Restructure .................................................. 33, 349
Restructured ............................................. 44, 419, 423
Restructuring ................................ 44, 104,349,406,417,418
Results-oriented ................................................. 123
Retail ................................ 54, 55, 220-223, 225, 238, 370, 400
Reward .............. 53,64,75,88,89,116,118,121,122,124-129,131-135,
137,140,141,171,198,199,207,233,240,308,317,327,328,
335,336,376,385,387,397,447, 460
Risk ............. 5,91-94,96-98,106-108,118,119,123,179,185,189,198,
199,253,258,310,311,314,315,318,332,335,336,378,
387,425,429,455
Roles ..................... 6,13,87,88,94,125,133,146,148,334,337,426
Routine ........... 5,8,17,20-23,25,27,40,45,87,108,169,171,175,188,
214,216,220-239,256,258,260,261,263,264,266-269,
275,283,288,294-300,302-312,313,315-319,330,
334,338-342,371,372,388,390-393,442,447,456
Rule-based ................................................. 19, 20, 31
Subject Index 475

Rules ........ 5, 7, 10, 13-16, 19, 20, 24-31, 34-36, 42, 45, 52, 64, 65, 73-75, 78,
98,101,102,104,118,119,124,127,132-134,146,157, 158, 173,
181,182, 185-187, 190-192,194-199,201,202,207,211,213,214,
216,225,226,230-234,237,238,259,275,282,284-286,295,
299-302,307,317,319,320,323,325,329,330,332,334-336,
338-341,347,348,384,385,391,393,412,425,447,460
Salary .................................................. 198,427,428
Samsonite .......... 165,166,171,175,180,182,186, 189, 197,202,241,242
Samsonite's ..................................................... 165
SAS ......... 2,7,41-45,68,69,74,75,77-81,85,104-108,110,138-140,143,
161,164,204-207,210,236-240,256,272-278,321,324,328,
332, 338-343, 355
Scandinavian Airline System (SAS) ................................ 41, 42
Scapegoating .................... 116, 121, 122, 124-126, 129, 132, 133, 135,
136,141,377,386,387
Semiautonomous ................................................. 445
Sensitivity analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 329, 344, 353-355
Service ...... 49,51,55,58,59,61,75,77-81,106, 108, 110, 123, 138, 154, 175,
177,179,217,220-223,225,226,234,236-238,240-242,255,
261,270-273,277,302,312,313,328,359,361-363,370,382,
387,389,391,395,404,410-412,439,440,445-448,456,458
Siemens ............................................ 366, 396, 404, 407
Simple configuration ........... 46-48, 69, 71, 75, 101, 159, 189, 192, 196, 197,
263,265,266,286-288,323,329
Situationfit ........................ 15,17,18,236,284,298,307,308,325
Situational misfits. . . . . . . . . . .. 31, 284, 285, 308, 309, 321, 329, 330, 332, 333,
340,341,351,353,392,393,409,432,444,449,452,460
Size ......... 1,2,13,15,17-19,22,24-28,30,31,35,39,47,63,76, 81, 84-86,
90,91,99,101,102,114,127,138,140,145-164,170,172,
179,183,192,195,201,203,211,212,214,220,226-228,234,
246,247,271,274,278,282,286-290,292-307,310,311,321,
323-325,328,329,332-334,339-341,353,354,360,362,364,
366,379,386,389,390,393,395,403,404,448,449,455,459
Skill ...... 64,71,76,145,151,152,195,212-215,217,218,223,312, 316, 329
Skill capabilities ................................................. 145
Socialization .................................................. 66, 87
Spaghetti Organization ................................ 355, 399, 419, 430
Spatial differentiation ............. 68, 69, 72, 73, 108, 296, 361, 382, 388, 456
Specialization ........... 3,45,46,48,51-53,55,59,61,68-70, 100, 101, 104,
136,146,148,155,156,159,172,182,185,186,188,
194,207,225,228,236,263,268,282,286,288,289,
296,298,300,316,323,336,337,381
Spies ........................................................... 262
Standardization. . . . . . . . . . . . . . . . . . .. 10, 24, 52, 66, 70, 73, 74, 158, 168, 179,
188,212-216,220,222,225,226,231,237,
476 Subject Index

