Strategy
T. Kretschmer
MN3119
2018
Undergraduate study in
Economics, Management,
Finance and the Social Sciences
This subject guide is for a 300 course offered as part of the University of London
undergraduate study in Economics, Management, Finance and the Social
Sciences. This is equivalent to Level 6 within the Framework for Higher Education
Qualifications in England, Wales and Northern Ireland (FHEQ).
For more information about the University of London, see: london.ac.uk
This guide was prepared for the University of London by:
T. Kretschmer, Professor of Business Studies, Ludwig-Maximilians-Universität, Munich,
Germany.
Acknowledgements
The author would like to thank Brooke Russell for her excellent assistance on the first
edition of the guide and especially her help with the extended activities. Thorsten Grohsjean,
Christina Finsterwalder and Sanja Rikanovic were very helpful in the revision of this guide,
in particular in putting together the guide to the in-chapter activities and the sample
examination questions.
This is one of a series of subject guides published by the University. We regret that due to
pressure of work the author is unable to enter into any correspondence relating to, or arising
from, the guide. If you have any comments on this subject guide, favourable or unfavourable,
please use the form at the back of this guide.
University of London
Publications Office
Stewart House
32 Russell Square
London WC1B 5DN
United Kingdom
www.london.ac.uk
Published by: University of London
© University of London 2013. Reprinted with minor revsions 2018.
The University of London asserts copyright over all material in this subject guide except where
otherwise indicated. All rights reserved. No part of this work may be reproduced in any form,
or by any means, without permission in writing from the publisher. We make every effort to
respect copyright. If you think we have inadvertently used your copyright material, please let
us know.
Contents
Contents
Chapter 1: Introduction........................................................................................... 1
1.1 Introduction to the subject........................................................................................ 1
1.2 Aims of this course................................................................................................... 1
1.3 Learning outcomes................................................................................................... 1
1.4 The structure of the course........................................................................................ 1
1.5 Use of the guide and hours of study.......................................................................... 2
1.6 Essential reading...................................................................................................... 2
1.7 Further reading......................................................................................................... 3
1.8 Online study resources.............................................................................................. 5
1.9 The examination and examination advice.................................................................. 6
1.10 The syllabus............................................................................................................ 7
Chapter 2: The evolution of strategic management as an interdisciplinary field...... 9
Learning outcomes......................................................................................................... 9
Essential reading............................................................................................................ 9
2.1 Introduction............................................................................................................. 9
2.2 Early theories .......................................................................................................... 9
2.3 Michael Porter and the industrial organisation paradigm......................................... 10
2.4 Organisational economics....................................................................................... 10
2.5 Resource-based view, dynamic capabilities and leadership theory............................ 11
2.6 A consensus definition of the field........................................................................... 11
2.7 Key concepts ......................................................................................................... 12
2.8 A reminder of your learning outcomes..................................................................... 12
Chapter 3: Analysis of market structure................................................................ 13
Learning outcomes....................................................................................................... 13
Essential reading.......................................................................................................... 13
Further reading............................................................................................................. 13
3.1 Introduction........................................................................................................... 13
3.2 Techniques of market definition............................................................................... 13
3.3 Market analysis with many firms............................................................................. 16
3.4 Key concepts.......................................................................................................... 18
3.5 A reminder of your learning outcomes..................................................................... 18
3.6 Sample examination questions................................................................................ 18
3.7 Extended activity: the commercial banking industry in the United States.................. 19
Chapter 4: Introduction to game theory and strategy.......................................... 21
Learning outcomes....................................................................................................... 21
Essential reading.......................................................................................................... 21
Further reading............................................................................................................. 21
4.1 Introduction........................................................................................................... 21
4.2 Static games........................................................................................................... 22
4.3 Dynamic games...................................................................................................... 27
4.4 Key concepts.......................................................................................................... 30
4.5 A reminder of your learning outcomes..................................................................... 30
4.6 Sample examination questions................................................................................ 30
i
MN3119 Strategy
Chapter 5: Oligopolistic models of competition................................................... 33
Learning outcomes....................................................................................................... 33
Essential reading.......................................................................................................... 33
Further reading............................................................................................................. 33
5.1 Introduction........................................................................................................... 33
5.2 Preliminaries.......................................................................................................... 34
5.3 Bertrand competition.............................................................................................. 34
5.4 Cournot competition............................................................................................... 35
5.5 Comparing Bertrand and Cournot........................................................................... 36
5.6 Stackelberg leadership............................................................................................ 37
5.7 Product differentiation............................................................................................ 38
5.8 Key concepts.......................................................................................................... 39
5.9 A reminder of your learning outcomes..................................................................... 39
5.10 Sample examination questions.............................................................................. 39
Extended activity: articles on Airbus-Boeing competition................................................ 40
Chapter 6: The resource-based view of the firm................................................... 41
Learning outcomes....................................................................................................... 41
Essential reading.......................................................................................................... 41
Further reading............................................................................................................. 41
6.1 Introduction........................................................................................................... 41
6.2 Competitive advantages and resources................................................................... 42
6.3 Some examples of resources as sources of competitive advantage........................... 43
6.4 The dynamics of resource and capability building..................................................... 46
6.5 Key concepts.......................................................................................................... 47
6.6 A reminder of your learning outcomes..................................................................... 48
6.7 Sample examination questions................................................................................ 48
6.7 Extended activity: Dell Computer Corporation ......................................................... 48
Chapter 7: Strategic asymmetries – persistent dominance over time.................. 69
Learning outcomes....................................................................................................... 69
Essential reading.......................................................................................................... 69
Further reading............................................................................................................. 69
7.1 Introduction........................................................................................................... 69
7.2 Firm asymmetries – long- or short-term?................................................................. 69
7.3 Traditional sources of persistent dominance............................................................. 71
7.4 Dynamic capabilities............................................................................................... 75
7.5 Key concepts.......................................................................................................... 76
7.6 A reminder of your learning outcomes..................................................................... 76
7.7 Sample examination questions................................................................................ 76
Extended activity: competition in the wide-body aircraft market..................................... 76
Chapter 8: Organisation design............................................................................. 77
Learning outcomes....................................................................................................... 77
Essential reading.......................................................................................................... 77
Further reading............................................................................................................. 77
8.1 Introduction........................................................................................................... 77
8.2 Strategy and structure............................................................................................ 77
8.3 Organisation design and competitive advantage..................................................... 83
8.4 Key concepts.......................................................................................................... 83
8.5 A reminder of your learning outcomes..................................................................... 83
8.6 Sample examination questions................................................................................ 84
ii
Contents
Chapter 9: Value chain analysis: vertical relations................................................ 85
Learning outcomes....................................................................................................... 85
Essential reading.......................................................................................................... 85
Further reading............................................................................................................. 85
9.1 Introduction........................................................................................................... 85
9.2 Double marginalisation........................................................................................... 85
9.3 Vertical foreclosure................................................................................................. 88
9.4 Key concepts.......................................................................................................... 89
9.5 A reminder of your learning outcomes..................................................................... 90
9.5 Sample examination questions................................................................................ 90
9.6 Extended activity: outsourcing at Eriksson and Sony and Loews............................... 90
Chapter 10: Vertical integration and transaction cost.......................................... 91
Learning outcomes....................................................................................................... 91
Essential reading.......................................................................................................... 91
Further reading............................................................................................................. 91
10.1 Introduction......................................................................................................... 91
10.2 Purchasing versus production costs....................................................................... 91
10.3 Coordination costs............................................................................................... 92
10.4 Proprietary knowledge.......................................................................................... 93
10.5 Transaction costs.................................................................................................. 94
10.6 Asset specificity.................................................................................................... 95
10.7 Alternatives to ‘make’ or ‘buy’............................................................................... 96
10.8 Key concepts........................................................................................................ 97
10.9 A reminder of your learning outcomes................................................................... 97
10.10 Questions for discussion..................................................................................... 97
Extended activity: athenahealth.................................................................................... 97
Chapter 11: Collusion.......................................................................................... 103
Learning outcomes..................................................................................................... 103
Essential reading........................................................................................................ 103
Further reading........................................................................................................... 103
11.1 Introduction....................................................................................................... 103
11.2 Stability of cooperation....................................................................................... 103
11.3 Extending the model........................................................................................... 105
11.4 Key concepts...................................................................................................... 107
11.5 A reminder of your learning outcomes................................................................. 107
11.6 Sample examination questions............................................................................ 107
Extended activity: the DeBeers diamond cartel............................................................ 107
Chapter 12: Strategic partnerships..................................................................... 115
Learning outcomes..................................................................................................... 115
Essential reading........................................................................................................ 115
Further reading........................................................................................................... 115
12.1 Introduction....................................................................................................... 115
12.2 Building capabilities........................................................................................... 115
12.3 Business and strategic partnerships..................................................................... 116
12.4 Equity ownership................................................................................................ 118
12.5 Key concepts...................................................................................................... 120
12.6 A reminder of your learning outcomes................................................................. 120
12.7 Sample examination questions............................................................................ 120
Extended activity: the EU aviation industry.................................................................. 120
iii
MN3119 Strategy
Chapter 13: Competitive dynamics..................................................................... 123
Learning outcomes..................................................................................................... 123
Essential reading........................................................................................................ 123
Further reading........................................................................................................... 123
13.1 Introduction....................................................................................................... 123
13.2 A framework to analyse competitive dynamics..................................................... 124
13.3 Key concepts...................................................................................................... 127
13.4 A reminder of your learning outcomes................................................................. 127
13.5 Sample examination questions............................................................................ 127
Chapter 14: Entry and entry deterrence.............................................................. 129
Learning outcome....................................................................................................... 129
Essential reading........................................................................................................ 129
Further reading........................................................................................................... 129
14.1 Introduction....................................................................................................... 129
14.2 Structural entry barriers...................................................................................... 130
14.3 Strategic entry barriers........................................................................................ 132
14.4 Summary............................................................................................................ 136
14.5 Key concepts...................................................................................................... 136
14.6 A reminder of your learning outcome.................................................................. 137
14.7 Sample examination questions............................................................................ 137
Extended activity: Dubai flowers and internet banking................................................. 137
Chapter 15: Research and development competition......................................... 143
Learning outcomes..................................................................................................... 143
Essential reading........................................................................................................ 143
Further reading........................................................................................................... 143
15.1 Introduction....................................................................................................... 143
15.2 Terminology........................................................................................................ 143
15.3 Innovation and market structure......................................................................... 146
15.4 Strategic issues in R&D....................................................................................... 151
15.5 Some further thoughts on R&D........................................................................... 155
15.6 Key concepts...................................................................................................... 156
15.7 A reminder of your learning outcomes................................................................. 156
15.8 Sample examination questions............................................................................ 157
Extended activity: discovering DNA............................................................................. 158
Chapter 16: Technology adoption........................................................................ 169
Learning outcomes..................................................................................................... 169
Essential reading........................................................................................................ 169
Further reading........................................................................................................... 169
16.1 Introduction....................................................................................................... 169
16.2 Adoption dependence......................................................................................... 169
16.3 Strategic technology adoption – option value...................................................... 170
16.4 Technology diffusion models............................................................................... 172
16.5 Key concepts...................................................................................................... 176
16.6 A reminder of your learning outcomes................................................................. 176
16.7 Sample examination questions............................................................................ 176
Extended activity: the adoption of Botox..................................................................... 177
Chapter 17: Network effects............................................................................... 181
Learning outcomes..................................................................................................... 181
Essential reading........................................................................................................ 181
Further reading........................................................................................................... 181
iv
Contents
17.1 Introduction....................................................................................................... 181
17.2 Network market structures.................................................................................. 182
17.3 Technology diffusion with network effects........................................................... 183
17.4 Generic strategies in network markets8................................................................................................................. 186
17.5 Fighting a standards battle................................................................................. 187
17.6 Key concepts...................................................................................................... 188
17.7 A reminder of your learning outcomes................................................................. 188
17.8 Questions for discussion..................................................................................... 188
Extended activity: Skype and digital cinema................................................................. 188
Appendix 1: Sample examination paper............................................................. 191
Appendix 2: Guidance on answering the Sample examination paper................ 193
v
MN3119 Strategy
Notes
vi
Chapter 1: Introduction
Chapter 1: Introduction
1.1 Introduction to the subject
Let us begin this course with a wager: I bet you that, when opening
the business section of your local newspaper, you will find the word
‘strategy’ at least once per page. A Google search of ‘strategy’ throws up
1,010,000,000 hits. Narrowing this down to ‘business strategy’ leaves
us with 622,000,000 hits. That’s a lot of business strategy! There are a
large number of definitions of strategy, and I will not attempt to write my
own. There are also entire research fields of ‘business strategy’ ‘corporate
strategy’ ‘strategy content’ ‘strategy process’ ‘management strategy’
‘competitive strategy’ and so on, and I will not get into the fine distinctions
between one and the other. Much more, I will try to boil down ‘strategy’
to what most definitions have in common, and more importantly, I will
introduce a number of techniques that will be helpful in formulating,
analysing and implementing a strategy.
1.2 Aims of this course
In this course, you will not learn ‘how-to’ recipes of how to react to specific
situations. What you will learn is a way of thinking about such situations.
In management, as in economics, the right answer to almost any question
is ‘it depends’. What you will learn in this course is what the right answer
depends on and, given a particular set of circumstances, how you can
analyse the situation.
1.3 Learning outcomes
Once you have completed the course and done the Essential reading and
activities, you should be able to:
•
use tools of strategic analysis and game theory to value and analyse
strategic options in real life.
In particular, you should be able to:
•
anticipate the actions of a rational (that is, individually profitmaximising) rival and act accordingly.
1.4 The structure of the course
The course is structured in six parts: after this and the following
introductory chapter, the basic building blocks of strategic analysis are
introduced: market analysis, game theory and oligopoly competition. We
will refer back to these chapters often in later chapters, so you are advised
to spend a significant amount of time on these and make sure you have
understood the basic principles and techniques of these chapters. The
third part introduces the sources of competitive advantage – resources and
capabilities, strategic asymmetries and organisation design. These chapters
will aim to explain why firms are different, in what way and what makes
some firms more successful than others. The fourth part will study firms
and their relations to other firms. Why are some firms vertically integrated
and some not, and what are the implications of this? When do firms
cooperate with rivals, and what do partnerships among firms typically look
like? The fifth part takes a closer look at competition among firms – over
1
MN3119 Strategy
time, and in specific situations like entry or research and development.
This is where we will use the tools of game theory we introduced in the
second part of the guide very intensively. In the final part we will look at
some of the particularities of high-technology industries. This part of the
guide may be considerably more demanding technically than the previous
ones, but by then you should have had enough opportunity to study, revise
and practise the concepts from previous chapters. See this final part as the
‘icing on the cake’ – after you have looked at many techniques and topics
in isolation, these last chapters give you an opportunity to look at some of
them in combination.
1.5 Use of the guide and hours of study
To help you get the most out of this course, you will be given a number
of examples and activities throughout each chapter. These will vary in
difficulty and style. When you read a chapter, try and do these as you go
along, and go back to the ones you had problems with the first time round
once you have completed the chapter.
At the end of the chapter, you will find a list of questions or exercises
designed to challenge you and to check if you have read and understood
the chapter. They are titled Sample examination questions. These precise
questions are unlikely to come up, but they should give you a general
impression of the level that is required in this course. If you are studying
this text in a group, you might want to consider discussing these questions
in a tutorial-style session at the end of each chapter.
As a further test of your skills, most chapters will have an Extended
activity in a separate section. This is a longer text, case study, interview,
quote etc. that illustrates some of the concepts in the chapter, and gives
some concrete questions at the end. These activities will test your overall
grasp of the chapter and are often not limited to one chapter or one topic.
The advice normally given to University of London students is that if they
are studying one course a year, they should allow at least six hours of
study every week. Most of the chapters are relatively short compared to
regular textbook chapters. Therefore, a chapter should be read all in one
go to give you a general idea of what it is about. After that, you should set
some time aside to work through the chapter properly. Times will vary for
every student and every chapter.
Sample examination questions and extended activities will probably
take very little time if you just glance over them and sketch them out in
your mind. It is recommended, however, that you write out some of your
answers as you may find that a casual thought will not look as convincing
if you write them out on paper and you need to have a clear, coherent and
logical argument.
1.6 Essential reading
For many chapters, the Essential reading is:
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) [ISBN 9780262032865].
The chapters in the book will often clarify points and go a little further,
but you will find most of the points covered in the subject guide chapter
are also covered in Cabral. Reading further often highlights some specific
aspects of the chapters or describes some of the research findings
2
Chapter 1: Introduction
explained in the chapter. If you would like to get another perspective on
strategy from an economics viewpoint, you should have a look at
Besanko, D., D. Dranove, M. Shanley and S. Shaefer Economics of Strategy. (New
Jersey: Wiley, 2009) fifth edition [ISBN 9780470484838].
or
Saloner, G., A. Shepard and J. Podolny Strategic Management. (New Jersey: Wiley,
2006) revised edition [ISBN 9780470009475].
Detailed reading references in this subject guide refer to the editions of the
set textbooks listed above. New editions of one or more of these textbooks
may have been published by the time you study this course. You can use
a more recent edition of any of the books; use the detailed chapter and
section headings and the index to identify relevant readings. Also check
the virtual learning environment (VLE) regularly for updated guidance on
readings.
Journals
Dierickx, I. and K. Cool ‘Asset stock accumulation and sustainability of competitive
advantage’, Management Science 35(12) 1989, pp.1504–511.
Dyer, J. and H. Singh ‘The relational view: co-operative strategy and sources of
interorganisational competitive advantage’, Academy of Management Review
23(4) 1998, pp.660–79.
Haskel, J. and C. Martin ‘Capacity and competition: empirical evidence on UK
panel data’, Journal of Industrial Economics 42(1) 1994, pp.23–44.
Lexecon Ltd ‘An introduction to quantitative techniques in competition analysis’,
Lexecon Ltd. publication, mimeo. www.crai.com/ecp/assets/quantitative_
techniques.pdf
Swaminathan, A. ‘Entry into new market segments in mature industries:
endogenous and exogenous segmentation in the US brewing industry’, Strategic
Management Journal 19(4) 1998, pp.389–404.
1.7 Further reading
Please note that as long as you read the Essential reading you are then free
to read around the subject area in any text, paper or online resource. You
will need to support your learning by reading as widely as possible and by
thinking about how these principles apply in the real world. To help you
read extensively, you have free access to the VLE and University of London
Online Library (see below).
For your ease of reference here is a list of all the Further reading for this
course.
Angelmar, R. ‘Market structure and research intensity in high-technologicalopportunity industries’, Journal of Industrial Economics 34(1) 1985, pp.69–79.
Arora, Alfonso Gambardella ‘Complementarities and external linkages: the
strategies of large firms in biotechnology’, Journal of Industrial Economics 38(4)
1990, pp.361–79.
Barney, J.B. ‘Firm resources and sustained competitive advantage’, Journal of
Management, 17 1991, pp.99–120.
Barney, J.B. ‘Resource-based theories of competitive advantage: A ten-year
retrospective on the resource-based view’, Journal of Management 6(2001a),
pp.643–50.
Benkard, L. ‘Learning and forgetting: the dynamics of aircraft production’,
American Economic Review 90(4) 2000, pp.1034–54.
Bryson, A., R. Gomez and T. Kretschmer Catching a wave: the adoption of voice and
high-commitment workplace practices in Britain, 1984–1998. CEP discussion
3
MN3119 Strategy
paper DP 0676. (London: Centre for Economic Performance, 2005) (http://cep.
lse.ac.uk/pubs/download/).
Cabral, L. ‘R&D competition when firms choose variance’, Journal of Economics and
Management Strategy 12(1) 2003, pp.139–50.
Cabral, L. and M. Riordan ‘The learning curve, predation, antitrust, and welfare’,
Journal of Industrial Economics 45(2) 1997, pp.155–69.
Camerer, C. ‘Redirecting research in business policy and strategy’, Strategic
Management Journal 6(1) 1985, pp.1–15.
Chen, M.-J. ‘Competitor analysis and interfirm rivalry: towards a theoretical
integration’, Academy of Management Review 21(1) 1996, pp.100–34.
Chen, M.-J. and D. Miller ‘Competitive dynamics: themes, trends, and a prospective
research platform’, The Academy of Management Annals 6(1) 2012, pp.135–210.
Church, J. and R. Ware Industrial Organisation: A Strategic Approach. (New York:
McGraw-Hill, 2000) [ISBN 9780071166454] Chapter 8, Classic Models of
Oligopoly
Csaszar, F. ‘Organizational structure as a determinant of performance: evidence
from mutual funds’, Strategic Management Journal 33 2012, pp.611–32.
David, P. ‘Clio and the economics of QWERTY’, American Economic Review 75(2)
1985, pp.332–37.
Dixit, A. and S. Skeath Games of strategy. (London: Norton & Company, 2004)
second edition [ISBN 9780393924992].
Emmons, W. and R. Prager ‘The effects of market structure and ownership on
prices and service offerings in the US cable television industry’, Rand Journal of
Economics 28(4) 1997, pp.732–50.
Farrell, J. and G. Saloner ‘Standardization, compatibility, and innovation’, Rand
Journal of Economics 16(1) 1985, pp.70–83.
Ferrier, W.J., K.G. Smith and C.M. Grimm ‘The role of competitive action in
market share erosion and industry dynamics: a study of industry leaders and
challengers’, Academy of Management Journal 42(4) 1999, pp.372–88.
Fudenberg, D. and J. Tirole ‘Preemption and rent equalization in the adoption of
new technology’, Review of Economic Studies 52(3) 1985, pp.383–402.
Geroski, P. ‘Early warning of new rivals’, Sloan Management Review 40(3) 1999,
pp.107–16.
Geroski, P. ‘Models of technology diffusion’, Research Policy 29(4–5) 2000,
pp.603–25.
Geroski, P. ‘Thinking creatively about markets’, International Journal of Industrial
Organisation 16(6) 1998, pp.677–95.
Gilbert, R. and D. Newbery ‘Preemptive patenting and the persistence of
monopoly’, American Economic Review 72(3) 1982, pp.514–26.
Griliches, Z. ‘Hybrid corn: an exploration in the economics of technological
change’, Econometrica 1957. Reprinted in Z. Griliches (ed.) Technology,
Education, and Productivity. (New York: Basil Blackwell, 1988) [ISBN
9780631156147] pp.27–52.
Hamel, G. and C.K. Prahalad ‘The core competence of the corporation’, Harvard
Business Review 68(May–June) 1990, pp.79–93.