244,250,265,269,294,295,299,316,370
Standards .................... 18,45,90,118,119,235,375,377,378,383,
384,398,425,451,460
Starkey ..................................................... 403, 404
Strategic ........... 2,8-11,38,39,42,54,57-59,75,79-81,85, 104, 107, 109,
142,178,183,201,202,206,208,209,242-245,247,248,250-
253,255,269-271,274,276-279,284-286,320,339,344,347,
349,350,357,360,363,366,371,378-380,395,397,410,452
Strategical ....................................................... 49
Strategies ............ 5-7, 52, 63, 64, 67, 70, 77, 78, 81, 91, 114, 151, 165, 166,
203,204,233,242,243,245,250,258,269,278,279,284,289,
291,312,313,325,333,350,384
Strategy . . . . . . . .. 1, 2, 6, 10, 13, 15, 17, 19, 22, 25-28, 30, 31, 35, 39-42, 46, 50,
63,64,66,77-81,84,85,89-91,99,105,106,110,114-116,
138-140,143,161,165,166,201-203,206,208,210,236-238,
241-243, 245, 247-249, 251-279, 284, 286-319, 321, 323-325,
328-330,333,334,338-342,345,349,350,354,364,
366,374,379,386,388-391,393,397,412,414,420,429,
431,432,451,452,455-457
Structuration theory .......................................... 230, 239
Structure ........ 1,2,7-28,31,32,38-40,44-46,48,55,59,60,67,68,77,78,
80-86,89-92,95,99, 101-105, 107-120, 124-137, 140, 141, 143,
145-149,151,153-155,158,160,162-168,171-174, 178-181, 185,
189,190, 197, 198,202,206-217,222,223,228,229,230,234-236,
239-245, 248-250, 252, 254, 255, 259, 260, 262, 263, 265-271,
275-278,282,283,285,286,288,293,296,298,323-325,328,330,
336-339,344,347,371,373,380,386,389,393,397,409-412,
415-419,426,427,429,430,440,441,446,447,449,451,
453,455,457,459,460
Subgroups ....................................................... 87
Subordinates ...................................... 71,92,444,447,459
Subsidiary ................ 10,44,67,105,276,361,375,382,401,404,408,
413,421,423,424,433
Subsidies ....................................................... 445
Substrategy ..................................................... 273
Subtasks ..................................................... 45, 52
Subunits ....................................... 53, 66, 71, 161, 290, 337
Support .......... 9,11,24,26,30,32,35,43,44,50,100,102,105,116,118,
119,122,141,148,155,187,191,195,200,225,231,232,242,
251,253,255,258,259,278,286-289,295,298,306,313,314,
316-319,325,338,348,353,363,369,372-374,378,382,413,
426,427,441,445,447,449,450
Targets ......................................... 5,6,123,336,337,349
Task ........... 3,5,6,9,20,21,35,43,47,50-52,56,69-71,89,94,104,107,
111, 118, 119, 127, 132, 134, 150, 152, 155, 158, 173, 181, 182,
Subject Index 477