Hitt, M., H.I.R. Volberda, R. Morgan, R. Hoskisson and P. Reinmoeller Strategic
Management: Competitiveness and Globalization. (Hampshire: Cengage Learning
EMEA, 2011) [ISBN 9781408019184].
Katz, M. and C. Shapiro ‘Systems competition and network effects’, Journal of
Economic Perspectives 8(2) 1994, pp.93–115.
Kay, J. Foundations of Corporate Success: How Business Strategies Add Value. (Oxford:
Oxford University Press, 1995) [ISBN 9780198289883], Chapters 5–8.
Klette, T. ‘R&D, scope economics, and plant performance’, Rand Journal of
Economics 27(3) 1996, pp.502–22.
Koski, H and T. Kretschmer ‘Survey on competing in network industries: firm
strategies, market outcomes and policy implications’, Journal of Industry,
Competition and Trade 4(1) 2004, pp.5–31.
4
Chapter 1: Introduction
Kretschmer, T. and P. Puranam ‘Integration through incentives within differentiated
organisations’, Organization Science 19(6) 2008, pp.860–75.
Leonard, R. ‘Reading Cournot, reading Nash’ Economic Journal 104(424) 1994.
Lieberman, M. ‘Market growth, economies of scale, and plant size in the chemical
processing industries’ Journal of Industrial Economics 36(2) 1987, pp.175–91.
McAfee, P., H. Mialon and M. Williams ‘What is a barrier to entry?’, American
Economic Review Papers and Proceedings (94) 2004, pp.461–65.
Monteverde, K. ‘Technical dialog as an incentive for vertical integration in the
semiconductor industry’, Management Science 41(10) 1995, pp.1624–38.
Ohashi, H. ‘The role of network effects in the US VCR market, 1978–86’, Journal of
Economics and Management Strategy 12(4) 2003, pp.447–94.
Peteraf, M.A. ‘The cornerstones of competitive advantage: A resource-based view’,
Strategic Management Journal 14(3) 1993, pp.179–91.
Porter, R. ‘A study of cartel stability: the Joint Executive Committee, 1880–1886’,
Bell Journal of Economics 14(2) 1983, pp.301–14.
Postrel, S. ‘Competing networks and proprietary standards: the case of
quadraphonic sound’, The Journal of Industrial Economics 39(2) 1990,
pp.169–85.
Sah R.K. and J.E. Stiglitz ‘The architecture of economic systems: hierarchies and
polyarchies’, American Economic Review 76(4) 1986, pp.716–27.
Saloner, G. ‘Modeling, game theory, and strategic management’, Strategic
Management Journal (12) 1991, pp.119–36.
Shapiro, C. and H. Varian Information rules. (Cambridge, MA: HBS Press, 1999)
[ISBN 97807875848631] Chapter 7 ‘Networks and Positive Feedback’.
Smith, K.G., W.J. Ferrier and H. Ndofor ‘Competitive dynamics research: critique
and future directions’ in M.A. Hitt, R.E. Freeman and J.S. Harrison (eds), The
Blackwell Handbook of Strategic Management. (Oxford: Blackwell Publishing,
2001) pp.314–61.
Spence, A. ‘The learning curve and competition’, The Bell Journal of Economics
12(1) 1981, pp.49–70.
Wernerfelt, B. ‘The resource-based view of the firm’, Strategic Management Journal
5(2) 1984, pp.171–80.
1.8 Online study resources
In addition to the subject guide and the Essential reading, it is crucial that
you take advantage of the study resources that are available online for this
course, including the VLE and the Online Library.
You can access the VLE, the Online Library and your University of London
email account via the Student Portal at:
http://my.londoninternational.ac.uk
You should have received your login details for the Student Portal with
your official offer, which was emailed to the address that you gave on
your application form. You have probably already logged in to the Student
Portal in order to register. As soon as you registered, you will automatically
have been granted access to the VLE, Online Library and your fully
functional University of London email account.
If you have forgotten these login details, please click on the ‘Forgotten
your password’ link on the login page.
The VLE
The VLE, which complements this subject guide, has been designed to
enhance your learning experience, providing additional support and a
sense of community. It forms an important part of your study experience
with the University of London and you should access it regularly.
5
MN3119 Strategy
The VLE provides a range of resources for EMFSS courses:
•
Self-testing activities: Doing these allows you to test your own
understanding of subject material.
•
Electronic study materials: The printed materials that you receive from
the University of London are available to download, including updated
reading lists and references.
•
Past examination papers and Examiners’ commentaries: These provide
advice on how each examination question might best be answered.
•
A student discussion forum: This is an open space for you to discuss
interests and experiences, seek support from your peers, work
collaboratively to solve problems and discuss subject material.
•
Videos: There are recorded academic introductions to the subject,
interviews and debates and, for some courses, audio-visual tutorials
and conclusions.
•
Recorded lectures: For some courses, where appropriate, the sessions
from previous years’ Study Weekends have been recorded and made
available.
•
Study skills: Expert advice on preparing for examinations and
developing your digital literacy skills.
•
Feedback forms.
Some of these resources are available for certain courses only, but we
are expanding our provision all the time and you should check the VLE
regularly for updates.
Making use of the Online Library
The Online Library contains a huge array of journal articles and other
resources to help you read widely and extensively.
To access the majority of resources via the Online Library you will
either need to use your University of London Student Portal login
details, or you will be required to register and use an Athens login:
http://tinyurl.com/ollathens
The easiest way to locate relevant content and journal articles in the
Online Library is to use the Summon search engine.
If you are having trouble finding an article listed in a reading list, try
removing any punctuation from the title, such as single quotation marks,
question marks and colons.
For further advice, please see the online help pages:
www.external.shl.lon.ac.uk/summon/about.php
1.9 The examination and examination advice
Important: the information and advice given in the following section
are based on the examination structure used at the time this guide was
written. Please note that subject guides may be used for several years.
Because of this we strongly advise you to always check both the current
Regulations for relevant information about the examination, and the VLE
where you should be advised of any forthcoming changes. You should also
carefully check the rubric/instructions on the paper you actually sit and
follow those instructions.
The examination will be a three-hour unseen written examination covering
all aspects of the syllabus. In the examination, you will be asked to:
6
Chapter 1: Introduction
•
reproduce some knowledge. (This will get you close to a pass grade,
although some application is needed).
•
apply knowledge to new situations. (This will lift you to the high lower
second, or low upper second marks (assuming you get all the above
questions right)) and
•
make new connections between topics and/or phenomena. (This will
enable you to obtain a first class mark in this course).
As with most examinations, try to allocate your time approximately
proportional to the marks available. If you are having problems with an
analytical question that worth very few points, it’s best to let that one go
and avoid losing time that you could use on another question.
Remember, it is important to check the VLE for:
•
up-to-date information on examination and assessment arrangements
for this course
•
where available, past examination papers and Examiners’ commentaries
for the course which give advice on how each question might best be
answered.
1.10 The syllabus
Basic game theory: Two-player games. Static and dynamic games and
some examples. Equilibrium concepts and solution mechanisms – Nash
equilibrium, dominant/dominated strategies, backward induction.
Oligopoly competition: Perfect competition and monopoly. Price
competition and the Bertrand paradox. Quantity competition. Reaction
functions. Bertrand versus Cournot.
Analysis of market structure: Describing market structure: C4-ratio,
Herfindahl index, Lerner index and market power. Market definition –
techniques and interpretation.
Collusion: Cartels and antitrust. Cartel stability and the discount factor.
Market dynamics and stability of collusion.
Strategic alliances: Portfolio test. Strategic and business partnerships.
Sources of complementarity. Resource accumulation. Absorptive capacity.
Organisation design: Organisational fit, Strategy and structure, Functional
organisation, Multidivisional structure, Worldwide structure.
Competitive Dynamics: Competitive dynamics, Competitive action, Resource
similarity, Market commonality, Awareness, motivation and capability.
Strategic asymmetries: Economies of Scale, sources and consequences.
Scope Economies: Airline Hubs. Learning or experience curve. Firm
strategies with EoScale/Scope/Learning. First-mover advantages. Market
structure with increasing returns.
Value chain analysis and vertical relations: Double marginalisation and
its remedies. Vertical foreclosure. Retailer competition and investment
externalities.
Vertical integration and transaction cost: Make or Buy. Contracts. RelationSpecific Assets and Hold-Up. Economic Rents and Quasi-Rents.
Entry and entry deterrence: Structural determinants of entry. Entry barriers
and exit barriers. Entry deterrence. Identifying entrants.
Research and Development: Market structure and R&D intensity. R&D
rivalry. Monopolists’ and entrants’ R&D incentives. Risk choice of R&D.
Benefits of the patent system. Sleeping patents. Spillovers.
7
MN3119 Strategy
Technology adoption: Preemption games. Option value and future
technological generations. Technology diffusion: Heterogeneity, epidemic,
and population ecology approaches.
Network effects: Direct and indirect network effects. Systems goods. Excess
inertia. Excess momentum. Firm strategies with network effects. Standards
Battles.
Realising intra-firm synergies: Incentive-setting: Free riding versus
cooperation. Interfaces: Modularity, task separation and task similarity.
Distinctive capabilities and competitive advantage: Identifying competitive
advantage. Sustainability of competitive advantage. Innovation,
architecture, strategic assets, reputation. Leveraging distinctive
capabilities.
All topics are supplemented in the subject guide with specially written case
studies.
8
Chapter 2: The evolution of strategic management as an interdisciplinary field
Chapter 2: The evolution of strategic
management as an interdisciplinary field
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
explain how the field of strategic management evolved over time
•
discuss the elements of the shared meaning in the strategic
management field.
Essential reading
Hoskisson, R.E., M.A. Hitt and D. Wan W.P. Yiu, ‘Theory and research in strategic
management: swings of a pendulum’, Journal of Management, 25(3) 1999,
pp.417–56.
Nag, R., D.C. Hambrick and M.-J. Chen, ‘What is strategic management,
really? Inductive derivation of a consensus definition of the field’, Strategic
Management Journal 28 2007, pp.935–55.
2.1 Introduction
Strategic management is a relatively young academic discipline. Among
the first publications are Alfred Chandler’s Strategy and structure (1962),
H. Igor Ansoff’s Corporate strategy (1965) and Kenneth Andrews’ The
concept of corporate strategy (1971). Since then the field has changed
its focus from business policy to strategic management and made the
transition from being a collection of toolkits developed by consulting
firms to a systematic, theory-driven academic field. In addition, right
from its beginning the area of strategic management was recognised as an
interdisciplinary research field. The field was – and is still – populated by
scholars from different disciplines like management, economics, finance,
marketing, psychology and sociology. For such a diverse field it might have
been difficult to develop a consensual meaning of what the discipline is all
about. However, such a shared understanding is important as all academic
fields are socially constructed and can only flourish if there is a shared
conception of its meaning. Thus, the purpose of this chapter is twofold:
first, we will sketch the evolution of the field before analysing the shared
understanding of the field.
2.2 Early theories
The field of strategic management or, to be more precise, the field of
business policy (as it was initially called) did not emerge before the 1960s
with the influential work of Alfred Chandler, Igor Ansoff and Kenneth
Andrews. These early writings were influenced by the work of Edith
Penrose, who developed a theory of the growth of the firm by emphasising
the importance of a firm’s resources and managerial capabilities for its
growth. Besides Penrose, these researchers were influenced by the work
of the Carnegie School, especially Herbert Simon, Richard Cyert and
James March, who developed the idea of bounded rationality to study the
decision-making process in firms. Deviating from standard economics,
Simon and his colleagues assumed that decision makers do not have
9
MN3119 Strategy
complete information and that the alternatives they are deciding upon
needs to be researched. The last important influence of the early writings
was the work by Philip Selznick and his idea on distinctive competences.
A distinctive competency is something that is unique to an organisation
and superior in some respects when compared with the competencies of
other organisations that offers some value to their customers. Influenced
by these ideas, Chandler, Ansoff and Andrews developed their theories of
strategy by placing an emphasis on the internal characteristics of a firm.
In Strategy and structure, for example, Alfred Chandler studied how large
organisations develop new administrative structures to accommodate
growth and how strategic change influences an organisation’s structure.
2.3 Michael Porter and the industrial organisation
paradigm
After relabelling the field ‘strategic management’ in the late 1970s, the
focus moved towards industrial organisation economics in both theory and
method. At this time research was aiming to develop and test hypotheses
derived from the structure-conduct-performance (SCP) paradigm. The
basic idea of this paradigm is that the performance of a firm is determined
by the industry in which it competes. The conduct of a firm is just a
reflection of the industry structure, so that the most important decision
a firm has to make is in which industry it wants to compete in. The SCP
paradigm led to a shift in focus from the firm to the industry or market
structure. The most influential scholar from this era is Michael Porter, who
is not only well known among researchers but also among practitioners.
He developed the so- called ‘five forces’ model that specified different
features of an industry and which determines its attractiveness and
facilitates competitor analysis. Porter also proposed the idea of generic
strategies – low cost leadership, differentiation and focus – to match the
characteristics of an industry and achieve a competitive advantage.
Besides the SCP paradigm, research at this time developed the idea of
strategic groups. Strategic groups are groups of firms in the same industry
who follow the same strategy, for example, all suitcase producers in the
high-price market like Louis Vuitton or Bottega Veneta.
A third important research theme, influenced by industrial organisation
economics, is competitive dynamics (which is subject of Chapter 13).
The basic interest of competitive dynamics is to investigate how firms
are jostling for competitive advantage by carrying out different types of
strategic and tactical actions.
2.4 Organisational economics
In the 1970s the field of strategic management was not only influenced
by the work of industrial economists but also by another subfield of
economics: organisational economics. Organisational economics tries to
open the ‘black box’ of the firm and looks at its inner structural logic and
functioning. Its most prominent theories are transaction cost economics
and agency theory.
Transaction cost economics, developed by the work of Oliver Williamson,
studies the question of why firms exist and which transactions are made
inside the market and which inside the firm. Building on the concept
of bounded rationality and asset specificity, Williamson argues that
transactions are made within the firm (or hierarchy, in the words of
Williamson) when the sum of all transaction costs is smaller than the price
10
Chapter 2: The evolution of strategic management as an interdisciplinary field
for the transaction in the market. He further argues that firms exit as an
efficient alternative to the market. Strategic management scholars have
applied transaction cost theory to study the make or buy decision of all
firms, including multidivisional and international firms.
Agency theory claims that the separation of ownership and control in
most organisations lead to a problem of divergence of interest between
shareholders (‘principals’ in the words of agency theory) and managers
(called agents). The idea is that managers try to maximise their utility,
which might not be in line with the maximisation of the long-term profits
of the firm. To avoid this problem, firms need contracts that rule the
relationship between principal and agents.
2.5 Resource-based view, dynamic capabilities and
leadership theory
One of the latest steps of the development in the field of strategic
management occurred with the emergence of the resource-based view and
its dynamic extension – the dynamic capability approach. This stream of
research is influenced by the work of Edith Penrose, who viewed the firm
as a collection of productive resources. The resource-based view further
argues that these resources are heterogeneously distributed among firms.
The differences in resources combined with the imperfect mobility of the
resources can explain the differences in performance among firms in the
same industry. The dynamic capability approach extends this logic and
argues that a firm needs dynamic capabilities to modify and extend its
resource base to achieve and sustain a competitive advantage or a series of
competitive advantages over time.
Finally, in 1984 Don Hambrick and Phylis Mason developed the strategic
leadership or upper echelons theory. They developed a theoretical
framework that proposed senior executives base their strategic choices
on their cognitive structures and values. Hence, strategic leadership
theory tries to explain differences in firms’ performance by, for example,
differences in past performances of their executives, top management
team size, composition and tenure.
2.6 A consensus definition of the field
As seen in this short historical overview, strategic management has been
advanced by researchers from different disciplines, especially economics,
organisation science or sociology. Such an interdisciplinary field often
lacks a clear understanding of or consensus on the subject. However,
without this consensus, a field is limited in its growth. As this is at odds
with the great success of the field of strategic management over the last 30
years, Nag et al. conducted a study among strategy scholars from different
disciplines asking them about their personal views of the field of strategic
management. In their large study among strategy researchers, they
found the following implicit definition of the field: The field of strategic
management deals with (a) the major intended and emergent initiatives;
(b) taken by general managers on behalf of owners; (c) involving
utilisation of resources; (d) to enhance the performance; (e) of firms; (f)
in their external environments. They therefore concluded that the success
of the field emerges from an underlying consensus that enables it to attract
multiple perspectives.
11
MN3119 Strategy
2.7 Key concepts
•
Strategic management
•
Business policy
•
Structure-conduct-performance paradigm
•
Transaction cost economics
•
Agency theory
•
Resource-based view
•
Dynamic capabilities.
2.8 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
12
•
explain how the field of strategic management evolved over time
•
discuss the elements of the shared meaning of the strategic
management field.
Chapter 3: Analysis of market structure
Chapter 3: Analysis of market structure
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
discuss the most common techniques used to define a market
•
describe a firm’s market power within a defined market.
Essential reading
Besanko, D., D. Dranove, M. Shanley and S. Shaefer. Economics of Strategy (New
Jersey: Wiley, 2009) pp.205–11.
or
Lexecon Ltd An introduction to quantitative techniques in competition analysis.
Lexecon Ltd. publication, mimeo. Available at: www.crai.com/ecp/assets/
quantitative_techniques.pdf.
Saloner, Shepard and Podolny Strategic Management. (New Jersey: Wiley, 2005)
Chapter 6.
Further reading
Emmons, W. and R. Prager ‘The effects of market structure and ownership on
prices and service offerings in the US cable television industry’, Rand Journal of
Economics 28(4) 1997, pp.732–50.
Geroski, P. ‘Thinking creatively about markets’, International Journal of Industrial
Organisation 16(6) 1998, pp­.677–95.
3.1 Introduction
In this chapter we will first present techniques which are used when
defining markets, for instance, for policy or firm strategy analysis. In
the second part of this chapter we introduce some methods of analysing
markets with many firms.
The process of market definition and analysis will be an input for the later
chapters in the second part where we will learn how to analyse tightly
structured problems, for example, by solving games having identified the
players, the strategies, the pay-offs and the rules.
3.2 Techniques of market definition1
Why should we want to define the market for a particular product or firm?
Geroski (1998) states that there are three ways of ‘thinking creatively
about markets’: trading markets, antitrust markets and strategic
markets. Trading markets are defined as geographical areas and a set of
products for which the law of one price holds to a certain degree. That is,
taking into account transport costs and slight differences in the product
offerings, we would expect a similar price to be charged for similar
products. Market definition from an antitrust perspective looks at the
likelihood that the market can be monopolised by a single firm, or group
of firms. That is, if there was some change in the degree of the market’s
competitiveness (for example, through mergers or collusive agreements),
would it be possible to initiate a ‘small and significant increase in prices’,
This section is based
mainly on Lexecon
Ltd An introduction to
quantatative techniques
in competition analysis.
1
13
MN3119 Strategy
or is market power restricted by other, closely related markets? Finally,
strategic market definition concerns itself with market boundaries
which are defined by firms’ product offerings. The argument here is that
companies not only adapt to but also create or segment markets in order
to maximise their profits.
We will introduce a number of techniques of market definition: estimating
the cross-price elasticity of demand, price correlation analysis, switching or
diversion ratio, shock analysis and bidding studies.
Example: Sony’s PlayStation 3 (PS3), a video game console
The PS3 was introduced in the year 2006 with the express intent of providing a successor
to the Playstation 2 and to challenge Microsoft’ s early mover position in the market for
256-bit game consoles. The simple definition of the market would suggest that the PS3’s
market is defined by ‘game consoles’. Hence, its main competitors are the Nintendo Wii,
the Microsoft X-Box 360, and Sony’s own Playstation 2 to some extent. Is this the whole
story? To start with, PCs nowadays have relatively advanced gaming features as well and
there is a large library of PC games available, which would suggest that PCs are also a
significant competitor also. Going even further, if young people no longer find playing
game consoles attractive, they could switch to other ways of entertaining themselves like
watching TV, reading books, playing board or ball games, etc. Another way of thinking
about the market for the PS3 is by looking at the buyers, who are more often than not
parents around Christmas time. What would they be spending their money on if not on
the PS3? This could be other consumer electronics as gifts or clothing, etc.
3.2.1 Cross-price elasticity of demand
You will recall the (own-price) elasticity of demand (e) from your
introductory microeconomics course: it is defined as the ‘relative change in
quantity (Q) demanded due to a price (P) change’. We can write the ownprice elasticity as:2
−
∂Qi
Qi
∂Pi
Pi
or −
∂Qi Pi
∂Pi Qi
The cross-price elasticity then measures the ‘relative change in quantity
demanded of one good, due to a one per cent price change of another
good’. We can write the formula as:
∂Qi Pj
∂Pj Qi
That is, if the price for another good goes up by one per cent, how will
demand for my own product change (in percentage terms)? If a product
is a close substitute, we would expect cross-price elasticity to be a large
positive number: a price increase of five per cent for a rival product will
redirect a lot of customers towards my own product, meaning that demand
goes up by more than five per cent. On the other hand, if a price decrease
for a potential substitute results in only small quantity losses, cross-price
elasticity is low and we can say that the products are distant substitutes.
This simple technique cannot be used to define or test a market. However,
it is a very powerful tool to assess the relationship of two products if the
data is available.
14
2
We leave out the
negative sign commonly
used in own-price
demand elasticity.
This is simply because
we expect cross-price
elasticity to be positive
for substitute products.
This is however just a
convention and nothing
should be read into it.
Chapter 3: Analysis of market structure
Activity 3.1
Give your estimates of the cross-price elasticities of the following product pairs and
explain why.
a. Two gas stations on opposite sides of the road
b. Coffee and tea
c. Hotels in Bahrain and New Zealand
Guidance on this activity can be found in the VLE.
3.2.2 Price correlation analysis
Quite often it is difficult to gather enough data to calculate the cross-price
elasticity of demand – in particular, obtaining a time series of quantity and
price data containing some small price changes and very little changes in the
general market parameters is typically difficult.
An (imperfect, but still acceptable) alternative may be tracking movements
of prices over an extended period of time. As we will see in Chapter 5, the
B and C competition model shows prices as strategic variables are strategic
complements, which means that, on the one hand, if one firm increases its
price, so will the other. In contrast, if quantities are the strategic variable
(and the strategies therefore are strategic substitutes), we expect market
prices emerging from the quantities set by the firms to move in unison for
related products. In the extreme case of perfect substitutes, the market price
for both will be the same, so an increase in price for one product implies a
price increase for the rival product.