194,206,213-217,219,229,231,250,259,267,282,312,327,
342,370,372,399,423,439,447,459
Team ........ 7,8,39,40,64,65,79,88, 106, 138, 139, 199,272,349,412,457
Teamwork .............................................. 111,123,386
Technology ........ 1,2,5-7,13,15,17,19-30,50,53,54,59,72,77,83-85,99,
108,114,116,140,150,152,154,160,162-165,169,172,176,
187, 194, 195,200-204,207,211-240,243,247-249,252-258,
260-269,271,275-277,283,284,286,288-300,302-319,321,
323-325,328-330,333,334,336,338-342,344,347,350,354,
357,360-365,368,370-374,383,384,388-398,402,404,
410-413,421,444,447,448,450,451
Technology change ................................................. 5
Telecommunication ....................................... 175, 176, 200
Template ................................................... 352, 354
Theory X .................................... 86,87,92,93,95, 107, 138
Theory Y ............................. 86,87,92,93,95,96, 107, 138,292
Top management .......... 9,49,52,54,55,57,69,75,80, 104, 107, 108, 121,
156,183,185,189,196,252,263,291,293,296,303,
305,310,311,350,377,378,381,385,391,399,401,
406,408,413,416,417,420,424,455,456
Total design fit ......................... 15, 18, 284, 285, 320, 321, 323, 342
TQM ................................................... 234, 235, 269
Training ......... 65,68,70,74,84,87, 104, 152, 160,215,237,296,312,316,
329,379,381,396,397,458
Traits ...................................................... 112,220
Transnational .............................................. 66, 67, 81
Trust .......... 106, 116, 118-120, 122, 124-129, 131-133, 135, 137, 138, 140,
141,300,303,317,318,321,375,386,457,461
Turbulent ............................................ 10,173,247,248
Tylenol ..................................................... 189, 197
Typology ........... 87,90,91,107,115,116,124,169,209,242,243,248-253,
255,277,279,388
Uncertain ........ 29, 39, 108, 169, 173, 177, 185, 200, 201, 230, 233, 284, 304,
312-314,321,323,325,329,330,333,341,344,367,444,451
Uncertainty ......... 3,5-8,26,29,38,63,64,77, 151, 166-188, 190-200, 202-
204,206,208-210,216,217,219,221,223,230,231,248,252,
259,267,283,284,287-290,292,293,296-302,304-311,313-
315,318,324,330,333,338-340,364,365,388,390,391,393,
445,448,457
Unemployment .................................................. 204
Unequivocal ................................................. 207,342
Unit production ........... 12, 212, 213, 216, 222, 223, 236, 269, 323, 371, 389
Unknown .......... 175,177,196,207,233,257,276,288,293,304,305,314,
316,335,381,382
Validation ................... 28, 32-38, 94, 99, 102, 103, 152, 154, 200, 323
478 Subject Index

Validation process ................................ 32,34-37,99, 152, 154

Validity ....................................... 32, 33, 119, 168, 231, 250
Velocity ................................................. 169, 208, 209
Vertical differentiation ......... 25,45,69,71-73, 100, 127, 130, 131,150,152,
182,186,213,214,227,228,232,296,298,323,
337,382,388,390,456
Viability ................ 4,46,53,207,245,249,252,267,282,332,334,388
Viable ....................... 4,17,32,46,53,166,245,269,276,282,287,
315,327,328,330,332,335
Video-conferencing ........................................... 230, 232
Wall Street Journal ............................................... 358
Weaknesses .......................................... 47, 173,349,404
Westinghouse ................................................... 440
Wholesale .................................. 220, 222, 223, 225, 238, 370
Widex .............................................. 404, 405, 407, 435
yahoo .......................................................... 358
This program contains proprietary information of Teknowledge Corporation and one or
more third parties. Its receipt or possession does not convey any rights to reproduce or
disclose its contents. Reproduction, disclosure or use without specific written
authorization of Teknowledge Corporation is strictly forbidden. Disassembly or
decompilation of the software is prohibited.

By purchasing this book the purchaser agrees to abide by the following Terms &
Conditions:

1. Agreement. This Agreement consists of these General Terms and Conditions.

2. License. Teknowledge Corporation (Teknowledge) grants to you a non-

exclusive, non-transferable license to use the M.4 software ("Product") in
concert with the Publication Knowledge Bases included with this book, only.
The license is granted to you subject to the following terms and conditions.

3. Ownership/Non-Disclosure. The Product is and shall remain the property of

Teknowledge. Teknowledge considers the Product to be and intends that it
remain the proprietary information and a trade secret of Teknowledge. You
are aware that this Agreement grants you no title or rights of ownership in the
Product. You agree to treat Product as the proprietary information and a trade
secret of Teknowledge.

4. Restrictions on Use. You agree not to modify or translate all or any part of
the Product. You agree not to distribute all or any part of the Product to
others except as expressly provided by Sections 7 and 8 or by addenda to this
Agreement, if any. You agree to maintain Teknowledge markings, including
any copyright notice or legend associated with the Product.

YOU AGREE NOT TO DISASSEMBLE, REVERSE ENGINEER OR

DECOMPILE THE PRODUCT, IN WHOLE OR IN PART. iF YOU DO
NOT COMPLY WITH THIS PROVISION, YOUR LICENSE IS
AUTOMATICALLY TERMINATED.

5. Limitation of Liability. TEKNOWLEDGE WILL NOT BE LIABLE FOR

ANY DAMAGES, INCLUDING ANY LOST PROFITS, LOST SAVINGS OR
OTHER INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT
OF THE USE OR INABILITY TO USE THE PRODUCT, EVEN IF
TEKNOWLEDGE HAS BEEN ADVISED OF THE POSSffiILITY OF SUCH
DAMAGES, OR FOR ANY CLAIM BY ANY OTHER PARTY. SOME
STATES DO NOT ALLOW THE LIMITATION OR EXCLUSION OF
LIABILITY FOR INCIDENTAL OR CONSEQUENTIAL DAMAGES AND,
TO SUCH EXTENT, THE ABOVE LIMITATION OR EXCLUSION MAY
NOT APPLY TO YOU.

6. No Warranty. EXCEPT AS OTHERWISE PROVIDED IN THIS

AGREEMENT, SHOULD THE PRODUCT PROVE TO BE DEFECTIVE,
YOU AGREE TO ASSUME THE ENTIRE COST OF ALL NECESSARY
SERVICING, REPAIR OR CORRECTION. YOU AGREE TO ASSUME
ALL RESPONSffiILITY FOR THE SELECTION OF THE PRODUCT TO
ACHIEVE YOUR INTENDED RESULTS, AND FOR THE
INSTALLATION, USE AND RESULTS OBTAINED FROM THE
 PRODUCT. TEKNOWLEDGE DOES NOT WARRANT THE PRODUCT
Wll.L MEET YOUR REQUIREMENTS OR THAT THE OPERATION OF
THE PRODUCT Wll.L BE UNINTERRUPTED OR ERROR FREE.

EXCEPT AS OTHERWISE PROVIDED IN THIS AGREEMENT THE

PRODUCT IS PROVIDED "AS IS" WITHOUT A WARRANTY OF ANY
KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT
LIMITED TO THE IMPLIED WARRANTIES OF MERCHANTABILITY
AND FITNESS FOR A PARTICULAR PURPOSE. SOME STATES DO
NOT ALLOW THE EXCLUSION OF IMPLIED WARRANTIES AND, TO
SUCH EXTENT, THE ABOVE EXCLUSION MAY NOT APPLY TO YOU.
YOU MAY ALSO HAVE OTHER RIGHTS WHICH VARY FROM STATE
TO STATE.

7. Transfer. You agree not to transfer this Agreement or this license to use all
or any part of the Product to any other person or entity except as permitted by
this Agreement. You may transfer this license to a third party if (a) such party
agrees to accept all of the terms and conditions of this Agreement; (b) the
Product, including all related materials, is transferred to the same party or any
materials not transferred are destroyed; and (c) such transfer complies with all
applicable United States export rules and regulations.

8. No Export. You understand and recognize that the Product may be subject to
the Export Administration Regulations of the United States Department of
Commerce and other United States government regulations relating to the
export of technical data and equipment. You are familiar with and agree to
comply with all such regulations, including any future modifications thereof.

9. Term and Termination. This Agreement is effective upon your receipt of the
product and will remain effective until terminated. This Agreement may be
terminated by (a) you at any time by providing written notice of termination
to Teknowledge and a certificate of destruction or return of the Product and all
related materials to Licensor, or by (b) Licensor at any time should you fail to
comply with any of the terms and conditions of this Agreement. Termination
subject to this Section 9 will create no obligation on the part of Teknowledge.

10. General. If any provision of this agreement is legally unenforceable or illegal,

such provision shall be severed from this Agreement and the balance of this
agreement shall continue in full force and effect. You agree to be responsible
for the payment of all state and local taxes (other than California income taxes)
incurred on account of this license. This Agreement shall be governed by
California law.

PUBLISHER'S NOTE

The Publisher is not liable for any errors in the program. This disclaimer will hold
the Publisher harmless against any costs, damages, and expenses caused by the
software.