A word of caution on the interpretation of price correlations: it is important
to rule out other explanations for the co-movement in prices. For example,
if the prices of ice cream and sunscreen are highly correlated, this does not
imply that the two are close substitutes – if anything, they are complements,
but demand for both of them is affected by the same (seasonal) components,
temperature and sun. Similarly, tyres and washing-up liquid are rarely
mistaken for substitutes, but changes in their prices are likely to occur at
somewhat similar times – both products use oil as a significant input3 and
are therefore likely to be affected by oil price changes in a similar way. The
point is that price correlations have to be interpreted with care and that
potential sources of ‘spurious correlation’ (i.e. correlation that is not down
to the reason stipulated) have to be taken into account. It is also important
to get a sense of the reasons for the price changes, even if they were not
‘spurious’ in the sense outlined above – were there any product or process
innovations (we will cover this later in the guide), was there a significant
shift in consumer preferences or did firms simply change their strategies?
3
A more comprehensive
list of products made
from oil is given on
www.anwr.org/features/
oiluses.htm
3.2.3 Switching/diversion ratio analysis
If a time series of prices is not available or if for some reason would not be
meaningful (for example, if one product is priced at zero, e.g. a software
programme available as freeware on the internet), it may be useful to ask
consumers directly for the products or services they perceive as the closest
substitutes to the products they are currently using.
Frequently, this will take the form of a survey of consumers of a particular
product or service who are asked a question to the tune of: ‘If the product
you have just purchased had not been available today, what other option
would you have chosen?’ Clearly, the more ‘votes’ a particular alternative
obtains, the closer a substitute it is likely to be.
15
MN3119 Strategy
3.2.4 Shock analysis
Another reliable method of determining the closeness of two products
or services is their reaction to a ‘shock’. Technically, this is a natural
experiment observing a sudden and unexpected change in one market
and analysing the reaction in the (possibly) related market. Shocks can
take many forms: entry of a new product, technological change, price
concessions, a change in input prices, natural crises, etc. For example,
when Sony released the Playstation 3 in 2006, board game manufacturers
made only limited changes to their strategies (and probably experienced
only a limited impact on their sales), whereas Nintendo and Microsoft
were hit relatively hard, suggesting that the game consoles market consists
of closer substitutes than the broadly defined ‘leisure’ industry.
3.2.5 Bidding studies
Finally, there may be cases where prices are not transparent or not
publicised, or where the number of transactions is relatively low so that
no meaningful correlations can be gathered from the data. Sometimes
therefore, the best that can be done is to determine ‘who bids for the
same business’ as a proxy of who competes in the same market. For
example, when the ‘big four’ accounting firms were still the ‘big six’, the
effect of merging PriceWaterhouse and Coopers & Lybrand was assessed
using a bidding study. On the one hand, it was found that in addition to
PriceWaterhouse and Coopers & Lybrand, the other four accountancy
firms were bidding for very similar projects or accounts, which meant that
competition was likely to be intense even after the two merged. On the
other hand, if the two merging partners had been a duopoly in a submarket, say, large manufacturing accounts in the north-east of England,
the merger would have led to a monopoly in the north-east manufacturing
sector. So, while bidding studies are a relatively crude way of determining
one’s competitors, they are still a useful exercise.
3.3 Market analysis with many firms
Suppose now that we have defined the market using one (or multiple)
of the techniques introduced above. What next? We need to find some
proxies for determining whether the market is competitive or not, in-order
to judge, for instance, how attractive the market is likely to be in the near
future, how likely it is that antitrust action will be taken, and if entry or
exit by rivals can be expected.4
Figure 3.1 gives the four most commonly used measures of
competitiveness and concentration in an industry. Their quality and
accuracy increase from number one to four:
The simplest way to measure the competitiveness of a market is by
counting the number of firms in an industry. In a market with many
firms, it is less likely that a single firm will have a significant amount of
market power. Further, more firms suggest rather low entry barriers (the
more firms there are in the industry, the lower the average size or market
share per firm), which is another indicator for a competitive market. If
there are some firms, however, that do have significantly higher market
power than others in the industry, a simple count would not do the trick.
16
4
We will discuss
entry in more detail in
Chapter 14. For now it is
sufficient to know that a
less competitive market
is likely to be more
profitable for potential
entrants, which in turn
implies that incumbents
will try and build barriers
to entry to maintain
their profitable position.
Chapter 3: Analysis of market structure
Measures of competitiveness and concentration
1
n
1. #firms
C≡
2. C4/5/8
C4 ≡ ∑ si
4
i =1
3. Herfindahl index
n
H ≡ ∑ si2
i =1
n
4. Lerner index
L ≡ ∑ si
i =1
p-MC i
p
Figure 3.1: Measures of competitiveness and concentration.
A more useful method, particularly if there are several larger firms, is to
calculate the Cn-ratio (mostly C4, C5 or C8) – the sum of the market shares
of the n largest firms. This gives some information about how strong the
biggest firms are likely to be. Again, however, this measure is not fully
satisfactory: Imagine a market with 4 firms – the C4-ratio will be 100 per
cent regardless of the distribution of market shares between the four firms.
Similarly, if the n+1th largest firm is almost as big as the nth firm, the
Cn-ratio will not pick this up. So even though we can do a little better than
simply counting the firms, not all information is utilised in the Cn-ratio.
The Herfindahl-Hirschman index (HHI) takes the market shares
of all the firms in an industry into account and sums their squares. This
solves a lot of the problems of the previous two concentration indices:
first, information of all firms is taken into account,5 and second, larger
firms feature more prominently in the index. The HHI is a standard tool
for antitrust economists and the US Department of Justice guidelines state
that ‘…markets in which the HHI is between 1,000 and 1,800 points are
considered to be moderately concentrated, and those in which the HHI is
in excess of 1,800 points are considered to be concentrated. Transactions
that increase the HHI by more than 100 points in concentrated markets
presumptively raise antitrust concerns under the Horizontal Merger
Guidelines issued by the US Department of Justice and the Federal Trade
Commission.’6
The one remaining problem with the HHI is that while it says a lot about
the size distribution of firms within an industry, it does not say anything
about the way and intensity with which these firms compete, or, in
economists’ parlance, their conduct. The Lerner index, then, takes
the HHI one step further and looks at a firm’s profit margins (i.e. price
– marginal cost) and weighs them by the firm’s market share. In other
words, if the largest firm charges a high price relative to its marginal cost
while smaller ones price relatively aggressively, the Lerner index will be
higher than if a small firm charged high mark-ups (e.g. because it operates
in a niche market). Note, however, that it is still possible for margins to be
high and profits to be low due to fixed costs. The main problem with this
measure, however, is that it is almost impossible to gather the necessary
information on prices and particularly costs for each firm.
In practice, this is
often impossible with
a large number of very
small firms. Antitrust
practitioners mostly
ignore the market shares
of all firms with a market
share of <1 per cent.
This seems acceptable
since they would have
a minute effect on the
HHI anyway. On the
other hand, if the fringe
of small firms can be
expected to act jointly
(or at least to react in
the same way to, say, a
price increase of a larger
firm), there may be a
case for squaring the
joint market share of all
the fringe firms. This will
depend on the particular
case in question.
5
See www.usdoj.gov/
atr/public/ testimony/
hhi.htm
6
17
MN3119 Strategy
3.4 Key concepts
•
Cross-price elasticity of demand
•
Price correlation analysis
•
Switching/diversion ratio analysis
•
Shock analysis
•
Bidding studies
•
Cn-ratio
•
Herfindahl-Hirshman index
•
Lerner index.
3.5 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
discuss the most common techniques used to define a market
•
describe firm’ market power within a defined market.
3.6 Sample examination questions
1. You are a product manager for a computer game and have been
asked to analyse whether a new game is a competitor and to suggest
techniques which could help inform your decision. The game you are
in charge of is called ‘Industry Giant’ and has the following product
description:
‘Begin in the year 1900 with little money but large ambitions and
through skilful decision-making you can build up an enormous
business empire. Make critical decisions on which of more than 200
products you should manufacture, where to gather the best raw
materials, where to sell them and how to effectively transport them
there. Recommended Retail Price (RRP): £29.99’
Industry Giant has a multiplayer option, in other words, you can play
it against others on the same PC, but it has no internet capabilities.
Your competitors are a combination of computer-controlled teams and
(if you play with friends) ‘real’ players. Internet Industry Manager
(IIM) is a relatively new computer game where players get to ‘manage’
a firm in an interactive environment by choosing prices, quantities
and advertising efforts. It is played over the internet and all players
are ‘real’ players, that is, every firm is controlled by somebody on the
internet. It is free to download and to play.
Is IIM a competitor? Why or why not? What techniques would you use
specifically to analyse the degree of competition between the products?
2.
You are working for a Japanese firm selling hot coffee in vending
machines. You have been asked to identify the market and the
potential competitors and substitutes for your product in Europe.
a. How would you design and conduct a study to do this (outline
your choice of techniques and the reason why you chose this
particular set of techniques)?
b. What is your first guess of your product’s relevant market?
Guidance on answering these Sample examination questions can be found
in the VLE.
18
Chapter 3: Analysis of market structure
Extended activity: the commercial banking industry in
the United States
Read the following and answer the questions at the end.
The US commercial banking industry consists of independent banks and
bank holding companies that provide firms and homes with depository
and lending products, as well as a range of other related financial services.
Simply put, the difference between the interest paid on deposits and the
interest earned on loans determines the profitability of commercial banks.
This sector provides a good example of how market structure affects
performance by influencing conduct – how the players compete with each
other.
Most banks are involved in retail banking, serving households and small
firms, as opposed to corporate banking, which serves large businesses and
offers more sophisticated financial services. The retail banking sector is
characterised by having many, relatively small and geographically local
markets. People tend to choose banks that are physically close to where
they work or live, and, once they become customers, they tend to view
their bank as a primary provider of a range of personal finance services.
Gaining customer trust and forming strong customer relationships are
critical to banks as access to and knowledge about customers enhances the
prospects of cross-selling other products and services.
Banks operating within markets with a higher concentration – where the
market is shared between fewer players – exercise a greater market power,
for example in setting fees and prices. Conversely, in local markets with
lower concentration – where the market is shared by many competitive
players – banks tend to have a lesser market power.
Historically, the banking industry has been shaped by regulations that
have restricted the geographical scope of operations. Banks, for example,
needed to maintain a head office in their local market where deposits were
collected and loans originated. Moreover, they were often not allowed
to form branches in other regions of their state, and were completely
restricted from establishing branches in other states.
In the 1990s many of these regulations were relaxed, prompting a
significant industry consolidation on a national scale. Forming these
larger banks was believed to yield many benefits including the ability to
reap greater returns on investments in technology and advertising, to
consolidate back-office functions, and to specialise employee functions in
conjunction with the diversification of products and services.
This merger activity within and across state boundaries had a great impact
on the concentration levels of banks on the national scale, where we have
seen much fewer – but significantly larger – players. In local markets,
however, the concentration levels did not change significantly due to
antitrust regulations that continued to restrict mergers between banks in
direct local competition. The changes in market structure on the national
level, however, do affect the competitive landscape of local markets in
that local branch players are often, through their larger bank parents, in
competition in several distinct local markets at the same time.
Industry consolidation has created a trend of ever-larger banks controlling
ever-larger market shares. Evidence shows that small banks are generally
less profitable than larger banks, but that there is little difference in
profitability within the large bank category. This suggests a threshold on
the spectrum of bank size beyond which further gains in efficiency and
ability to charge higher prices cannot be sustained.
19
MN3119 Strategy
Source
Pilloff, S.J. ‘Commercial banking’ in Adams W. and J. Brock (eds) The structure
of American industry. (New Jersey: Prentice Hall, 2001), pp.224–54. [ISBN
9780130189166] pp.224–54.
Questions
1. How would different market concentration levels affect bank prices
and efficiency through the respective levers of market power and
competition? Map and describe the relationships between these
factors. If you were a bank owner, would you prefer to operate in a
market with high or low concentration? Why?
2. Analyse the factors influencing how banks compete in today’s local
markets. Consider market concentration levels, bank size, the types
of products and services that can be offered, the ‘stickiness’ of the
customer relationship, entry barriers.
3. What do you think are the potential implications of national banks
competing through branches in several distinct local markets where
the conditions and relative competitive positions may differ. How can
they exploit their national muscle for local gain, and what might the
limitations to this be?
20
Chapter 4: Introduction to game theory and strategy
Chapter 4: Introduction to game theory
and strategy
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
describe simple games, for instance the prisoner’s dilemma, using the
normal form and extensive form
•
define the concept of Nash equilibrium and find equilibria for a game.
Essential reading
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) Chapter 4.
Further reading
Camerer, C. ‘Redirecting research in business policy and strategy’, Strategic
Management Journal 6(1) 1985, pp.1–15.
Dixit, A. and S. Skeath Games of strategy. (London: Norton & Company, 2004)
second edition Part II.
Saloner, G. ‘Modeling, game theory, and strategic management’, Strategic
Management Journal (12) 1991, pp.119–36.
Saloner, G., A. Shepard and J. Podolny Strategic management. (New Jersey: Wiley,
2005) Appendix.
4.1 Introduction
There are about as many definitions of strategy as there are textbooks
on the topic. The definitions often have very little in common and are
frequently coloured by the preferences of the author. Instead of advancing
another definition with a claim for universal applicability, let us identify
a number of features the many definitions have in common. Specifically,
strategies are often:
•
long-term
•
about securing competitive advantage
•
coupled with actions, not just a plan.
First of all, a strategy is supposed to have a long-lasting effect on a firm
– in other words, we are not talking about small, day-to-day transactions
(even though these may be in accordance with an overall strategy).
Second, a strategy is often aimed at securing an advantageous position
relative to other firms in the market. That is, there is an element of
competition in many of the definitions of strategy. Finally, strategies are
often a set of actions rather than a set of lofty ideas and visions.
What would count as a strategy? Surely, the most basic strategy must be
whether you want to enter a (geographical or product) market or not.
Once you have entered, setting capacities or production volumes
will determine the long-term success of your firm. Once this is settled, you
have to sell your product. How to sell? Setting prices, choosing (to the
extent that you can do this) qualities and advertising levels would
21
MN3119 Strategy
be the most obvious levers a firm can pull in order to sell its product.
There are other decisions a firm has to make, however: should the firm
integrate vertically (i.e. with suppliers or buyers) or horizontally
(with competitors), should the firm spend resources on research
and development, and if so, what kind of research should be done?
Common to these is that the profitability of these actions depends on
what your competitors do: lowering your price will only be successful if
your competitor does not lower his by the same amount. Introducing a
high-quality version of your product will only draw consumers from your
competitors if they do not counter this by launching their own high-end
version as well. Entering a market on your own is much more profitable
than with a competitor, and so on. For formulating a successful strategy,
it is therefore important to at least try and anticipate what (actual or
potential) competitors are going to do.
4.2 Static games
4.2.1 Anticipating rivals’ moves
In strategic analysis, it seems important to be able to figure out what one’s
rival is going to do, that is, to anticipate a rival’s moves. How can we do
this? Strategists (both professional ones (i.e. managers and consultants)
and ‘strategists in the making’ (i.e. students)) often assign probabilities
to the different actions a rival might take. But can we do better than this?
Game theory tells us we can (most of the time)! Let us use an example to
illustrate this.
Prisoner’s dilemma – Advertising wars
Consider the following situation. P&G and Colgate Palmolive sell
competing brands of toothpaste – Crest and Colgate – in a market. The
brands share the market equally, that is, both firms have a 50 per cent
market share. The overall market for toothpaste is fixed – let’s assume
total sales for toothpaste will be €10m per year. Both firms now have
the option of launching an advertising campaign for one year at a cost
of €2.5m. While advertising does not increase total sales for toothpaste,
advertising if the other firm does not advertise would increase market
share to 80 per cent. What should both firms do? In fact, can we use the
information above to make a prediction of what each of the players is
going to do?
The first step to take is identifying the players (which is typically a
simple but nonetheless important step). In this case, P&G and Colgate
Palmolive are the players. The second step is to identify their strategies,
that is, what are the choices they have? Here, P&G and Colgate Palmolive
decide whether to run an advertising campaign or not. Third, we have to
specify the rules of the game. We will go into this in more detail later on,
but here the rules are that both players have to decide simultaneously to
run or not to run an advertising campaign. Finally, we have to specify the
pay-offs for each of the possible outcomes of the game. In this case, both
firms not running an ad campaign results in them sharing the market and
making sales of €5m. If one of them runs an advertising campaign and the
other does not, the first makes sales of €8m (80 per cent of €10m), but has
to pay advertising costs of €2.5m, leaving net sales of €5.5m. The other
(non-advertising) firm makes sales of €2m. If both firms advertise, their
sales will be €5m again (as the market for toothpaste is fixed), but they
again incur an advertising cost of €2.5m, resulting in net sales of €2.5m.
22
Chapter 4: Introduction to game theory and strategy
We can now represent the game in matrix, or normal form. First,
however, a word on convention: The first pay-off (or the leftmost) will be
the pay-off of the row player – that is, the player choosing the row (top or
bottom), Colgate in this case. The second pay-off (or the pay-off further to
the right) will be the column player’s pay-off (i.e. P&G).
P&G
No Ad
€5m
No Ad
€5m
Colgate
Ad
€5.5m
€2m
€2m
Ad
€5.5m
€2.5m
€2.5m
Figure 4.1: Advertising wars.
The situation P&G and Colgate are facing is what game theorists call a
prisoner’s dilemma: overall profits would be higher if both did not
advertise, but both have an individual incentive to go ahead and advertise
anyway – regardless of what the other player does (advertise or not
advertise).
This makes advertising a dominant strategy – a strategy that does
better than all others for any strategy chosen by the other player(s). This
gives us the first way of predicting our rival’s behaviour: if a strategy
always maximises my rival’s pay-offs, he will play it.
Eliminating dominated strategies – pizza wars
It would be nice if we could always make a clear prediction about our
rival’s behaviour. We can often simplify games by finding dominant
strategies as demonstrated above. However, consider the following game.
There are two restaurants in a small town, Dave’s Deep Dish and Paul’s
Pizza Pies. They are in competition with each other for customers and can
choose their prices: high (H), medium (M) or low (L). The city has 1,000
customers, of which 300 only ever buy at Dave’s, and 300 only buy at
Paul’s. The other 400 are price-sensitive and always buy the cheaper pizza
and choose at random if they charge the same price. Both places make a
margin of £12 per pizza if they charge high prices, £10 per pizza if they
charge medium prices, and £5 for low prices. Both Dave and Paul cannot
observe what the other player has chosen before they choose themselves.
Can we draw this in a pay-off matrix? We know players, strategies and
rules. What about calculating pay-offs? We can calculate profits by
multiplying the number of customers with the margin per customer. For
example, if Dave charges a medium price and Paul a high price, Dave will
sell to his 300 ‘loyal’ customers and the 400 ‘price sensitive’ customers at
a margin of £10 each, giving him pay-offs of £7,000. Paul only sells to his
300 loyal customers, but at a margin of £12 per pizza, giving him profits
of £3,600.
23
MN3119 Strategy
Paul
High Medium Low
High
Dave Medium
Low
60/60
36/70 36/35
70/36
50/50 30/35
35/36
35/30 25/25
Figure 4.2: Pizza wars.
Can we solve this game by finding dominant strategies? For none of the
players can we find one strategy that does better than the other two for
all strategies chosen by the rival. For example, M outperforms H if the
other price is H or M, but not if it if L. Dominant strategies will not get
us very far then. We can, however, find a strategy that never does better
than another one – a dominated strategy: charging L is worse than H or
M for all strategies chosen by the other player. So if Dave and Paul are (as
we assume) trying to maximise their profits, they will never charge lower
prices. We combine our outcomes on dominant and dominated strategies
for the following set of predictions.
Result
a. A dominant strategy should always be played.
b. A dominated strategy will never be played.
Let’s now go one step further. Having simplified the game, we can in fact
find a dominant strategy again. Playing M always does better than H,
knowing that L will never be played by either of the players. This process
is called the iterated elimination of dominated strategies. This
makes the game a good deal simpler, and you can now solve the simplified
game with just M and H for Dave and Paul. What is the solution?
4.2.2 Nash equilibrium
Hopefully you found that the solution to the simplified Pizza game is
(M,M) – that is, both players will charge a medium price. Looking for
dominant strategies in the simplified game gave you this outcome. This
combination of prices is a Nash equilibrium (NE). How did you try to ‘solve’
the game? Assuming that both players are interested in maximising their
pay-offs, you tried to predict what they would do given the other player’s
strategy. As it turned out, in the simplified game, playing M was the
best solution, regardless of the strategy of the other player – a dominant
strategy. This means that both players playing M is an equilibrium – none
of the players would want to change their behaviour given the other’s
strategy. Therefore, a Nash equilibrium can be formally defined as follows.
Nash equilibrium – definition
A Nash equilibrium is a combination of strategies such that no individual player can
deviate unilaterally from his/her current (equilibrium) strategy to improve his/her pay-offs.
This means that a game has a solution at a Nash equilibrium in the sense
that no player would change given the other player’s strategy. What does
this imply about the relationship between Nash equilibrium, dominant and
dominated strategies? Here are a few questions to consider.
24
Chapter 4: Introduction to game theory and strategy
Activity 4.1
Answer the following questions:
a. Is a Nash equilibrium the same as a dominant strategy?
b. Can a Nash equilibrium contain dominated strategies?
c. Will every Nash equilibrium contain dominant strategies?
Guidance on this activity can be found in the VLE.
Hopefully, your answer to all these questions was ‘No’. Why? As to the first
question, a Nash equilibrium refers to a combination of strategies, which
by definition means it’s more than just a single strategy by one player.
(Another objection is that not every Nash equilibrium must consist of a
combination of dominant strategies – as we discuss in c). The second ‘No’
comes directly from the first prediction we made: if a strategy is never
played, it cannot be a resting point – there is always something better for
the player to do. The final ‘No’ becomes clear when we study the Pizza
game. Looking at the full game, we could see that none of the strategies
is dominant, but we still found a Nash equilibrium in (M,M). Here are
another two questions.
Activity 4.2
Answer the following questions:
a. Will every game have a Nash equilibrium (in pure strategies)?
b. Can a game have more than one Nash equilibrium?
Guidance on this activity can be found in the VLE.
The answer to the first question is again ‘No’ (Hint: think of a game
of ‘matching pennies’, where one player wins if two pennies are both
heads or both tails, and the other wins if one is heads and the other is
tails.) Let us tackle the second question by using an example: Science
and Humanities faculties have to decide on the renovation of an unused
lecture hall. Both want something to be done with it, but have their own
preferences: Science prefers a laboratory, whereas Humanities prefers a
theatre. At the main meeting of all the faculties, both have to endorse one
of the projects. At the preliminary meetings within the individual faculties,
both have to decide which project to endorse. The meetings take place
simultaneously – or, if you are more comfortable with this interpretation,
the meetings take place without the faculties knowing what the other
faculty has decided.
First, let’s represent the game in matrix (or normal) form:
Humanities
Lab
Theatre
Lab
Science
Theatre
Figure 4.3: Coordination game.
25
MN3119 Strategy
(Try to put in sensible pay-offs in the four quadrants – remember
that what’s most important to both players is spending the evening
together, even though each of the two players has a preference for one
of the activities.) What will be the Nash equilibrium of the game? Well,
predicting the other player’s move is tough: putting ourselves in Science’s
shoes, if Science thinks that Humanities will endorse the lab, Science will
want to vote for the lab as well. But when would Humanities vote for the
lab? If it thinks that Science will vote for the lab as well – which will be
the case if Science thinks that Humanities will…the story could go on and
on, and we could run the same procedure for voting for the theatre!
So will this mean that the game has no Nash equilibrium? In fact, there
are two: (L,L) and (T,T) – both want to coordinate on one of the projects
and they have no incentive to deviate unilaterally (i.e. taking the other
player’s strategy as given) if they have coordinated. Note that this does
not mean that the faculties are indifferent between the two solutions –
Science would rather see a lab built than a theatre, but not at the expense
of risking any project being realised. We will deal with ways of solving the
conundrum of which of the two equilibria to choose shortly, but for now
we know that some games can have multiple Nash equilibria.
Activity 4.3
Represent the following situation as a strategic game and solve it, that is, look for the
Nash Equilibrium.
Two players, Arthur and Bob, drive towards each other in their cars. Both can swerve (i.e.
steer to the left or right) or go straight. Whoever swerves first loses and is considered
‘chicken’ (i.e. not very courageous) and the rival wins. If both swerve, they are both
chickens (but they do not face the abuse by the winner of the game). If they both go
straight, the cars collide and Arthur and Bob are seriously injured.1
Guidance on this activity can be found in the VLE.
Some games, as we have seen, have multiple Nash equilibria, with no
obvious way of selecting among them. Of course, one might think that
Scientists are a stubborn bunch who would only ever vote for a lab
anyway, but then who is to say that the Humanities faculty is not just as
stubborn? And if that is the case, would the Science faculty not want to
compromise instead of seeing the unused lecture theatre go unused for
another year?
We will now go through a few mechanisms or ways in which games
with multiple Nash equilibria can be approached. Note, however, that
most of these will imply changing the rules of the game in some way –
in other words, you have to make assumptions in addition to the initial
specifications of the game.
Hierarchy. If there is a ‘junior’ and a ‘senior’ player in a strategic game,
the game is played as if the senior player’s preferences are more important.
For example, if there is a merger or a takeover, firms will typically try and
agree on a common corporate culture (i.e. a coordination game). Although
it is possible in principle that either firm’s corporate culture will prevail, it
is most common that the larger or more powerful partner will dictate the
eventual culture.
Commitment. If one of the players commits to playing a certain strategy,
it is in the other’s interest to follow, that is, play the game as if the
committed player had already moved. For example, in a ‘chicken game’,
such as entry into a natural monopoly market, if one firm commits to
entering this market (for example, by building production facilities specific
26
1
This is a variant of the
'chicken game' played
in James Dean's Rebel
without a cause.
Chapter 4: Introduction to game theory and strategy
to this market or hiring experts for a particular technology on long-term
contracts), other firms might be deterred by this and refrain from entering
themselves.
History. If there has been a ‘traditional way’ of playing a particular game,
new versions of the game will automatically ‘converge’ on the historical
equilibrium. For example, there is no real method of choosing whether to
drive on the right side or the left side of the road – apart from selecting the
option others in the country choose.2 Also, if a coordination game between
players has traditionally been solved by taking turns of one party’s
preferred option and then the other party’s, future versions of the game
would take this into account.
Mixed strategies. If none of the above work, the best shot players
might have of ending up with an acceptable outcome is to randomise their
behaviour. We will not go into this in much detail, but the intuition is to
take into account the likely actions of the other player and one’s own payoffs and choose accordingly. For the moment, it is sufficient to recognise
that sometimes playing ‘randomly’ is the best way of maximising one’s
expected pay-offs from a game with multiple equilibria. In fact, mixing
behaviour such that each player is indifferent between the possible options
is another equilibrium concept – a mixed strategy equilibrium. This
is in contrast to the Nash equilibria we have been looking for so far – socalled pure strategy equilibria, which involve non-random behaviour,
but deterministic choice of one strategy.
2
Of course, there is also
a legal issue here, but
even without any legal
recourse, no individual
driver would have an
incentive to choose
a different side than
all others in the same
region.
4.3 Dynamic games
In the remainder of this chapter we will briefly discuss dynamic games,
and in particular the difference between static and dynamic games.
Dynamic games are, put simply, games with a time aspect in them.
For example, if one firm acts before the other, this has quite important
implications for playing the game: the second firm can play the game
knowing what the first firm has done, whereas the first firm has to make
its decision without the requisite knowledge about the follower. Some
games simply don’t make much sense to play sequentially – paper/scissor/
stones, for example, would not be very exciting if one player knew what
the other player has chosen.3 Some games, on the other hand, could be
played either simultaneously or sequentially. Setting prices, for example,
will be done without knowledge of rivals’ prices some of the time (making
it a simultaneous game), but in other situations sequential moves might be
more relevant.
Representing a sequential game is usually done by drawing a game
tree, where the first decision starts the game, and every decision point
represents a node from which the decisions of subsequent players branch
out accordingly. We illustrate this with an example. Assume that two
firms selling mineral water have to decide on their advertising budget.
For simplicity, there are only two levels of advertising, High and Low.
Harrogate Spa chooses its advertising level first, and Vittel chooses after
that. If both firms choose H, profits are zero (because all the money is
spent on advertising), if both choose L, their profits are £1m each, and if
one chooses H and the other L, the firm running an intensive advertising
campaign makes £1,250k, while the other one makes profits of £500k. A
game tree (or extensive form) of this situation would then look like
Figure 4.4.
For those unfamiliar
with paper/scissor/
stone, have a look at the
website www.netlaputa.
ne.jp/~tokyo3/e/
janken_e.html for an
explanation and its
equivalent in many
different cultures.
3
27
MN3119 Strategy
Game Tree
H
L
Harrogate Spa:
H or L?
Vittel:
H or L?
H
0/0
L
H
1,250k/500k 500k /1,250k
L
1m/1m
Figure 4.4: Game tree.
How should one analyse a sequential game? Analysing sequential games
has a very similar objective to analysing simultaneous ones: predict sensible
behaviour and an eventual outcome of the game. The way to do this in a
simultaneous game is by eliminating dominated strategies and/or playing
dominant strategies. (While this may enable us to simplify the game, it may
still not lead us to a Nash equilibrium. In this case we would have to look for
Nash equilibria after having simplified the game as much as we could.) One
way of simplifying a sequential game is by backward induction. Backward
induction works by eliminating strategies at the final node of the game (i.e.
the point when the last player makes a decision, based on the decisions
previously taken), and working one’s way forward, that is, closer to the
start of the game. The strategies we can eliminate are moves that would not
maximise an individual’s profit at that point. A rational player should never
select these moves, which means that a player anticipating rivals’ moves
should not expect these moves to be chosen.
Looking at the game referred to above we can see that Vittel would not
choose H if Harrogate Spa plays H, and they would not choose L if Harrogate
Spa chose L. This then enables Harrogate Spa to anticipate that playing H
gives profits of £1,250k, while playing L yields profits of £500k – Harrogate
Spa should then choose high advertising expenses, which Vittel will react to
by choosing low expenses.
Activity 4.4
Represent the previous game as a normal-form game, if Harrogate Spa and Vittel select
advertising expenses at the same time. What is/are the Nash equilibrium/a?
Guidance on this activity can be found in the VLE.
Commitment
We already mentioned commitment in the context of selecting among
different possible equilibria in a static game. Commitment, however, can also
be used by an agent to choose the preferred course of action that he would
not otherwise choose. Again, we use an example to illustrate this point.
28
Suppose a monopolist (M) in a market faces a potential entrant (E). The
entrant can choose whether to enter (e) or not (ne). If E does enter, the
monopolist can choose to fight (f) or to accommodate (a). If E does not
enter, M experiences business as usual. Suppose M currently makes profits of
PM = 50 (and E makes zero profits in this market), in the case of entry and
accommodation the entrant makes PE = 10, the incumbent PM = 20, and if
entry is followed up by fight (think of this as the monopolist starting a price
war flooding the market), profits are PM = PE = – 10.4 Figure 4.5 represents
this game in game tree form.
Note that the profits
should be considered
long-term profits and
the entry is a one-off
opportunity.
4
Chapter 4: Introduction to game theory and strategy
Game Tree
e
E: e or ne?
M: f or a?
f
-10/-10
ne
0/50
f
10/20
Figure 4.5: Game tree.
By backward induction we now find that the Nash equilibrium is for E to
enter and for M to accommodate – if entry does occur, M will choose payoffs of 20 over –10, and, knowing this, will prefer entering to not entering.
The monopolist will not be particularly happy with this outcome: of
course, given the choice, he would rather keep the entrant out – for
example, by committing to fight in the case of entry. To this end, M could
issue a statement along the following lines:
‘Should entry occur in our market, we will fight aggressively to
protect our market position.’
This seems sensible enough, and should go some way towards convincing
the entrant not to enter. Or should it? The entrant would have to believe
that M would indeed prefer to fight. This would imply, however, that in
the case of entry M would choose a sub-optimal action, namely one that
gives him pay-offs of –10 rather than 20. In game-theoretic terminology,
fighting after entry is not sub-game perfect – a sub-game starts at one
player’s decision node and covers all the decisions that follow on from this
node, and an action that does not maximise pay-offs at that decision node
should not be played. Since post-entry fighting is not sub-game perfect, the
entrant should not believe the monopolist if he makes this statement.
Let us now consider another strategy by M. Suppose that M sign and
publicise a long-term contract with one of suppliers that states: ‘if we, M,
ever purchase less than the current quantity, we will incur contractual
penalties of 40.’5 This seems like an odd move to begin with, since all
that it achieves is to lower pay-offs in some cases, but it never increases
pay-offs. However, it changes things around in the particular game we
are analysing. Accommodating implies sharing the market, that is, selling
less and consequently buying less from one’s supplier. This means that
it is now less attractive to accommodate, since M would have to incur
contractual penalties, taking down profits from accommodating from +20
to –20. This changes the pay-offs and the way the game is being played –
it is now sub-game perfect to fight after entry, which means that the
entrant has to choose entering and incurring losses or keeping out and
have unchanged profits. The entrant will choose to stay out, leaving the
incumbent with profits of PM = 50.
We assume here that
it is not possible to
renegotiate this contract.
5
What happened? By limiting his options, the monopolist was able to
commit to playing the game differently, which accordingly made the
entrant play the game differently as well, taking M’s expected reaction into
account. If we represent this in our game tree (Figure 4.6), the monopolist
changed the pay-offs in one game outcome and subsequently the sub-game
perfect equilibrium.
29
MN3119 Strategy
(New) Game Tree
No contract
Contract
e
e
ne
ne
0/50
f
-10/-10
a
10/20
0/50
f
a
-10/-10 -10/-20
Figure 4.6: (New) Game tree.
In this game, committing credibly was worth 30 – the difference between
the outcome of the game with commitment and the outcome without.
Committing can be done by limiting one’s options, or lowering pay-offs in
some (undesired) states of the game.
Activity 4.5
What are other examples of credible commitments? Explain how an organisation (or an
individual) limits his/her options by lowering pay-offs in one particular state of play.
Guidance on this activity can be found in the VLE.
4.4 Key concepts
•
Prisoner’s dilemma
•
Dominant/dominated strategies
•
Iterated elimination of dominated strategies
•
Nash equilibrium
•
Backward induction
•
Normal and extensive form games
•
Sub-game perfect equilibrium.
4.5 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
describe simple games, for instance the prisoner’s dilemma, using the
normal form and extensive form
•
define the concept of Nash equilibrium and find equilibria for a game.
4.6 Sample examination questions
1. Can you find some examples from business life of coordination games?
Of ‘chicken’ games?
2. Simple coordination games have two pure-strategy Nash equilibria.
What are ways of selecting among these equilibria?
3. Consider the advertising game in Section 4.3. As you can see, there is a
first-mover advantage for Harrogate Spa. Develop an advertising game
in which there is a second-mover advantage.
4. Simultaneous games and Nash equilibria. Consider the
following games. The first two are zero-sum games in that one player’s
30
Chapter 4: Introduction to game theory and strategy
gain is the other’s loss, which is why we can write only player 1’s
pay-offs. The other two are non-zero sum. Find all pure-strategy
Nash equilibria, and demonstrate if you could identify dominant and
dominated strategies.
a)
b)
c)
d)
31
MN3119 Strategy
5. Producing cars.6 Consider a normal form game between three major
car producers, C, F, and G. Each producer can produce either large
cars or small cars, but not both. That is, a car producer’s actions
(strategies) are ai = SM or LG. We denote profits for firm i as πi(aC, aF,
aG), i.e. they depend on the decisions of all three car producers. The
profit function is defined as follows:
πi =
{
γ if all firms play the same strategy (i.e. all SM or all LG)
α if i plays a strategy on its own (i.e. ai = SM,aj = LG or
vice versa)
β if i play a strategy with one other firm, but the
remaining firm does not
Answer the following questions:
i.
Does there exist a Nash equilibrium when a > b > g > 0?
Prove your answer.
ii. Does there exist a Nash equilibrium when a > g > b > 0?
Prove your answer.
6. Wage dispute. Consider a situation where an employer and a union
are involved in an industrial dispute – a union member is claiming
unfair dismissal and demands compensation of £10k. The case has
gone to a tribunal which will decide in favour of one or the other with
50 per cent probability. The tribunal can only award the full £10k to
either of the parties and cannot propose a compromise. Further, this
verdict is binding, that is, both parties have to abide by it. Both parties
now can choose a strategy (L) to increase their chances of winning
– hiring a lawyer, gathering evidence for or against unfair dismissal,
bribing the judge, etc. Playing L comes at a cost of £4k per player.
Choosing L gives a success probability of 100 per cent only if the other
player does not play L too. If both players choose L, the advantages
cancel out and the chances of winning are unchanged at 50 per cent.
Represent this game in normal form and find the Nash equilibrium.
Guidance on answering these Sample examination questions can be found
in the VLE.
32
This exercise is
taken from Shy,
Industrial organization.
(Cambridge: MIT Press,
1995), p. 41.
6
Chapter 5: Oligopolistic models of competition
Chapter 5: Oligopolistic models of
competition
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
describe and compare the Bertrand and Cournot models of
simultaneous, oligopolistic competition
•
explain the Stackelberg leadership model of sequential competition.
Essential reading
Besanko, D., D. Dranove, M. Shanley and S. Shaefer Economics of Strategy. (New
Jersey: Wiley, 2009) pp.221–29.
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) Chapter 7.
Haskel, J. and C. Martin ‘Capacity and competition: empirical evidence on UK
panel data’, Journal of Industrial Economics 42(1) 1994, pp.23–44.
Further reading
Church, J. and R. Ware Industrial Organisation: A Strategic Approach. (New York:
McGraw-Hill, 2000) Chapter 8.
Leonard, R. ‘Reading Cournot, reading Nash’, Economic Journal 104(424) 1994.
5.1 Introduction
A 1999 Wall Street Journal article said that:
…in a strategic shift in the US and Canada, Coca-Cola is gearing
up to raise prices by about 5%. The price changes could help
boost Coke’s profit…Important to the success of Coke is how
Pepsi-Cola responds. The No. 2 soft-drink company could well
sacrifice some margins to pick up market share on Coke, some
analysts said.
It is clear from the preceding quote that the profitability of one firm’s
strategy (Coke raising prices) depends crucially on the response of its
closest rival (Pepsi reciprocating or not). We can think of many situations
like this: capacity expansion and the following production increase will
only be profitable if there is sufficient demand, which in turn depends
on the strategies of producers of competing products. Lower prices will
only generate the expected market share gains if other firms do not
simultaneously lower their prices. The list could easily be extended – more
interesting, however, is the fact that these problems have a strikingly
similar structure to the problems we addressed in the previous chapter:
There are players (firms) with a certain set of strategies or actions
(prices, capacities, etc.), whose pay-offs (profits) are interdependent,
that is, they depend on one’s own actions and those of one’s rivals. In
short, such problems of firms interacting in (narrowly defined) product
markets could be modelled as a strategic game.
Such situations are called oligopolies – loosely translated as markets
with few players. That is, they are not monopolies, markets with only
33
MN3119 Strategy
a single player, or perfectly competitive markets, where one player’s
action has no impact on the profits of others (because every single firm is
assumed to be too small to matter in the grand scheme of things).
The literature on oligopoly competition is vast, and we cannot hope to
cover all the details of the numerous different models. We will focus on
the two most basic models – Bertrand and Cournot competition – and
then discuss two extensions that are especially relevant for applications in
strategy – Stackelberg leadership and product differentiation.
5.2 Preliminaries
The first step in any problem of oligopoly competition will be to derive the
profit function. Profits are simply Revenues – Costs. Revenues are price x
quantity, and costs are average variable cost multiplied by quantity plus
fixed cost. For this chapter, we will assume that marginal cost is constant
and that there are no fixed costs. Denoting marginal cost as c, this then
gives us the following profit function:
Pi = piqi – ciqi = (pi – ci)qi
Of course, this all seems very simple, but we need to keep in mind that
the price a firm can achieve for its product may depend on the overall
quantity produced in the market (so that the firm is not a price-setter, but
a quantity-setter among other quantity-setters). On the other hand, firms
could be setting prices, in which case the quantity they sell would depend
on the prices charged by their competitors.
Also, to fix ideas, in the first part of the chapter, we will use a simple
demand function:
Q=5–P
Note that this is a market demand function – Q is the total quantity
produced and demanded, while we will write q for the quantity an
individual firm produces and sells. Where convenient, we will rearrange it
to P = 5 – Q (the inverse demand function). For most of our discussions,
we will assume marginal costs of MC = 2.
5.3 Bertrand competition
When asked about the key strategic variable by firms, most students would
opt for prices. After all, product quality is typically given, demand by
consumers derives directly from it, so firms can only try and set the price
that maximises profits. The basic Bertrand model of competition captures
this in a very simple setting: two firms produce identical products and
choose the price simultaneously, and then produce to satisfy demand.
Since the products are identical, consumers can be expected to purchase
purely on the basis of price – whoever charges the lower price will obtain
all the market demand Q, and if prices are identical the consumers are
distributed evenly across both firms. This gives us a three-part profit
function, depending on one’s own price and the rival’s.
π1 =
34
{
(p1 – c)Q
if
p1 < p2
(p1 – c)Q/2
if
p1 = p2
if
p1 > p2
0
If they only interact once, this makes the incentives of the two players
quite clear: If player 1’s price exceeds player 2’s, he makes no sales –
clearly not desirable. So lowering the price to match player 2’s might be an
option as player 1 could take half of the market. However, lowering prices
Chapter 5: Oligopolistic models of competition
even further by the smallest currency unit available would hardly mean
a significant decrease in the price, but an enormous increase in quantity
– player 1’s market share would increase from half the market to the full
market.
Both players will therefore want to undercut their rival by the smallest
amount possible for any range or ‘normal’ prices. But what do we mean by
normal? Let us determine the upper and lower bound of the ‘undercutting’
regime, that is, the range for which firms will keep on undercutting their
rivals by the smallest amount possible.
The upper bound is the monopoly price – let us use an extreme
example. If firm 2 charges a price of 15, which would lead to zero sales in
any case, firm 1 has no incentive to marginally undercut it. In fact, since
firm 2 has priced itself out of the market, firm 1 is best off by behaving
as a monopolist and charging (for the specific demand functions we have
specified) a price of 3.5.1
Activity 5.1
Check that the monopoly price is indeed 3.5.
(Hint: derive the profit function and maximise it.)
Guidance on this activity can be found in the VLE.
The lower bound is own marginal cost. If firm 2 charges marginal
cost and marginal costs are symmetric, undercutting would mean selling
to the entire market, but losing money on every unit that sold. Therefore,
undercutting will stop at P = MC.
We make a distinction
here between
'charging lower prices'
and 'undercutting'.
'Undercutting' means
that a firm charges a
price that is lower than
the rival's price, but by
the smallest possible
amount.
1
What does this all mean for the market? We can see that no firm would
want to charge a price above the monopoly price, and no firm wants to
be undercut in the range between monopoly price and marginal cost – a
decrease in price just below the rival’s price would give higher profits.
The only sensible Nash equilibrium is for both firms to charge marginal
cost: increasing prices (assuming the other firm remains at marginal cost)
leads to zero sales and therefore zero profits, while decreasing prices
would increase sales but every unit will incur a loss because price is below
marginal cost. If both firms charge marginal cost, therefore, there is no
incentive to deviate unilaterally – the condition for a Nash equilibrium.
We can therefore state the following:
Result
In a price-setting market with two symmetric firms selling undifferentiated goods, the
unique Nash equilibrium is marginal cost, which is the competitive outcome.
This is slightly troubling since we quite frequently see two firms active in
a market, obviously setting prices but making quite healthy profits. As we
will see, changing some of the assumptions means that the competitive
outcome need not always emerge in duopolistic markets.
For a formal (and
very complicated)
analysis of this problem,
see Kreps, D. and J.
Scheinkman 'Quantity
precommitment and
Bertrand competition
yield Cournot outcomes',
Bell Journal of
Economics 14, 1983,
pp. 326–37.
2
5.4 Cournot competition
We now look at the same situation (i.e. same demand curve, same
symmetric marginal cost), but where firms set quantities rather than
prices. We can think of this as a two-step setting: firms set quantities first
and then prices. It can be shown that firms will produce the quantity they
expect to sell in the second stage, and they set prices in the second stage to
clear everything they produced in the first period.2
35
MN3119 Strategy
The major difference to the previous model is that there are no ‘jumps’
in the profit function – if my rival produces slightly more than I do,
market prices will be slightly lower, but sales or prices will not jump
discontinuously as in Bertrand competition.
Firm 1’s profit function is determined by the quantity firm 1 produces, but
also by firm 2’s quantity (via the market price, which depends on both
quantities).
π1 = (5 – q1 – q2 – c)q1
We then take the first derivative of the profit function with respect to our
decision variable, q1 and set it equal to zero to find the maximum.
δ1
δq1
q*1 =
= 3 – q2 – 2q1 = 0
3 – q2
2
;q2* =
3 – q1
2
⇒ q1* = q*2 = 1
We find that, not surprisingly, my optimal quantity depends on the
quantity the other firm sets – after all, every unit of output lowers market
price, and if my rival is already producing a lot, high output by two players
might depress market price even more.
We then have to derive and solve the same problem for firm 2 to find
the Nash equilibrium. Recall that the definition of a Nash equilibrium
is a situation when both players have no incentive to deviate from their
current behaviour given the other player’s behaviour. That is, if
player 2 sets quantity q2, player 1’s best response will be to set q1*, and vice
versa. After all, if your quantity is a best response to the quantity I set as a
best response to your quantity, neither player has an incentive to deviate.
As we can see, quantities in the Cournot case are 1 for both players, which
implies a price of 3, which exceeds marginal cost. We can conclude the
following:
Result
In a quantity-setting market with two symmetric firms selling undifferentiated goods,
prices will be above marginal cost and below the monopoly price.
5.5 Comparing Bertrand and Cournot
Perhaps the most striking difference is that despite being based on the
exact same demand and cost curves, price competition does not generate
any profits, while quantity competition does. This can be interpreted in a
number of ways, but the main intuition is that firms react differently to a
change in the other firm’s strategy.
In the case of price competition, a change in price by firm 1 will lead
to a change in price by firm 2 in the same direction. That is, if one firm
increases their price, the other will reciprocate by increasing theirs,
but possibly by a slightly lower amount. This is called strategic
complements – not to be confused with (product) complements, which
describes the nature in which the demand of two products are related to
each other.
On the other hand, if firms compete in quantities, if firm 2 increases its
quantity, the best response by firm 1 is to decrease its quantity. Analogous
to the previous situation, this is called strategic substitutes. This
reaction may seem strange initially, but inspection of the reaction curve
36
Chapter 5: Oligopolistic models of competition
q1* = (3 – q2)/2 shows that as q2 goes up, q1 should go down. A more
intuitive reason for this is that firms are still trying to maximise their own
profit, not market share or anything else. This means that if one firm
commits to producing large quantities of a product, competitors will not
(should not) be tempted to further lower market price by putting out large
quantities of output as well. Instead, their incentive to maximise profits
will dictate a decrease in their own quantity.
Strategic complements and substitutes – definition
If the best response to an increase in the strategic variable of a rival is an increase in
one’s own strategic variable, the strategic variables are strategic complements.
If the best response to an increase in the strategic variable of a rival is a decrease in one’s
own strategic variable, the strategic variables are strategic substitutes.
Activity 5.2
Think of the following strategic variables. Which ones are strategic complements or
substitutes, and under which circumstances?
a. Advertising spend
b. Research and development effort
c. Entry decisions in small niche markets
Guidance on this activity can be found in the VLE.
5.6 Stackelberg leadership
Quite frequently, firms do not set prices or quantities (or R&D efforts,
advertising efforts, or any other strategic variable) simultaneously, but
sequentially. For example, an early entrant in a market will have an
opportunity to set production capacities before any potential entrants. We
will now analyse oligopoly games with sequential moves, that is, games
with so-called Stackelberg leadership.
The mechanics of a Stackelberg game are relatively simple: instead of
two firms choosing quantities/prices simultaneously and having to infer
the other’s actions, Stackelberg games have one leader choosing first,
and the leader’s choice will then be observed by the follower, who reacts
accordingly (i.e. following her best-response function).
We use Cournot competition as an example, so that the Stackelberg leader
sets quantities first and the follower reacts to it. We have learned from the
previous chapter that sequential games have to be solved by backward
induction, that is, by solving the follower’s problem first.
The follower’s problem is relatively easy to solve because we have
done it before – by finding the best-response function to any quantity by
the other player. Therefore, we know that if the leader sets quantity qL, the
follower’s best response will be
q*F =
(3 – q1 )
2
Now on to the leader’s problem. The leader does not have to secondguess what the follower would do – knowing the reaction function is
enough to conjecture that the quantity chosen by the follower will be
qF* = (3 – qL)/2. This leaves the leader with a residual (inverse)
demand curve of
37
MN3119 Strategy
which gives a profit function of:
ПL =
(
)
3 qL
–
q
2
2 L
Conveniently, this is a function of just a single variable – the leader’s
quantity. This lets us compute the optimum quantities for leader and
follower:
3; * 3
qL* =
q =
4
2 F
Looking at the optimal quantities just derived, we can see that with
uniform market prices and symmetric marginal cost, the leader makes
higher profits than the follower since it has a higher market share (in fact,
twice as much). It is also interesting to note that overall quantities are
higher in the Stackelberg case (3/2 + 3/4 = 9/4) than in the symmetric
Cournot case (1 + 1 = 2) – in other words, competition is more aggressive
with Stackelberg leadership (and prices are lower). This confirms our
intuition about strategic substitutes: if one player commits to a high
quantity (the leader), the other player would not want to reciprocate with
high quantity, but lower quantities to offset the effect to some extent. The
offset is not complete however – the follower will not decrease quantities
to the same extent that the leader increases its own.
With Stackelberg leadership in quantity competition we can therefore say
that there are first-mover advantages. In the exercises, you will be asked
to show that this does not hold for price competition – but for now we
note that first-mover advantages depend on the nature of competition, in
particular, if strategic variables are strategic substitutes or complements.
5.7 Product differentiation
Finally, we go back to one of the implications that generated the extreme
result in the price-setting model, namely that the lower-priced firm takes
all the demand. This is a result of the assumption that products offered by
the firms are identical. Rational consumers should therefore buy from the
lowest-priced firm or choose at random if prices are equal. Realistically,
there will always be slight differences, either real or perceived ones. For
example, some customers might prefer a certain type of salesperson,
or they prefer a closer shop to a more distant shop, or they prefer one
package design to the other, even if the product is essentially the same.
The important point to note is that the products will not lose all demand if
they have a slightly higher price than that of their rival.
Suppose for example, that there are different coefficients on ‘own price’
and ‘rival price’ in the case of imperfect substitutes:
40 4
2p
q1 =
– p + 2
3
3 3 1
This demand function indicates that an increase in own price has a
negative effect on demand (the coefficient is – 4/3), while an increase in
the rival’s price has a positive effect on demand, which seems intuitive as
the two products are assumed to be substitutes. However, the coefficient
on own price is larger in magnitude than the rival price coefficient, which
indicates that a change in the rival’s price will not have an effect of the
same magnitude as a change in own price. In the example above, a one
unit increase in own price will decrease the quantity sold by 4/3, while
a one unit increase in the rival’s price will increase units sold only by
2/3. Thus, the extent of differences in magnitudes indicate how close
substitutes two products are – a zero coefficient on the rival’s price
38
Chapter 5: Oligopolistic models of competition
obviously means that demand does not depend at all on the price of
another good.
We can now solve the game in much the same way as our previous
oligopoly games, that is, deriving the profit function for both firms,
taking the derivative and setting it equal to zero to find the best-response
function, and inserting one best-response function into the other to
find the Nash equilibrium. It turns out that for the demand function we
specified and zero marginal cost, we obtain the following result:
q1 =
40 4
40 4
2
2p
– p + 2; q2 =
– p + p
3 3 1
3
3 3 2 3 1
δ1
40 8
2p
=
– p1 + 2;
3 3
3
δq1
p1 = 5 + p2 ; p2 = 5 + P1
4
4
As prices are above marginal cost (20/3 versus 0), profits will be positive,
even with price competition. This is because now both firms have an
incentive to keep prices high even if the rival undercuts. Consider for
example a firm facing a rival setting a price of zero. According to the bestresponse function we have derived, the firm should set a price of 5 since
p1* = 5 + p2/4. That is, firm 1 would willingly concede consumers to firm
2 and make positive profits on the ones that remain.
To summarise, we can interpret the coefficient on own prices as ‘price
sensitivity’ and the relative magnitude of the rival price coefficient as an
indicator for the degree of substitution between the two products.
5.8 Key concepts
•
Bertrand competition
•
Cournot competition
•
Bertrand v Cournot
•
Strategic complements and substitutes
•
Stackelberg leadership.
5.9 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
describe and compare the Bertrand and Cournot models of
simultaneous, oligopolistic competition
•
explain the Stackelberg leadership model of sequential competition.
5.10 Sample examination questions
1. Consider the demand system in Section 5.7 (Differentiated goods).
Solve a game in which firm 1 is a Stackelberg (price) leader and firm 2
a follower. Is there a first-mover advantage?
2. Boeing and Airbus produce more or less homogeneous aircraft for the
world market. Let p be the price for a wide-body aircraft. Boeing’s
output level is denoted by qB, and Airbus has output level qA. The
aggregate industry output is denoted by Q, where Q = qB + qA.
39
MN3119 Strategy
The aggregate industry demand curve for this product is given by
p = 2000 – Q. Assume that the unit cost of firm 1 is cB = 300 and the
unit cost of firm 2 is cA = 500.
a. Derive the reaction functions of both firms. Then solve for
the Nash equilibrium when firms set quantities (i.e. Cournot)
simultaneously.
b. Solve a quantity-setting model where Airbus is a Stackelberg
leader. Highlight the differences to a).
c. Which assumptions do you consider unrealistic in this model? How
do you think the model would change if they were incorporated?
4. a. Show that in quantity competition a firm would produce the
monopoly quantity if the other firm were to exit the market.
b. In a quantity-setting model, show that prices cannot go below the
competitive level.
Guidance on answering these sample examination questions can be
found in the VLE.
Extended activity: articles on Airbus-Boeing competition
Read the following and then answer the question at the end.
Reading
‘For their next trick’, The Economist 25 March 1999. Available at:
www.economist.com/node/193875
‘Mumbo Jumbo’, The Economist, 23 September 1999. Available at:
www.economist.com/node/242880
Question
Read the two articles on the Boeing/Airbus competition. Airbus is
proposing to launch a new aircraft with higher capacity than Boeing’s 747
(Jumbo). How might this affect competition between the two? Frame your
answer (also, but not exclusively) in terms of price or cost leadership, cost
asymmetries and changes in the demand function.
40
Chapter 6: The resource-based view of the firm
Chapter 6: The resource-based view of
the firm
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
link resources and distinctive capabilities
•
identify some sources of competitive advantage
•
describe the process of capability building.
Essential reading
Barney, J.B. ‘Firm resources and sustained competitive advantage’, Journal of
Management 17 1991, pp.99–120.
Dierickx, I. and K. Cool ‘Asset stock accumulation and sustainability of competitive
advantage’, Management Science 35(12) 1989, pp.1504–11.
Kay, J. Foundations of Corporate Success: How Business Strategies Add Value.
(Oxford: Oxford University Press, 1995) Chapters 5–8.
Wernerfelt, B. ‘The resource-based view of the firm’, Strategic Management Journal
5(2) 1984, pp.171–80.
Further reading
Barney, J.B. ‘Resource-based theories of competitive advantage: A ten-year
retrospective on the resource-based view’, Journal of Management 6(2001a),
pp.643–50.
Peteraf, M.A. ‘The cornerstones of competitive advantage: A resource-based view’,
Strategic Management Journal 14(3) 1993, pp.179–91.
6.1 Introduction
Although the field of strategic management is occupied by researchers
from different disciplines – like management, economics, psychology or
sociology – they all share an interest in answering one question: Why do
some firms consistently outperform their competitors? In line with our
analysis, a traditional industrial organisation economist, who believes in
the structure-conduct-performance paradigm, would answer that firms
who picked the ‘right’ industry are able to outperform other firms (where
‘right’ means an industry in which a firm can earn monopoly rents).
(Remember that the profit in the monopoly setting is always higher than
the profit in a duopoly or perfect competition.) However, if you ask a
management scholar interested in strategy, they would probably argue
that explaining variation in firm performances by differences in the
competitiveness of industries is rather dull and that it is more interesting
to study why performance differences exist between firms in the same
industry. Thus, to be a bit more precise, management scholars are more
interested in the question: Why do some businesses succeed in the same
industry environments where others fail? rather than the general question
stated above. Over the last 20 years the most prominent management
literature dealing with this question has focused on the resource-based
view (RBV) of the firm, which is the topic of this chapter.
41
MN3119 Strategy
6.2 Competitive advantages and resources
Before we explain the main ideas of the resource-based view of the firm,
we will first define what a competitive advantage is.
Competitive advantage – definition
A firm possesses a competitive advantage if it can consistently outperform other players
in the industry.
This definition may seem rather obvious, but it is important to note
two elements. First, it refers to other players in an industry. That is,
competitive advantage is by definition a comparative concept. Second, and
most importantly, it refers to consistently outperforming other players.
As we have already mentioned, one of the most influential management
theories to explain competitive advantage is the resource-based view of
the firm. Although some of its ideas can be found in the seminal work of
Edith Penrose from the late-1950s, it was not until the mid-1980s that the
idea was made popular by the work of Jay Barney and Birger Wernerfelt.
The core idea of the RBV is that firms can be seen as a collection of
heterogeneous and immobile resources and capabilities and that this
heterogeneity of resources and capabilities can explain differences in a
firm’s performances.
A resource is anything that can be thought of as a strength or weakness
of a firm. More formally, a firm’s resources can be defined as the assets
that are tied to the firm. Examples of resources include machinery, capital
and production facilities as well as brand names, management skills
and organisational routines. Resources can be divided into two broad
categories: tangible and intangible resources. Tangible resources, on the
one hand, are assets that can be observed and quantified, such as financial,
organisational, physical and technological resources. Intangible resources,
on the other hand, are assets that are deeply rooted in the history of the
firm, accumulated over time and which are difficult to quantify or observe.
Intangible resources include human resources, innovation or reputational
resources. Further capabilities include an organisation’s capacity to deploy
resources.
As not every resource and capability leads to superior performance, the
argument behind the RBV is that a resource must fulfil four different
criteria to result in a competitive advantage. Resources and capabilities
give sustained competitive advantage when they are: valuable, rare,
imperfectly imitable and not substitutable. First, to be valuable, resources
must yield a superior product or service or lower costs. Second, they must
be rare to ensure that the resource holders cannot compete away the
value they create. Third, resources and capabilities must be imperfectly
imitable (i.e. competitors must not be able to imitate them). Resources
and capabilities are harder to imitate if tacit knowledge is involved in
the resource or its application or if they are complex. Fourth, resources
and capabilities must be not substitutable (i.e. there are no strategic
equivalents, using either the same resource or an equivalent one). The
four characteristics mentioned are individually necessary, but do not
comprise sufficient conditions for a sustained competitive advantage.
In the next section we will talk about four specific examples of resources
and how they contribute to a firm’s competitive advantage. Kay (1995)
thinks of these resources as the main sources of competitive advantage.
42
Chapter 6: The resource-based view of the firm
6.3 Some examples of resources as sources of
competitive advantage
6.3.1 Contractual architecture
Architecture – definition
Architecture is the set of structure, style and routines which a firm employs in dealings
with employees and other firms.
The architecture, in other words, is the set of spot, long-term and
relational contracts that a firm is involved in. Roughly, we could divide
legal contracts (spot and long-term contracts) as defining the structure
of the firm, and relational contracts as defining the style and routines of
a firm. This is not completely accurate as the choice of contract will also
guide the structure of the firm to some extent.
Example
Volker Finke is the longest-serving football manager in German professional football.
He has been manager of SC Freiburg, a team from a mid-sized German city near the
French and Swiss border, from 1992 to 2007 – an eternity in professional football. After
a few years of signing formal contracts for one or a few years, Volker Finke and the club
president, Achim Stocker, stopped signing a formal contract governing Finke’s role in the
club, preferring instead a handshake at the start of each season. This absence of a formal
(i.e. spot and long-term) contract is indicative of the style of the club, which has for a
long time cultivated its image as ‘alternative’, with open, informal structures, no ‘stars’, a
relatively flat salary structure and many non-pecuniary perks.
On the other hand, Bayern Munich, Germany’s most successful club, is
known to insist on formal contracts being signed between player and club.
Several years ago, one of their star players, Mehmet Scholl, lost his place
in the first team and threatened to leave if he did not get to make more
appearances. The response of Uli Hoeness, Bayern’s manager at the time,
was that, having signed a contract, Scholl should realise that he was a
Bayern player, whether on the field, on the bench or in the stands (i.e. out
of the match day squad).
We have to distinguish between internal and external architecture. Internal
architecture encompasses formal employment relations and career paths,
but also in corporate culture and job security – that is, the ‘soft’ factors
of working for an organisation. External architecture captures supplier
contracts, but also implicit agreements about flexibility over and above
what has been agreed in the contract, knowledge sharing arrangements
and/or ‘first-point-of-contact’ assurances.1
We should note that even though spot and long-term contracts can
secure a firm a superior position in its industry, architecture is likely
to be sustainable because that architecture includes a whole set of
interdependent contracts that cannot easily be identified and copied
by competitors, and the informal or relational contracts a firm has
are even more difficult to replicate.
‘First-point-of-contact’
assurances promise that
if a new opportunity
in a similar field arises,
the existing supplier
within that field will be
approached first.
1
6.3.2 Reputation
Reputation – definition
Reputation is an aggregate of many personal and collective judgments about a
company’s or an industry’s credibility, reliability, responsibility and competence based on a
common set of values.
43
MN3119 Strategy
It is easy to imagine why reputation might be a source of competitive
advantage. It takes a long time to build up an asset stock like reputation,
and it might be very difficult, possibly even prohibitively expensive, to try
and match the reputation of the market leader. The precarious nature of
reputation as a competitive advantage is also reflected in the following
quote:
If you lose dollars for the firm by making bad decisions, I will be
very understanding. If you lose reputation for the firm, I will be
ruthless.
(Warren Buffet, remarks made to Salomon Brothers managers,
1991)
Why is reputation so important and in what markets? It turns out that
it is mainly the nature of the product in question that determines how
important reputation is.
More precisely, reputation is invaluable in markets where product quality
is important, and can only be determined through long-term experience.
These goods are commonly called ‘experience goods’, and a firm’s or a
product’s reputation can serve as a signal for quality without a consumer
having to purchase the product in the first place. Similarly, ‘search
goods’ are goods where quality is important and can only be ascertained
by inspecting the goods personally. This implies that if there are a lot of
variants around, ‘searching’ for the right product might prove prohibitively
costly and reputation can serve as a substitute for the search. We do this
virtually every time we go shopping and choose for a special brand, a
specific supermarket, a favourite restaurant, because sampling all options
available would be too time consuming.
Activity 6.1
Which of these are experience or search goods?
•• a cure for baldness
•• a hamburger
•• stocks
•• funeral services
•• skirts.
6.3.3 Innovation
We distinguish between two concepts here: innovation and innovativeness.
Innovation – definition
Innovation relates to a new product that is valuable to consumers and provides a firm
with an opportunity to outperform others in the market.
Innovativeness – definition
Innovativeness is a firm’s ability to continuously invest and market new products.
At first sight, you would expect that innovativeness is more sustainable
than innovation because it refers to a capability that it is possible
(although difficult) to build up and copy. However, it might also be the
case that a firm with a constant stream of innovations is not able to
protect its innovation and consequently will not be able to outperform
its competitors consistently. The key, therefore, for both innovation and
innovativeness is how to safeguard the value created from an innovation.
There are several alternatives.
44
Chapter 6: The resource-based view of the firm
First, legal protection through patents and copyrights gives inventors
a (usually limited) time period during which they have exclusive rights to
bring their invention to the market. Second, building up a reputation
as the first or most innovative product on the market can protect the
value of an invention. Third, standard-setting can be a very powerful
mechanism for securing the benefits of an innovation.2 Finally, if none of
these is likely to work for an innovator, another way is to adopt a policy
of licensing the innovation. That way, revenues can be generated
through license fees instead of own sales.
Activity 6.2
Find an example for each of the four mechanisms of protecting the value created by an
innovation.
However, all the mechanisms mentioned above are risky – they may be
circumvented by other firms or simply may not work as well as envisaged.
The best way therefore to derive value from an innovation is to support it
with other capabilities. For example, supporting innovation with a set of
contracts, both internal and external, is likely to generate a sustained
advantage. Launching a product with large, concurrent investments in the
firm’s reputation will create a first-mover advantage that is hard to copy
with another innovation. Finally, strategic assets like patents or the
long-term monopoly over a key input for the new product are useful ways
to obtain income from innovations.
2
If a firm sets the
industry standard,
other firms either have
to adopt the standard
by licensing specific
components from the
innovator, or they have
to engineer around
the industry standard,
leading to inferior
products in comparison
with the original.
The profits to be gained from innovativeness are often dependent on a
firm’s architecture: relational or formal employment contracts that reward
‘innovation from within’, in other words, the workforce, is likely to enable
the firm to generate innovations continuously. Similarly, an external
architecture consisting either of very flexible spot contracts, which do not
tie a firm in to a particular product line or supplier over a long period,
or relational contracts, which encourage suppliers to react quickly and
constructively to sudden changes in requirements, are complementary to
an innovative environment within the firm.
6.3.4 Strategic assets
Strategic asset – definition
A strategic asset enables a firm to maintain a favourable market position because its
competitors must use less productive or profitable substitutes.
Firms therefore benefit from the ownership of strategic assets, but in
contrast to the other sources of competitive advantage, there is no reason
why another firm – if these assets were transferred – would not do equally
well. It is therefore not entirely clear if strategic assets should be classed
as a capability in the first place, as it does not really describe a capability
that is unique to the firm. What it certainly is, however, is a source of
competitive advantage, which is why we include it here.
There are numerous strategic assets a firm might try to capture. The
important point is that once a firm has ownership of the asset, other firms
that are otherwise just as capable would be in a worse competitive position
than that firm.
Activity 6.3
The following are all examples of strategic assets. What would you do as a competitor if
one of your rivals had access to such an asset and you did not?
45
MN3119 Strategy
•• A patent for a painkilling substance. Assume you do not have a comparable substance
ready.
•• A taxi licence – all taxi licences are currently allocated (and not traded).
•• A prime location for a retail site for which your competitor has signed a long-term
lease.
•• Your rival has entered a market and is making profits of £100,000 at the moment. It
would cost you £70,000 to enter the market, and profits if both firms were active in
the market would be £60,000 each.
6.4 The dynamics of resource and capability building
We now have a look at the process of capability building. A firm’s
capabilities are very much like an asset – in balance sheets, the term
‘goodwill’ often represents the amount of brand equity or intangible assets
a firm possesses. In some important ways, however, a firm’s capabilities
differ from, say, a piece of production equipment. First, capability cannot
be built up immediately. Second, its returns are uncertain. Third, having
built up some capabilities, it may become easier to build up even more.3
We now describe the process of capability building (or asset stock
accumulation in Dierickx and Cool’s terms) in more detail.
An important distinction we have to make is between stocks and
flows. While the current stock of a firm’s capabilities will give firms
different strategic options, another important goal of strategy has to be
the building up of stocks by regulating the flow in and out of the resource
pool. Numerous established firms, for example, neglect to replenish their
capability pool, which eventually leads them to lose their competitive
advantage. On the other hand, young firms that are doing their utmost
to catch up with the industry leaders will find it hard to match their
capability level immediately. The important distinction to make is the
following:
Rule
While flows can be adjusted immediately, stocks cannot. Hence, the imitability of stocks
of resources depends on how resources are accumulated.
Bearing in mind that current stock levels affect a firm’s ability to
implement some strategies successfully, it is easy to see that a firm’s
capabilities may be a sustainable source of competitive advantage – after
all, it will take a challenger some time to build up the same capabilities.
Asset stock accumulation has a number of special features, which we will
discuss in turn.
In building up a firm’s capabilities, success often breeds success, which is
called asset mass efficiencies by Dierickx and Cool (1989). If you are
a successful firm, it is easier to attract good employees, which is likely to
make you even more successful, and so on. Similarly, if your technology
gathers a sufficiently large following, you will receive user feedback that
will enable you to improve your technology further (this is similar to
network effects – as you become more successful, you are likely to become
even more successful). There is, however, also a flipside to this – when
starting from scratch, it is very difficult to get a ‘foot in the door’ since the
first step is usually the most difficult and most expensive to make.
Asset attrition will work against the stability of an asset stock. As
mentioned before, if a firm fails to reinvest in replenishing its current asset
stock, it will deteriorate over time and its reputation wanes, employees
46
3
Of course, these
properties may still
exist for physical assets
to some extent, but
generally they are
regarded as being more
common for capabilities
or knowledge assets
than physical ones.
Chapter 6: The resource-based view of the firm
leave, the product goes out of date, and so on. Possessing a competitive
advantage cannot be taken for granted; as difficult as it may be for a
challenger, it is also crucial to keep up a firm’s own investment.
As we have seen before, stocks cannot be immediately adjusted – the
reason for this is time compression diseconomies whereby it is less
efficient to invest double the amount of resources in one period than it
is to spread investment over two periods. This means that a firm trying
to catch up with an incumbent will either have to spend considerably
more than a firm with an existing asset stock or accept that it is going to
take an long time of taking small steps to finally match the incumbent’s
asset stock. Anyone who has tried to study for an exam by revising all the
material two days before the exam will soon realise that it is much more
efficient to study a moderate amount every day or every week, since the
knowledge gained has more time to be absorbed. This will also be the case
for a firm’s ability to innovate – if all the possible innovation paths are
pursued at once, it is most likely to experience some shortages in terms of
researchers, coordination problems regarding research findings or perhaps
a lack of cross-fertilisation regarding unsuccessful research trajectories that
might be duplicated.
Also, asset stocks may be interconnected with each other. For
example, the ability to innovate will be inextricably linked to the ability
to design successful products for the market and the ability to roll out
such products to end consumers. If either of the three steps falls short,
the value of the others will decrease. For example, an engineering-focused
firm might be better off building up marketing capabilities instead of
investing yet more in the perfection of the production process for two
reasons: first, returns to investment in marketing might be much higher
than returns to further investment in the production process, and second,
a well-manufactured product is likely to benefit considerably from being
sold effectively – that is, the asset stocks are interconnected.
Finally, every resource accumulation process has a random element,
in particular for cases of capability building – for example, in firms
where employees are a large part of a firm’s capabilities, because it is not
always possible to predict when employees will leave the firm, this may
have a dramatic effect on the firm’s competitiveness. Although this is not
something a firm can actively influence, it is at least important for this
to be taken into account, for example, by scenario planning or making
allowances for periods of high or low asset stock growth.
6.5 Key concepts
•
Competitive advantage
•
Resources
•
Capabilities
•
Architecture
•
Reputation
•
Innovation
•
Strategic asset
•
Asset mass efficiency
•
Asset attrition
•
Time compression diseconomies.
47
MN3119 Strategy
6.6 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
link resources and distinctive capabilities
•
identify some sources of competitive advantage
•
describe the process of capability building.
6.7 Sample examination questions
1. ‘Reputation is hard to build up, but easy to lose.’ Does this mean that
reputation is an even stronger source of competitive advantage or a
weaker source than we discussed in this chapter?
2. Celebrity endorsements are often used to substitute for consumers’
experience with a particular product. What are the advantages and
risks of using celebrities as ‘reputation-builders’?
3. In light of the sources of competitive advantage outlined in this
chapter, explain the success of Japanese Keiretsu and Korean Chaebols.
Do you think that this source of competitive advantage will be stable
over time? Which factors could destabilise it?
4. Firms in network industries can often sustain a long-term advantage
over their competitors. Does this imply that network effects are
another source of competitive advantage that has been missed?
Guidance on answering these Sample examination questions can be found
in the VLE.
Extended activity: Dell Computer Corporation
Read the following and answer the questions at the end.
Dell Computer Corporation – an analysis of the PC industry.
The personal computer industry: History
Electronic computers emerged from military research undertaken during
the Second World War. In 1949, the magazine Popular Mechanics predicted
that ‘Computers in the future may perhaps only weigh 1.5 tons’. For the
following three decades, large mainframe and minicomputers, produced
by vertically integrated firms like IBM and Digital Equipment Corporations
(DEC) dominated the market. In 1977, Kenneth Olson, founder of mini
computer maker DEC, said, ‘There is no reason for any individual to have a
computer in their home’ (MacIntyre, 1977).
However, electronic hobbyists were already purchasing mail order and
retail kits, which allowed them to assemble simple computers at home.
Between 1975 and 1981, a group of firms began to offer increasingly
integrated, pre-assembled personal computers (PCs) (Das Narayandas
and Kasturi Rangan, 1996). Start-ups such as Apple Computer, MITS and
Commodore led the early market and were gaining popularity among
hobbyists and educational institutions with their easy-to-use machines.
Established firms including Texas Instruments, Hewlett-Packard, Zenith,
Nec, Xerox, IBM, Toshiba, Sanyo, Sony, Olivetti, Wang and DEC soon
joined the entrepreneurs and began to produce PCs.
IBM launched its first PC in 1981 and, two years later, held 42 per cent
of the market. On the mainframe market, IBM produced many of the
components and commanded a 61 per cent market share (Steffens,
48
Chapter 6: The resource-based view of the firm
1994a). In launching its PC, however, IBM purchased many components.
It commissioned a start-up software firm Microsoft, to write the operating
systems for its PCs and adopted a microprocessor architecture designed
by Intel. Most of the industry simply followed the IBM standards and, by
1983, the major alternative standard, a proprietary system by Apple, held
only 20 per cent of the market (Steffens, 1994b).
IBM used its huge sales force to sell PCs to large corporate accounts. To
serve small businesses and individuals IBM turned to retail stores such as
Sears and Computerland (Chposky and Leonsis, 1988). These resellers not
only sold PCs, they also guided the customers though the purchase of what
was still an unfamiliar product.
As the demand for IBM PCs grew rapidly, other firms began to offer ‘IBM
clones’. Compaq entered the market with a low-priced PC in 1982 and
booked a $100 million revenue during its first year. A large number of
start-ups followed the Compaq example, and entered the market with
IBM clones. Among these entrants was Dell Computer Corporation (Dell),
founded in 1984.
By 1986, IBM realised that it had set the standard, but that it had
outsourced the most valuable components of the PC, the microprocessor
and the operating system, to Intel and Microsoft. In 1986, IBM declined to
adopt Intel’s third-generation microprocessor, the 386 chip. By introducing
its PS/2 line of computers in 1987, IBM tried to make the PC more
proprietary. Compaq adopted the 386 chip and led a group of nine clone
makers in affirming the existing industry standards (Microsoft, Intel).
Though IBM eventually accepted the 386 microprocessor, its market
share had fallen from 37 per cent in 1985 to 17 per cent in 1989 (Das
Narayandas and V. Kasturi Rangan, 1996).
Microsoft released its new operating system Windows 3.0 in 1990, and
over the next four years, the user-friendly Windows became the PC
standard. ‘Wintel’, the combination of the Windows operating system and
Intel’s x86 microprocessor architecture dominated the market. By 1991,
between 85 and 90 per cent of computers sold conformed to Microsoft/
Intel standards, with the remainder using the proprietary Apple operating
system and a Motorola microprocessor.
The initial surge in sales of the PCs slowed down in 1990, just as a
recession hit the United States. The result was vigorous price competition.
For example, Dell ran advertisements in 1990 showing that its prices were
much lower than Compaq’s prices. Compaq usually discounted its PCs well
below the list price, but the advertising campaign was highly effective. In
response, Compaq slashed its prices by as much as 32 per cent, introduced
41 new products in 1992 and added a new distribution channel (Steffens,
1994). The inevitable price war followed.
The demand growth recovered in the mid-1990s, caused by strong
economic growth and the emergence of new, popular services involving
computer networks. The growing popularity of email and the world wide
web gave individual customers new reasons to buy a PC. However, PC
prices continued to decline. Compaq offered a powerful personal computer
for less than $1000 in 1997. By December 1998, the prices of the least
expensive models had dropped below $500.
Source
Kramer, N. under the supervision of Dr T. Kretschmer ‘Dell Computer
Corporation – an analysis of the PC industry’, LSE case study, July
2003.
49
MN3119 Strategy
Current situation and trends
Hardware components such as housings, keyboards, memory chips,
motherboards, disk drives, monitors, modems and connectors can be
purchased in highly competitive global markets, served by numerous
companies. In contrast, microprocessors and operating systems are
supplied by only a handful of companies. Standardisation permits
compatibility and interchangeability between different computer and
software systems. At the moment the standard is Wintel. Microsoft
Windows is installed on 97 per cent of all computer systems (due to
Network Effects), compared with just 1.43 per cent for Apple Macintosh
and 0.26 per cent for Linux (www.wininformant.com; September 2002).
Intel microprocessors are on 82.8 per cent of all PCs (http://news.com).
The hardware and software that makes up a PC are often sold as an
integrated bundle. PC makers such as IBM, Compaq, and Dell deliver
computers with a Microsoft operating system already installed and, in
turn, pay a fee to Microsoft.
Increasingly, PCs are being delivered with pieces of ‘application software’
already installed. A number of vendors offer application software such
as word processors, spreadsheets, database management systems, web
browsers and electronic messaging software. In this market Microsoft
holds a strong position as well, with a share of 80 per cent of the market
for the so-called ‘office productivity applications’ and approximately 10 per
cent of the overall market for application software (Microsoft Corporation
Annual Report, 2001).
As the processing costs decline, the lines between PCs and other devices
are becoming blurred. At the lower end of the processing and memory
spectrum handheld electronic organisers are beginning to compete with
PCs for applications such as email and portable computing.
Furthermore, an observer of the distribution channel reported that a number
of distributors, resellers and retailers are exploring ‘strategic alternatives’ as
the result of the bad availability of certain models of PCs (Anastasi, Johnson,
Drennan Lane and Spinner, 1998). MicroAge, one of the resellers that has set
up assembly line operations, has established the brand name ‘Pinacor’ for its
distribution operation (Lyons, 1998). The retailer CompUSA has started to
sell inexpensive computers under its own brand name (Lyons, 1998).
Customers
PC buyers are usually divided into four categories: large and mid-size
businesses and government institutions; small businesses and offices;
individual consumers and educational institutions.
1. Large and mid-size businesses and government institutions
Large and mid-size businesses and government institutions usually have
significant Management Information Systems (MIS) departments that
purchase maintain and support PCs in a centralised fashion. Staff members
are highly knowledgeable about PCs and are charged with providing a
reliable network of high-performance computers while also controlling
the information system costs. Once a PC has been purchased, the MIS staff
train the users and help the users when they encounter problems.
2. Small businesses and offices
Such organisations typically lack MIS staff. However, virtually all
businesses have extensive experience with PCs. Reliability, performance,
support, service, price, brand and channel recommendations play an
important role in the choice of a PC.
50
Chapter 6: The resource-based view of the firm
3. Individual consumers
Individual consumers purchase PCs for home or home-office use.
Individual buyers are very diverse, but tend to be more sensitive to price
and more interested in a computer’s brand name than business buyers
(Kay, 1999). Some customers also pay attention to the brand of the
microprocessor. This is why Intel has spent an estimated $4 billion on
brand advertising for its microprocessors since 1990 (Intel Corporation
Annual Reports 1990–2001). Among the individual consumers in the US,
30 per cent of the purchasers were first time buyers in 1998. By 2000 this
figure had dropped to 16 per cent (International Data Corporation, 2000).
4. Educational institutions
Apple was more successful in selling its PCs to educational institutions,
since it was the most user-friendly program in the 1990s. However,
more and more educational institutions are purchasing Wintel PCs at the
moment, since it is the industry standard.
Channels
PCs flow from manufacturers to customers via four channels: retail stores,
distributors (working with small resellers), integrated resellers and direct
distribution (Kasturi Rangan and Bell, 1998).
1. Retail stores
Retailers such as Circuit City and CompUSA in the United States and Time
Computers in Europe take delivery of PCs directly from manufacturers.
Machines then pass through distribution centres owned by the retailers on
their way to the stores. In these stores, retail displays and sales people play
an important role in helping customers select models and manufacturers.
According to a survey 93 per cent of the customers accepted the retailer’s
recommendations concerning computer purchases (Fattah, 1998). The
distributors and retailers have marked up the hardware by a total of
5–7 per cent in recent years (Fortuna and Pappachan, 2001). Moreover,
shelf space is limited and even large superstores carry only three to five
brands of PCs. These retailers operate on very thin margins. CompUSA
for example had a gross margin of roughly 7–9 per cent on computers
(CompUSA Annual Reports, 1998–2001).
2. Distributors (working with small resellers)
A handful of large distributors such as Ingram Micro and Tech Data supply
a full range of computer hardware and software to nearly 100,000 resellers
(Fattah, 1998). The resellers are typically small owner-managed firms.
3. Integrated resellers
Integrated resellers such as MicroAge and Vanstar operate distribution
centres, field extensive sales and service organisation, and in some cases,
manage the PC network of clients on an ongoing basis.
4. Direct distibution
A fourth and final channel leads directly from the PC manufacturer to the
customer. Only a few PC manufacturers take orders direct from the customers,
either over the telephone and internet or by an internal sales force. They then
deliver PCs via third party shippers like UPS or TNT Logistics.
Most PC manufacturers (HP, IBM, Fujitsu Siemens and NEC) use indirect
channels to deliver the PCs to the customers (first three combinations).
The general approach to selling PCs through indirect channels is shown in
Figure 6.1.
51
MN3119 Strategy
Component
Component
manufacturer
Order
Forecast
PC
manufacturer
Corporate
customer
Distributor
Product
Product
Figure 1: PC sales through indirect channels.
Manufacturers usually agree to buy back channel inventory that cannot
be sold. In addition, they provide price protection to resellers and
distributors, which means that if the price of the computer falls while it is
in the distribution channel, the manufacturer will reimburse the reseller
or distributor accordingly. By one estimate, inventory buy-backs and price
protection cost PC manufacturers 2.5 cents on every dollar of revenue
(Fortuna and Pappachan, 2001).
Dell Computer Corporation – history
While a first year undergraduate at the University of Texas at Austin,
18-year-old Michael Dell started a part-time business in his dorm room.
He formatted hard disks for PCs and added extra memory, disk drives and
modems to IBM clones, selling them for as much as 40 per cent less than
the comparable IBM machines. Reluctant to reveal his distraction from his
studies, Dell hid his PCs in his roommate’s bathtub when his parents came
to visit (www.dell.com).
When revenues reached $80,000 per month in 1984, Dell dropped out of
college and founded Dell Computer Corporation. Already companies such
as Exxon and Mobil were clamouring for 50 to 100 of Dell’s machines at a
time (Das Narayandas and V. Kasturi Rangan, 1996). In 1985, Dell shifted
from upgrading the machines of other manufacturers to assembling Dellbranded PCs. Revenue rose each subsequent year.
Between 1994 and 2003, the revenue of Dell rose from $3.5 billion to
$35.4 billion and profits increased from $149 million to $2.1 billion (as
can be seen from Exhibit 3, p.162). Between 1994 and 1998, Dell grew
twice as fast as its major rivals in the PC market and tripled its market
share (www.dell.com).
In the 2003 Forbes magazine list of the richest Americans, Michael Dell,
the 37-year-old founder of Dell, was ranked 11th with an estimated worth
of $11.2 billion (Forbes, 2003).
However, the basic elements of Dell’s Direct Model that came together
early in the company’s history still remained in place in 2003. The
company deals direct with end customers. It serves primarily corporate
customers and offers them high-performance PCs at relatively low prices.
PCs are customised to buyer specifications and assembly commences only
after Dell receives an order.
To quote Michael Dell: ‘It was too late to challenge the technical standard
and the dealer network had been done already. Compaq was very strong
in retail. A new marketing and distribution strategy was something new,
however’ (Michael Dell quoted in John R. Halbrooks, How to really deliver
superior customer service).
The direct business model
The core of Dell’s competitive advantage is the direct business model. The
underlying principles of the model will be analysed more extensively now.
52
Chapter 6: The resource-based view of the firm
Start here with the customer
Dell customers communicate and buy from Dell in three ways:
www.dell.com
Voice-to-voice
Face-to-face
Kitting
Based on the Traveler, all internal parts and components required to make the
system are picked and placed into a tote.
Just in-time inventory
Dell receives only the materials that are immediately ready to use and only those
specified as desirable to that particular customer.
Custom-designed computers
The Traveler is a sheet that contains all of the customer’s unique configuration
information. This document travels with the system throughout the assembly and
shipping.
Built-to-order
A team of workers uses the kit to assemble and initally test the
entire system
Testing and system integration
Systems are then extensively tested using Dell diagnostics. Standard or custom
hardware is factory installed and tested.
Boxing, shipping and delivery
The build-to-order cycle takes less than five hours from start to finish.
Figure 2: The direct sell and build to customer-model (Ng and Lovelock,
2000).
Dell pioneered a new business model that focused on speed of execution
and minimum inventory. To this end, the company bypassed the dealer
channel, selling products direct to customers over the phone or web
(compare Figure 3 with Figure 1).
Component
Order
Dell Computer
Corporation
Component
manufacturer
Distributor
Corporate
customer
Product
Figure 3: The Dell direct model.
53
MN3119 Strategy
The direct model has several key advantages. By eliminating the
intermediaries, Dell can dramatically reduce its channel costs. Exhibit 2,
p.170, shows that channel costs amount to 2 per cent under the direct
model compared with 13–15 per cent for indirect sales. Besides, Dell
relies on Information Technology (IT) to tightly control its value chain and
achieve a high degree of coordination.
Procurement
Dell works closely with suppliers to arrange just-in-time delivery of parts.
From 1992 to 1998, Dell had reduced the number of suppliers for its
Austin facility from 204 to 47 (Kasturi Rangan and Bell, 1998). With the
remaining suppliers, Dell maintains close electronic links, communicating
replenishment needs to some vendors on an hourly basis (Dell annual
report 2001). Most of the suppliers keep the components warehoused less
than 20 minutes away from Dell’s factories. Moreover, suppliers assign
engineers to Dell’s design teams, who are stationed in Dell plants during
new product introductions.
The electronic links allow Dell to direct some suppliers’ shipments straight
to the customers. Computer monitors supplied by Sony, for instance, never
pass through Dell’s facilities (www.dell.com). Dell trusts Sony to provide
high-quality monitors, so it is unnecessary to truck them to Round Rock
(Austin, Texas), test them and then ship them (with the rest of the PC) to
the customer. ‘That’s big waste of time and money, unless we get our jollies
from touching monitors, which we don’t’, says Dell (Dell, M., 1998).
Dell utilises the information to get its suppliers in line with its highvelocity model. This is achieved either directly, by improving the logistics,
or indirectly, by enhancing quality. For example, when a defective part
is identified, Dell lets the supplier know what was wrong with the part,
so it can quickly redesign the part and the poor-quality problem is fixed
immediately.
Operations
Dell operates facilities in Austin, Texas; Limerick, Ireland; Penang,
Malaysia; Alvorada, Brazil and Xiamen, China. Dell puts a strong emphasis
on its operations. Every PC manufactured by Dell is specifically configured
to a customer’s order. Once received, an order is electronically sent to
the appropriate manufacturing facility. In the facilities the Dell PCs are
manufactured to order. This means that there are virtually no finished
goods in inventory and little work in progress. Since materials move faster
in Dell factories, the latest technology is introduced faster than through
slow-moving indirect channels. As Dell remarks, ‘If I’ve got 11 days of
inventory and my competitor has 80, and Intel comes out with a new
microprocessor, that means I’m going to get to market 69 days sooner’
(Dell, M., 1998). The direct model reduces the high risk of obsolescence
associated with products in a rapidly changing technological market (Dell
annual report 2000). This is very important, since component prices
decline very rapidly. The prices of vital components dropped approximately
25 per cent per quarter over the last few years (Aizcorbe, 2002).
Components arrive from suppliers just in time for manufacturing through
the factory’s cargo doors. Manufacturing is synchronised to avoid storing
parts or finished systems. Teams build systems from start to finish.
Team members have profit sharing incentives, and hourly data on their
performance are posted in large monitors on the factory floor, so each team
knows if they are achieving their targets. Since there are no finished goods
in inventory a system is taken out of the factory as soon as it is complete.
54
Chapter 6: The resource-based view of the firm
Outbound logistics
PCs are shipped directly to the consumers. As mentioned above, items such
as monitors never pass through Dell facilities. Dell trusts Caliber and UPS
to manage the logistics so the PC and the monitor are properly matched
even though the former comes from Austin (Texas), and the latter from
Sony’s factory in Mexico.
Marketing and sales
Dell sells direct to the consumers. The direct model eliminates the need to
support an extensive network of wholesale and retail dealers. The dealer
mark-ups and the higher inventory cost associated with the wholesale and
retail channels are avoided. The model also avoids the competition for
shelf space. As mentioned before, retail shelf space is limited, and even
large superstores typically carry only three to five brands.
Segmentation
Figure 4 shows Dell’s customer segments. They can be classified as
‘transactional’ or ‘relationship’.
Small customers
(Business/Customer)
Large customer
Global
enterprises
Global
enterprises
Large
companies
Mid-size
companies
Mid-size
companies
Government
Federal
State/
Local
Education
Higher
education
K-12
Small
customers
Small
business
Consumer
Figure 4: Dell Customer Segments (2001).
Transactional customers are individuals or businesses who make
transaction-buying decisions. These customers focus on the economics
of the purchase, looking at factors such as performance, specifications,
features, reviews and awards. On the lookout for the best PC for a
particular application, they shop from a variety of vendors and use an
array of information sources, including reviews, advertising and word of
mouth when making their purchase decision. Dell does best with more
experienced computer buyers. First-time buyers tend to be in the market
for low-priced PCs, have more intensive service needs and feel more
comfortable with a salesperson. Consequently, Dell focuses on ‘computerliterate’ customers.
Relationship customers, in contrast, consider the cost over the life cycle
of the PCs. Price is only a secondary concern (several studies have shown
that the total cost of ownership of a PC over its life cycle is 5–10 times
its purchase cost). Most of them are business, government or education
accounts in medium-to-large scale organisations. These customers focus
on attributes such as service, reliability, vendor reputation and product
standardisation.
55
MN3119 Strategy
Forty per cent of customers, mainly large corporate accounts, are classified
as relationship customers. Thirty per cent consisting of mainly consumer
and small businesses are classified as transactional. The remaining 30
per cent is a mix of the two. This customer segmentation drives the
organisational structure of the company.
Demand management
Dell’s close relationship with its customers improves its demand
forecasts. Dell tries to understand the future computing needs of large
accounts by discussing their company’s needs and jointly planning the
company’s infrastructure. With transaction accounts, the telephone sales
representatives try to steer the customer towards product configurations
that are readily available. For example, if a customer requests a certain
configuration over the phone, the sales representative may suggest that
for a small incremental payment, the customer will get a component with
significantly higher quality (and faster), because the component is already
in stock. Sales representatives’ commissions are based on gross margins.
Dell designed its information systems so the gross margin is calculated for
each system and displayed on the sales representative’s computer screen.
Sales organisation
Dell’s sales organisation consists of teams of field account executives
and telephone sales representatives. Account executives are responsible
for selling products and services tailored to the needs of customers
in their geographical area. The telephone sales representatives are
responsible for processing orders and handling sales calls. Relationship
accounts are assigned to both account executives and telephone sales
representatives. Transactional customers are solely handled by telephone
sales representatives. The roles of the sales representative varies with
the customer segment served. For example, customers in the consumer
segment only rely on the sales representative for technical advice, so
they can impact the purchase decision. On the other hand, when dealing
with large firms, whose purchasing departments predetermine system
configurations, the sales representative’s role is reduced to order-entry.
www.dell.com
Michael Dell got interested in the internet in the early nineties. Listening
to industry ‘buzz’ and noting that he could order t-shirts online, it struck
him that a customer could order anything over the internet, including
a computer. Dell said: ‘We think about internet commerce as a logical
extension of our direct model… I’m only half joking when I say that the
only thing better than the internet would be telepathy. Because what
we’re all about is shrinking the time and the resources it takes to meet
customers’ needs. And we’re trying to do that in a world where those
needs are changing’ (Dell, 1998).
Dell had a number of natural advantages for selling on the internet.
Corporations and computer-literate customers that were willing to buy
over the phone were among the first to be ready to make purchases on
the internet. Further, unlike rivals Compaq, HP, and IBM, Dell did not have
an existing indirect channel that would feel threatened by direct internet
sales, and the discipline imposed by tracking campaigns and direct
telephone sales was applicable to internet as well.
Dell’s internet site: www.dell.com, was launched in July 1996. By
December 1996, ‘Dell’s direct sales reached $1 million a day. Dell’s webbased sales continued to grow, and by the end of 2002, the company’s
56
Chapter 6: The resource-based view of the firm
online sales generated $44 million in daily revenue, about half of the total
sales (Bob Kaufman, spokesperson Dell; www.dell.com).
The web pages on www.dell.com mirror Dell’s segmentation strategy,
with the business unit in charge of each customer segment having the
autonomy to deliver the content for its segment. The same applies to
international customers. More than three dozen regional Dell websites are
customised for particular countries or regions.
Dell’s website was initially geared to its transaction customers, who were
enthusiastic about making purchases by clicking a button for ordering
the configuration they had just put together. The internet proved to
be a natural and more efficient extension of Dell’s direct model. ‘I
couldn’t imagine a more powerful creation for extending our business’
(Dell, 1999). The internet enabled Dell to decrease the direct cost of
configuration, ordering, tracking and support for its transactional business
by approximately 15 per cent.
Part of these savings come from the increased efficiency of the sales
process. Under the traditional direct model, Dell would mail out
catalogues to prospective buyers. On average, about 10 per cent of the
catalogues mailed would lead to sales calls, with about 20 per cent of
calls resulting in actual sales. About 0.5 per cent of the site visits resulted
in sales through the internet channel, and about 5 per cent resulted in
telephone calls (these statistics are available to Dell, because the website
lists separate toll-free numbers, enabling tracking). Because customers
who called after browsing the website already have the necessary
information and are more inclined to buy, the time to close a sale is shorter
and, in addition, the probability of a sale is almost twice as high. As a
result, sales representatives working phone orders that originated from
web browsers are about 50 per cent more productive than Dell’s pure
telephone-based sales representatives.
The internet also increases Dell’s service efficiencies. For example,
customers who wanted to track their shipments call Dell about three times
on average to find order status. Dell’s web-based order tracking system
replaced more than two-thirds of these phone calls (www.dell.com).
At the beginning, convincing large corporate customers to buy through
the web was a challenge and some customers felt Dell was asking them
to radically change the way they purchased computers. To overcome this
resistance, Dell studied how large customers evaluated and acquired
systems, discovering a diversity of practices with no ‘one size fits all’
solution. In autumn 1997, Dell developed ‘Dell Premier Pages’ that could
scale by the thousands using software tools that allowed a sales team to
develop a new customised Premier site in less than a day. This enabled
Dell to extend the program to medium- and small-sized enterprises. These
Premier Pages allow the customer to carry out standard transactions
such as configuration, price quotes and purchasing, but also track order
and inventory in systematic detail and provide him with online asset
management support. By September 2000, Dell had more than 50,000
customised Premier sites.
After-sales service
Dell offers custom-tailored service and support systems (Dell annual report
2000). Dell offers a broad range of service and support programs through
its own technical personnel and its direct management of specialised
service suppliers. After a sale, Dell supports its products in several ways.
Dell uses www.dell.com as an electronic backbone of the after-sales service
57
MN3119 Strategy
and offers approximately 60,000 pages of customer support information.
A customer with a problem could also reach a technical support staff of
1,500 representatives via a hotline that is manned 24 hours a day. Using
the diagnostic software installed in the factory, the customer and the
specialist can resolve the problem over the phone in approximately 90 per
cent of cases.
Dell employs third-party maintenance providers like Unisys, Wang and
Decision One Consulting, who send out technicians to tackle problems
that require on-site support. These problems are usually solved within
24–48 hours. Even though field service is outsourced, Dell maintains
accountability for customer service. This way Dell can provide service
somewhat comparable to that of the resellers, but at lower cost.
Leaving the direct business model
In 1990 Dell departed from its direct business model and entered the
retail channel. The move, Michael Dell said, would ‘provide us with the
opportunity to generate significant new business and increase Dell’s market
penetration’, especially among ‘PC customers – particularly at the entry
level – who want to physically “feel and touch” a unit before they buy’ (Dell
Press Release, 1990). Accordingly, Dell produced two lines of standard PCs
and reached distribution agreements with computer superstores such as
CompUSA. Sales through the retail channel were considerable, but Dell
soon found out that it was losing money on it. Dell’s operation income was
–3 per cent of the revenues through the retail channel (versus a +5 per cent
operation income via direct sales (Das Narayandas and Kasturi Rangan,
1996). In 1994, Dell wisely withdrew from retail sales.
Surprisingly, last July, Dell started to experiment with kiosks in shopping
malls. Since launching the initiative, it has opened more than 60 kiosks in
nine states. Recently, Dell announced it is ramping up its kiosk presence
by placing them in Sears stores. These kiosks are mini-stores, about 10–12
feet wide, with basic inventory and Dell salespeople. A Dell Spokesman,
Bob Kaufman, explains that the kiosks enable mall shoppers to ‘go in
and touch and feel some of our product and then either order right there
(online of course) or go back home and order’ (Maquire, 2003).
Isn’t it ironic that a part of a giant’s expansion effort is geared towards the
traditional retail, again? Some call it an odd hybrid: a bricks-and-mortar
mini-store with an e-commerce option. Analysts of the sector note that one
of the advantages of Sears kiosk placement is that, in comparison to the
Dell site, the department store’s foot traffic contains a higher percentage of
women and shoppers who are over 55.
Competition
Exhibit 2, p.60, shows that Dell was the only top five computer
manufacturer to grow steadily, in both turnover and market share, in the
five years to 2002. Furthermore, Exhibit 3 to 9 pp.61–67, show that Dell
reported a higher net profit margin than its major competitors (Apple,
Fujitsu Siemens, Gateway, Hewlett-Packard and NEC) over 2001 and 2002.
Only IBM performed better than Dell. However, most profits and revenues
of IBM are generated in departments other then its Personal and Printing
Systems Department. The Personal and Printing Systems Department
represented 14 per cent of IBM’s revenues and reported a pre-tax income
of $57 million and a pre-tax profit margin of 0.5 per cent in 2002 (IBM
Annual Report). Despite the bad performance of the Personal and Printing
Systems Department in 2002, the corporate net profit margin was a
reasonable 6.6 per cent.
58
Chapter 6: The resource-based view of the firm
To summarise, it can be said that Dell outperforms its competitors in the
PC industry. Needless to say, Dell’s rapid growth and superior financial
results caused its rivals to take both notice and action.
Competitors realised that Dell’s sales increase was caused by the
explosive growth of the direct and online sales. In October 1998, HP
launched a modest effort to set up direct sales on the world wide web.
HP Shopping Village (www.hpshopping.com), a web service previously
providing refurbished HP computers to individuals, was expanded to allow
consumers to buy new PCs direct from HP. Business customers could use
a similar website to purchase HP PCs (also similar to IBM’s first website),
except HP required businesses to complete purchases though resellers.
Lew Platt, former CEO of HP, hinted that business customers might not
have to go through resellers in the future ‘You can’t ignore what Dell has
done…I could give you a list of names of really large customers who have
said to HP: “Either do business with us directly or you are not going to do
business with us”.’ (Exhibit 4, p.172, provides financial information for
HP).
Gateway 2000, founded in an Iowa farmhouse in 1985, is the world’s
second largest direct marketer, and like Dell, only sells direct to
customers. Gateway takes orders from the customers, produces PCs to
their specifications, loads software onto the PCs, and ships the machines
to customers. Its sales surged at an annual clip of 39 per cent from 1991
to 1996 and surpassed Dell’s US sales briefly in 1994. In 1997, however,
sales growth slowed to 25 per cent, and net income fell by half. Until the
year 2000, Gateway’s net revenue grew steadily, but since then sales have
dropped sharply in 2001 and 2002 (Exhibit 5, p.173, provides financial
information for Gateway).
Following the lead of its rivals, IBM introduced its first program to enable
businesses to buy a small set of products direct: by the end of 1998, the
Netfinity Direct program allowed large enterprises to purchase a particular
line of IBM servers without going through resellers. IBM began selling its
full line of PCs via the internet in May 1999. In December 1999, IBM’s
online sales totalled $38 million, but only 7 per cent of that came from
small businesses and consumers. Consequently, IBM planned a direct
sales assault in 2000. IBM built up its ‘build-to-order’ manufacturing
capabilities, which it considered a crucial element in selling PCs and
servers directly and started a huge marketing campaign to promote direct
selling.
Despite various attempts by competitors to lessen Dell’s position in the
direct channel, Dell remains the dominant player. In the US for example,
Dell shipped more than half of the PCs sold direct. A strong position
considering the fact that for the first time in history direct sales accounted
for more than half of the shipments in the US in the second quarter of
2002 (Gartner Dataquest, 2002).
Up until now Dell’s senior management appears unconcerned about the
efforts of rivals. Concerning competitors’ plans to sell direct, Michael Dell
once joked, ‘It is like we’re the best baseball player and Compaq is the best
basketball player. Now they want to play baseball’ (Serwer, 1997).
59
MN3119 Strategy
Exhibit 1. Channel economics: direct v indirect channels
Channel mark-up
Indirect
Direct
5.0–7.0%
0.0%
Co-op marketing
3.0%
0.0%
Financing
1.0%
1.0%
Price protection*
2.5%
0.0%
Obsolescence
1.5%
1.0%
Total
13.0–15.0%
2.0%
Source: Presentation of Mike Winkler, Senior VP, Compaq Computer, 2001
*
Price protection is a scheme whereby the PC manufacturer protects
resellers from losses on unsold computers when their price is reduced.
Exhibit 2. Wordwide PC market shares (units sold)
2002
Rank
Vendor
2002
1999
1998
1997
1
Dell
15.2% 12.9% 10.6% 9.8%
7.9%
5.5%
2
HP*
13.6% 6.9%
7.6%
6.4%
5.8%
5.3%
3
IBM
5.9%
6.2%
6.8%
7.9%
8.2%
8.6%
4
Fujitsu Siemens**
4.3%
4.5%
n/a
–
–
–
5
NEC***
3.3%
3.6%
4.6%
5.2%
4.3%
5.1%
Compaq*
–
11.1% 12.7% 13.2% 13.8% 13.1%
Others
57.7% 54.8% –
–
–
–
136
114
93
81
Total units (millions)
2001
134
2000
127
Source: International Data Corporation, January 2003
*
2002 Data for Hewlett-Packard includes shipments for the new
Hewlett-Packard (including Compaq merger) starting in Q2 2002,
and only Hewlett-Packard data for Q1 2002. 2001 data includes only
Hewlett-Packard data (excluding Compaq). This reflects the legal
status of the companies, which merged in Q2 2002
** The 50–50 joint venture was formed in 1999 through the merger
of the European computer operations of global giants Fujitsu and
Siemens
*** Data for DEC includes Packard Bell
Shipments are vendor-branded shipments into distribution channels or
direct to end users
Data for all vendors are reported for calendar periods
60
Chapter 6: The resource-based view of the firm
Exhibit 3. Dell Computer Corp. financial performance
(in $; B=Billion, M=Million, K= Thousand)
Income statement
2002
2001
2000
1999
1998
1997
Net revenue
35.4B
31.2B
31.9B
25.3B
18.2B
12.3B
Expenses
32.6B
29.4B
29.2B
23.0B
16.2B
11.0B
Income before taxes
3.0B
1.7B
3.2B
2.5B
2.1B
1.4B
Income taxes
905M
485M
958M
785M
624M
424M
Income after taxes
2.1B
1.2B
2.2B
1.7B
1.5B
944M
EPS diluted, incl.
0.8
0.46
0.79
0.61
0.53
0.32
8.9B
7.9B
9.7B
7.7B
5.8B
3.9B
Balance sheet
Current assets
Long-term assets
15.5B
13.5B
13.7B
11.5B
6.9B
4.3B
Current liabilities
8.9B
7.5B
6.8B
5.2B
3.7B
2.7B
Long-term debt
506M
520M
509M
508M
512M
17M
Total liabilities
10.6B
8.8B
8.0B
6.2B
4.6B
3.0B
Avg. shares outstanding
2.6B
2.6B
2.6B
2.5B
2.5B
2.6B
Shares outstanding
2.6B
2.6B
2.6B
2.6B
2.5B
2.6B
Total equity
4.9B
4.7B
5.6B
5.3B
2.3B
1.3B
Cash from operations
n/a
3.8B
4.2B
3.9B
2.4B
1.6B
Cash from investing
n/a
–2.3B
–757M –1.2B
Cash from financing
n/a
–2.7B
–2.3B
–695M –812M –898M
Profit margin (%)
5.99
4.00
7.01
6.59
8.00
7.66
Operating margin (%)
8.55
5.55
10.02
9.70
11.42
11.10
Return on equity (%)
43.55
26.54
39.77
31.39
62.90
73.01
Return on assets (%)
13.72
9.21
16.36
14.52
21.23
22.12
Debt/equity
0.10
0.11
0.09
0.10
0.22
0.01
P/E
34.24
58.68
32.46
42.79
48.72
78.30
Current ratio
1.00
1.05
1.43
1.48
1.57
1.45
–930M –657M
Ratios
Source: Dell Annual Reports
Note: Fiscal year-end January
61
MN3119 Strategy
Exhibit 4. Hewlett-Packard Company financial performance
(in $; B=Billion, M=Million, K= Thousand)
Income statement
2002
2001
2000
1999
1998
1997
Net revenue
56.6B
45.2B
48.9B
42.4B
39.4B
35.5B
Expenses
57.6B
43.8B
44.8B
38.6B
36.0B
32.1B
Income before taxes
–1.1B
702M
4.6B
4.2B
3.7B
3.6B
Income taxes
–129M 78M
1.1B
1.1B
1.0B
1.1B
Income after taxes
–923M 624M
3.6B
3.1B
2.7B
2.5B
EPS diluted, incl.
0.32
0.32
0.32
0.32
0.3
0.26
36.1B
21.3B
23.2B
21.6B
18.5B
20.9B
Balance sheet
Current assets
Long-term assets
70.7B
32.6B
34.0B
35.3B
31.7B
31.7B
Current liabilities
24.3B
14.0B
15.2B
14.3B
11.9B
11.2B
Long-term debt
6.0B
3.7B
3.4B
1.8B
2.1B
3.2B
Total liabilities
34.4B
18.6B
19.8B
17.0B
14.8B
15.6B
Avg. shares outstanding
2.5B
1.9B
2.0B
2.0B
2.1B
2.1B
Shares outstanding
3.0B
1.9B
1.9B
2.0B
2.0B
2.1B
Total equity
36.3B
14.0B
14.2B
18.3B
16.9B
16.2B
Cash from operations
5.4B
2.6B
4.7B
3.0B
5.2B
3.9B
Cash from investing
3.1B
–561M –1.4B
–628M –608M –2.6B
Cash from financing
–1.6B
–1.2B
–5.3B
–1.0B
–3.7B
–1.1B
Profit margin (%)
–1.63
1.38
7.29
7.33
6.79
7.09
Operating margin (%)
–1.86
1.55
9.46
9.90
9.37
10.06
Return on equity (%)
–2.55
4.47
25.06
16.97
15.83
15.57
Return on assets (%)
–1.31
1.92
10.47
8.79
8.45
7.92
Debt/equity
0.17
0.27
0.24
0.10
0.12
0.20
P/E
NA
49.66
8.89
10.40
12.35
13.04
Current ratio
1.48
1.53
1.53
1.51
1.56
1.87
Ratios
Source: Hewlett-Packard Annual Reports
Note: Fiscal year-end October
62
Chapter 6: The resource-based view of the firm
Exhibit 5. Gateway Inc. financial performance
(in $; B=Billion, M=Million, K= Thousand)
Income statement
2002
2001
2000
1999
1998
1997
Net revenue
4.2B
6.1B
9.6B
9.0B
7.7B
6.3B
Expenses
4.7B
7.3B
9.1B
8.4B
7.2B
6.1B
Income before taxes
–475M –1.3B
408M
663M
541M
203.M
Income taxes
–178M –275M 155M
235M
194M
93M
Income after taxes
–297M –1.0B
253M
427M
346M
109M
EPS diluted, incl.
–0.95
–3.2
0.73
1.32
1.09
0.35
2.0B
2.1B
2.3B
2.7B
2.2B
1.5B
Balance sheet
Current assets
Long-term assets
2.5B
3.0B
4.2B
4.0B
2.9B
2.0B
Current liabilities
940M
1.1B
1.7B
1.8B
1.4B
1.0B
Long-term debt
0
0
0
3M
3M
7M
Total liabilities
1.3B
1.4B
1.8B
1.9B
1.5B
1.1B
Avg. shares outstanding
324M
323M
321M
314M
311M
307M
Shares outstanding
324M
324M
323M
319M
313M
308M
Total equity
1.2B
1.6B
2.4B
2.0B
1.3B
930M
Cash from operations
NA
–270M 288M
731M
907M
442M
Cash from investing
NA
109M
–953M –831M –356M –360M
Cash from financing
NA
405M
20M
54M
23M
153K
Profit margin (%)
–7.14
–16.69
2.64
4.77
4.50
1.74
Operating margin (%)
–11.41
–21.22
4.26
7.40
7.03
3.24
Return on equity (%)
–23.88
–64.81
10.64
21.22
25.77
11.81
Return on assets (%)
–11.86
–33.96
6.06
10.82
11.98
5.38
Debt/equity
0.00
0.00
0.00
0.00
0.00
0.01
P/E
NA
NA
2.91
1.68
2.06
6.41
Current ratio
2.08
1.85
1.37
1.49
1.56
1.54
Ratios
Source: Gateway Annual Reports
Note: Fiscal year-end December
63
MN3119 Strategy
Exhibit 6. International Business Machines Corp. financial
performance
(in $; B=Billion, M=Million, K= Thousand)
Income statement
2002
2001
2000
1999
1998
1997
Net revenue
81.2B
83.1B
85.1B
87.5B
81.7B
78.5B
Expenses
74.4B
73.2B
76.0B
77.8B
72.5B
69.4B
Income before taxes
7.5B
11.4B
11.4B
11.8B
9.0B
9.0B
Income taxes
2.2B
3.3B
3.5B
4B
2.7B
2.9B
Income after taxes
5.3B
8.1B
7.9B
7.7B
6.3B
6.1B
EPS diluted, incl.
0.59
0.55
0.51
0.47
0.43
0.39
Current assets
41.7B
42.5B
43.9B
43.2B
42.4B
40.4B
Long-term assets
96.5B
90.3B
88.3B
87.5B
86.1B
81.5B
Current liabilities
34.5B
35.1B
36.4B
39.6B
36.8B
33.5B
Balance sheet
Long-term debt
20.0B
16.0B
18.4B
14.1B
15.5B
13.7B
Total liabilities
73.7B
66.9B
67.7B
67B
66.7B
61.7B
Avg. shares outstanding
1.7B
1.7B
1.8B
1.8B
1.9B
2.0B
Shares outstanding
1.7B
1.7B
1.7B
1.8B
1.9B
1.9B
Total equity
22.8B
23.4B
20.6B
20.5B
19.4B
19.8B
Cash from operations
13.1B
14.0B
9.0B
10.1B
9.3B
8.9B
Cash from investing
–6.9B
–5.9B
–4.0B
–1.7B
–6.1B
–6.2B
Cash from financing
–7.3B
–5.3B
–6.4B
–8.6B
–5.0B
–3.1B
Profit margin (%)
6.57
9.81
9.25
8.81
7.75
7.76
Operating margin (%)
9.27
13.78
13.41
13.43
11.07
11.50
Return on equity (%)
23.41
34.74
38.18
37.60
32.56
30.75
Return on assets (%)
5.53
9.02
8.91
8.81
7.35
7.48
Debt/equity
0.88
0.68
0.89
0.69
0.80
0.69
P/E
26.01
17.35
18.29
19.15
24.14
26.38
Current ratio
1.21
1.21
1.21
1.09
1.15
1.21
Ratios
Source: IBM Annual Reports
Note: Fiscal year-end December
64
Chapter 6: The resource-based view of the firm
Exhibit 7. Apple Computer Inc. financial performance
(in $; B=Billion, M=Million, K= Thousand)
Income statement
2002
2001
2000
1999
1998
1997
Net revenue
5.7B
5.4B
8.0B
6.1B
5.9B
7.1B
Expenses
5.7B
5.7B
7.5B
5.8B
5.7B
8.2B
Income before taxes
87M
–52M
1.1B
676M
329M
–1.0B
Income taxes
22M
–15M
306M
75M
20M
0
Income after taxes
65M
–37M
786M
601M
309M
–1.0B
EPS diluted, incl.
0.18
–0.07
2.18
1.81
1.05
–4.14
5.4B
5.1B
5.4B
4.3B
3.7B
3.4B
Balance sheet
Current assets
Long-term assets
6.3B
6.0B
6.8B
5.2B
4.3B
4.2B
Current liabilities
1.7B
1.5B
1.9B
1.5B
1.5B
1.8B
Long-term debt
316M
317M
300M
300M
954M
951M
Total liabilities
2.2B
2.1B
2.7B
2.1B
2.6B
3.0B
Avg. shares outstanding 355M
345M
324M
286M
263M
252M
Shares outstanding
359M
350M
335M
321M
270M
255M
Total equity
4.1B
3.9B
4.1B
3.1B
1.6B
1.2B
Cash from operations
89M
185M
868M
822M
775M
154M
Cash from investing
–252.0M
892M
–972M –988M –543M –499M
Cash from financing
105M
42M
–31M
11M
19M
23M
Profit margin (%)
1.13
–0.69
9.85
9.80
5.20
–14.76
Operating margin (%)
1.52
–0.97
13.68
11.02
5.54
–14.76
Return on equity (%)
1.59
–0.94
19.14
19.36
18.82
–87.08
Return on assets (%)
1.03
–0.61
11.55
11.65
7.20
–24.69
Debt/equity
0.08
0.08
0.07
0.10
0.58
0.79
P/E
79.78
NA
6.03
6.96
12.47
NA
Current ratio
3.25
3.39
2.81
2.77
2.43
1.88
Ratios
Source: Apple Computer Inc. Annual Reports
Note: Fiscal year-end September
65
MN3119 Strategy
Exhibit 8. NEC Corp. financial performance
(in $ Millions)
Income statement
2002
2001
2000
Net revenue
38,446
43,099
48,461
Cost of goods sold
27,833
29,725
33,044
Gross profit
10,613
13,375
15,416
Gross profit margin
27.6%
31.0%
31.8%
SG&A expense
9,325
9,907
11,811
Depreciation & amortisation
1,769
1,993
2,533
Operating income
–481
1,475
1,072
Operating margin
–1.25%
3.40%
2.20%
Total net income
–2,352
451
101
Net profit margin
–6.12%
1.5 %
0.21%
Cash
2,847.3
3,081.9
3,630.7
Net receivables
7,064.0
8,607.0
8,772.6
Inventories
4,899.4
6,597.3
7,258.3
Total current assets
18,126.8
19,644.8
20,908.1
Total assets
37,767.0
38,429.8
44,747.2
Balance sheet
Short-term debt
6,199.3
4,356.8
6,126.7
Total current liabilities
16,659.3
17,508.8
18,991.8
Long-term debt
11,297.0
9,730.1
12,875.1
Total liabilities
3,509.2
31,139.6
35,263.1
Total Equity
4,257.8
7,290.1
9,484.0
Source: NEC Annual Reports
Note: Fiscal year-end March
66
Chapter 6: The resource-based view of the firm
Exhibit 9. Fujitsu Siemens financial performance
(in $ Millions)
Income statement
2002
2001
Net revenue
4,728
5,174
Cost of goods sold
3,959
4,441
Gross profit
769
732
Gross profit margin
16.3%
14.2%
SG&A expense
734
750
Operating income
35
–18
Operating margin
0.74%
–0.35%
Total net income
-6
–57
Net profit margin
–0.13%
–1.10%
Cash
82.4
134.7
Net receivables
806.1
1,122.1
Inventories
337.7
385.9
Total current assets
1,529.7
1,731.1
Total assets
1,697.5
1,873.0
Short-term debt
133.8
—
Total current liabilities
1,044.4
1,212.6
Long-term debt
154.6
158.3
Total liabilities
1,571.1
1,736.3
Total equity
126.3
136.7
Balance sheet
Source: Fujitsu Annual Reports
Note: Fiscal year-end March
Note: The 50–50 joint venture was formed in 1999 through the merger of the European computer
operations of global giants Fujitsu and Siemens.
References
Anastasi, R., N. Johnson, J. Drennan Lane and S. Spinner ‘The computer sales
channel: third quarter demand survey’, Robinson-Humphrey 23 September
1998.
Aizcorbe, A. ‘Why are semiconductor prices falling so fast, industry estimates and
implications for productivity measurement’, Federal Reserve Board March 2002.
Chposky, J. and T. Leonsis Blue magic. (New York: Facts of File Books, 1988), pp.
56–66.
CompUSA Annual Reports, 1998–2001.
Dell, M. ‘An interview with Michael Dell’, Harvard Business Review March/April
1998, pp.72–82.
Dell, M. Direct from Dell: strategies that revolutionized an industry. HarperBusiness,
1999.
Dell Computer Corporation Annual Report 2000.
Dell Computer Corporation Annual Report 2001.
Dell Computer Corporation Press Release ‘Dell PCs to be offered through soft
warehouse superstores – expands reach to individuals and small businesses’, 10
September 1990.
Fattah, H. ‘1998 Forecasts: the channel fights back’, MC Technology Marketing
Intelligence January 1998.
‘The Forbes 400: the richest people in America’, Forbes website, 2003.
67
MN3119 Strategy
Fortuna, S. and P. Pappachan Compaq Reengineers the Channel; Will It Be Enough to
Slow Dell’s Momentum?, Deutsche Morgan Grenfell Technogy Group, 2001.
Gartner Dataquest 2002, Most US PC Buyers Purchase Directly from Manufacturers,
internal report by Kitagawa and Smulders, 13 August 2002.
www.wininformant.com/Articles/Index.cfm?ArticleID=26567 (September 2002).
http://news.com.com/2100-1001-947317.html (June 2002).
www.dell.com/
Intel Corporation Annual Reports 1990–2001.
International Data Corporation, 2000–03.
Kasturi Rangan and Bell, M. ‘Dell Online’, 1998, HBS Case 598-116.
Kay, R., analyst at International Data Corporation, 22 April 1999.
Lyons, D. ‘Games dealers play’, Forbes 19 October 1998, pp.132–34.
MacIntyre, J. ‘Fact of life: computers’, Spirit 1997, p.170.
Maguire, J. Case Study: Dell.com, 3 March 2003 (http://ecommerce.Internet.com/
news/insights/trends/article/0,,10417_2013731,00.html).
Michael Dell quoted in John R. Halbrooks, How to really deliver superior customer
service. (Inc Publishing, 1996) [ISBN 1880394286].
Microsoft Corporation Annual Report 2001.
Narayandas, D. and Kasturi Rangan ‘Dell Computer Corporation’, 1996, HBS Case
596-058.
Ng, P., P. Lovelock and A. Farhoonand ‘Dell selling directly, globally’, 2000,
University of Hong Kong Case Study.
Serwer, A. ‘Michael Dell turns the PC world inside out’, Fortune 8 September 1997,
pp.76–86.
Steffens, J. Newgames: strategic competition in PC revolution. (Oxford: Pergamon
Press, 1994a), p.181.
Steffens, J. Newgames: strategic competition in PC revolution. (Oxford: Pergamon
Press, 1994b), p.253.
Yoffie, D. ‘Apple Computer 1992’, Harvard Business School Case 792-081, p.4.
Questions
1. What are Dell’s distinctive capabilities?
2. Dell has maintained their position of superior profitability for quite
some time now. Combining information from the case and the chapter,
explain why other firms did not imitate Dell’s business model?
3. List some other industries in which Dell’s business model might also
work. Would it make sense for Dell to diversify in other businesses?
68
Chapter 7: Strategic asymmetries – persistent dominance over time
Chapter 7: Strategic asymmetries –
persistent dominance over time
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
explain sources of firm asymmetries
•
describe available strategies firms can employ to increase their
industry dominance.
Essential reading
Besanko, D., D. Dranove, M. Shanley and S. Shaefer Economics of strategy.
(New Jersey: Wiley, 2009) Chapter 2.
Further reading
Benkard, L. ‘Learning and forgetting: the dynamics of aircraft production’,
American Economic Review 90(4) 2000, pp.1034–054.
Cabral, L. and M. Riordan ‘The learning curve, predation, antitrust, and
welfare’, Journal of Industrial Economics 45(2) 1997, pp.155–69.
Klette, T. ‘R&D, scope economics, and plant performance’, Rand Journal of
Economics 27(3) 1996, pp.502–22.
Lieberman, M. ‘Market growth, economies of scale, and plant size in the
chemical processing industries’, Journal of Industrial Economics 36(2) 1987,
pp.175–91.
Spence, A. ‘The learning curve and competition’, The Bell Journal of Economics
12(1) 1981, pp.49–70.
7.1 Introduction
Some firms are simply better than others at what they do. There are also
firms that play their game better than others or that have been able to
secure an advantage over their rivals. This chapter looks at persistent
asymmetries among firms and assesses features of a market that might
lead to a continuing asymmetric outcome.
7.2 Firm asymmetries – long- or short-term?
We have seen in the previous chapter that some firms may possess
distinctive capabilities – characteristics that may enable them to
outperform their competitors over an extended period of time. There are,
however, reasons other than core competencies as to why a firm might be
doing better than its rivals at any point in time.
The regulatory setting in an industry might allow some firms to be
more successful than their rivals. For example, state-owned enterprises
might be protected by favourable contracts, licences might protect existing
firms from entry and so on. This is similar to the ownership of strategic
assets, which might not necessarily be linked to a firm doing anything
particularly well, but rather with the firm having been allocated (or having
acquired) a particularly scarce asset that gives it market power.
69
MN3119 Strategy
Second, firms may also be successful due to historical factors. Having
been present at the outset of an industry, for example, will typically give
a firm an advantage over latecomers – these are the elusive ‘first-mover
advantages’ that might have their roots in consumer habits over branding,
technology employed, learning, network effects and lots of other factors.
The main point is that in the past, some factors have contributed to a firm
maintaining a dominant market position even though the factor as such is
no longer a key determinant of success.
Third, an advantage may also be down to luck. For example, a firm
may get lucky with a new product design – for example, Bart Simpson
phones (in the times before there were mobile phones) sold significantly
better than other ‘gimmick’ phones in the shape of other popular cartoon
characters (see also the example below).
Example: Courvoisier and rappers
‘We weren’t expecting cognac to be associated with those type of people,’ said JeanMarie Macoin, a 55-year-old cognac producer after viewing Busta Rhymes’ ‘Pass the
Courvoisier’ music video. To the horror of many in the close-knit cognac industry, Busta
Rhymes, Jay-Z and the hip-hop community have helped to resurrect this perishing industry
with an ongoing obsession for lyrics praising Courvoisier and Hennessy.
To most, cognac symbolises getting drunk – a far reach from the ideals of the small
French community of approximately 20,000 where hip-hop’s favourite alcoholic drink
originates. An economic crisis in 1998 caused cognac’s number one market, Asia, to
dramatically reduce demand for the drink – almost wiping out the cognac industry. But it
was rappers who eventually brought the bottle back to life.
MCs as diverse as Common and Snoop Dogg have long hailed cognac as a drink of
choice, with more recent references in last year’s hit single ‘Pass the Courvoisier’ which
featured a video showcasing Busta, Pharrell and P. Diddy partying it up in a bar filled with
beautiful women and hundreds of gleaming bottles of Courvoisier. Songs like ‘Courvoisier’
have produced a following among urban youths, who have been known to mix the drink
for new concoctions such as ‘Thug Passion’ and ‘French Connection’. Jay’s newly opened
40/40 Club even includes a ‘Remy Room’ in homage to Remy Martin, his favorite cognac
brand. ‘Cognac is a classy, sophisticated and really smooth thing to drink,’ Jay-Z told The
Wall Street Journal.
The cognac industry is now prospering thanks to exports to the US which have almost
tripled in the last 10 years. Furthermore, Americans spent approximately $1 billion on the
French drink last year. Interestingly, cognac’s surging US sales are parallel to hip-hop’s
mainstream growth. Americans imported 3.7 million cases of cognac last year, 36 per cent
of the worldwide market compared with 1.3 million in 1993. According to The Wall Street
Journal, Hennessy, America’s biggest cognac brand with 53 per cent of the market, claims
that young black people represent 85 per cent of US sales.
Though America’s infatuation with cognac has salvaged this industry, the flip side is that
French grape growers responsible for cognac are dumbfounded by rap’s use of the brand.
‘It’s not quite the same world,’ Anne-Sophie Louvet, a 44-year-old woman who cultivates
her great-grandfather’s 113-year-old vineyards told The Wall Street Journal. ‘In this region,
you don’t show your wealth if you have some, and you don’t talk about money,’ she
added.
Last April, Courvoisier took steps to educate 900 farmers about the US market, and part
of the lesson featured a presentation of the aforementioned ‘Pass the Courvoisier’ video.
‘They didn’t know what to make of it’, a Courvoisier spokewoman told The Wall Street
Journal. To further familiarise themselves with their biggest consumers, Hennessy intends
to fly half a dozen grape growers to tour New York City night clubs this autumn. where
they will get a chance to get lifted hip-hop-style.1
70
1
Source: www.sohh.com/
Chapter 7: Strategic asymmetries – persistent dominance over time
The sources of firm asymmetries – core competencies, regulation, luck and
history – all help to explain the differences between firms at any point in
time. More importantly, however, is to assess which asymmetries can be
sustained over a longer period of time. For example, we would expect the
current popularity of cognac to be short-lived – after all, the next big drink
might just be around the corner.
Figure 7.1 shows the evolution of the wide-body aircraft market. In the early
half of the 1970s, there was vigorous competition between Boeing (B747),
McDonnell Douglas (DC10) and Lockheed (L1011), but after 1975 Boeing
surged ahead and strengthened its dominance (until Airbus challenged its
position), meaning the market has now become a very competitive duopoly.
100
B747
DC10
L1011
Market share (%)
80
60
40
20
0
65
70
75
Year
80
85
90
Figure 7.1: Market evolution of the wide-body aircraft industry.
Source: ‘Competing down the learning curve’. Case Study, Luis Cabral and Tobias
Kretschmer.
This example is just one of many where firms start out with a relatively
modest advantage over their rivals, but then manage to extend their lead
over time and end up with vastly superior market shares and profitability.
For further details, please see the extended activity for this chapter.
7.3 Traditional sources of persistent dominance
We will discuss three of the most important sources of increasing
dominance – economies of scale, economies of scope and economies of
learning.
7.3.1. Economies of scale
The formal definition of economies of scale is as follows.
Economies of scale – definition
ACQ > ACQ ' , Q < Q ' , AC =
FC + VC
Q
This implies that as more output (Q) is produced, the average cost per unit
goes down. As we can see from the definition of average cost (AC), this
effect can originate either from FC – fixed cost – or VC – variable cost.
Fixed cost degression is the simplest and most intuitive form of economies
of scale. If there is some fixed cost that has to be paid as soon as
production starts – say, the cost of equipment, a building, some central
administration – the more output is produced, the more the cost can
be spread over a larger output. Fixed-cost degression is likely to exist
71
MN3119 Strategy
in practically any firm as some upfront investments simply have to be
made in order to start producing, but the effect of more units and the
contribution of fixed cost to overall unit cost will decrease as output
increases. The effect is therefore most pronounced in industries where the
ratio of fixed cost to variable cost is very high.
Activity 7.1
In which industries would you expect fixed cost degression to be especially relevant?
Rank them and give an estimated ratio of fixed to variable cost in an average firm in this
industry.
•• PC software
•• Hairdressing
•• Handmade candles
•• Car production.
Ever since Adam Smith’s visit to the pin factory the specialisation of
equipment or activities has been regarded as another source of scale
economies. Larger quantities will allow workers to split up tasks and
concentrate on a single one or will justify the purchase of a specialised
piece of production equipment to automate production. (There is a nice
link to fixed cost degression here – for smaller production quantities
purchasing such a machine would imply a high fixed cost per unit, while
large quantities will render such a machine economical.)
Some products possess production or inventory economies of scale. For
example, to build a pipe with double the capacity, less than twice the
material is needed, creating production or technological economies of
scale for liquid-based products. Similarly, queuing theory tells us (and
intuition hopefully confirms) that to serve twice the number of customers
arriving at random times, less than twice the inventory is required. The
following example illustrates this: a chip shop will still experience some
peaks in demand regardless of how many customers turn up on average.
These demand peaks are not twice the size if the average number of
customers is twice as high.
Advertising can also demonstrate economies of scale – larger firms
with more outlets will have two advantages over their smaller
counterparts. First, their brand recognition is likely to be higher, so that
an advertisement reaching a consumer is more likely to be recognised if
it comes from a large firm. Second, even a consumer who recognises the
advertisement of the smaller counterpart and would like to purchase the
product will have more difficulties finding the product since it will have
fewer outlets, less shelf space and so on. Both of these then will translate
the same advertising dollar into more sales pounds for a larger firm –
advertising economies of scale.2
Purchasing power also increases with the size of the buying firm. That is,
if the input purchased by one buyer makes up a significant portion of sales
by the supplier, bargaining power is high for the buyer and prices paid
are likely to be low. Realising this, buyers have increasingly teamed up to
form purchasing consortia – Letsbuyit.com (now defunct) is an example
of individual consumers purchasing jointly to achieve better prices, and
one of the main benefits of airline alliances is the ability to purchase
everything jointly, in particular aircraft.
There are other important sources of economies of scale, such as
economies of R&D, but most of these will fall under one of the headings
72
2
We should note here
that advertising is often
regarded as a fixed cost
and as such an economy
of scale is created by
distributing advertising
cost over a larger
output.
Chapter 7: Strategic asymmetries – persistent dominance over time
above. In fact, some of the economies of scale will have their roots in
several of the factors mentioned above.
Activity 7.2
Explain possible causes for economies of scale in the following industries or functions:
a. Research and development
b. Desktop publishing
c. Banking
d. Teaching.
7.3.2 Economies of scope
In many ways, economies of scope are similar to economies of scale and
the distinction is sometimes rather blurry. However, for the sake of clarity
we will treat the two concepts separately, but keep in mind that many
people will use these two almost interchangeably.
The formal definition of economies of scope is as follows.
Economies of scope – definition
TC(q1,q2) < TC(q10), + TC(0,q2)
It is cheaper to produce positive quantities of two different goods (q1 and
q2) in the same firm than it is to produce them separately. Again (and this
may be where the confusion with economies of scale comes from), this
could originate from fixed cost or from variable cost (TC = FC + VC). We
will discuss the most important causes in turn.
An economy of scope would result from the waste of one production
process being the input for another product, or input/output relations.
For example, chemical processes often generate waste that can be used
for other, unrelated products, thus creating a motive for merging or at
least collaborating across different product classes. Similarly, as we will
see later, vertical relations between firms are often subject to considerable
(direct and indirect) costs, which could be avoided if the supplier and
manufacturer were in the same firm. Be aware, however, that these
cost savings are not always considered economies of scope as they refer
to contractual incompleteness and market inefficiencies rather than
technological or logistical economies; the definition of economies of scope
still allows for such input/output relations to fall under it.
Complementary assets are the key source of economies of scope.
Many mergers of two firms from unrelated industries will be justified on
the grounds that assets can be used productively in both industries. For
example, banking and insurance divisions in the same firm could have
access to each other’s client lists and might conceivably make the most
of cross-selling opportunities – that is, selling banking clients insurance
and insurance clients banking services. Complementary assets can also
be physical assets, even buildings. For example, even though academic
departments may not necessarily interact much, they would still be using
the same lecture theatres and teach the same students. So while not
every management student will take an active interest in anthropology
lectures, there may be some each year that would like to take a course in
anthropology, thus enabling a lecturer to teach a course that might not
have been taught with anthropology students alone.
73
MN3119 Strategy
Umbrella branding, or reputation stretching in general, is the
process of using the reputation built up in one line of business to launch
another one. This new line of service may or may not have anything to do
with the first. What is important here is that it is offered under the same
brand name as the first one, and as such benefits to some extent from
the established reputation of the first product, as the following example
illustrates.
Example: Virgin – king of brands
Virgin, a UK-based company, is involved in planes, trains, finance, soft drinks, music,
mobile phones, holidays, cars, wines, publishing, bridal wear. What ties all these
businesses together are the values of our brand…Virgin stands for value for money,
quality, innovation, fun and a sense of competitive challenge.3
A rather specific source of economies of scope is the running of hub-andspoke-networks by airlines. By gathering all consumers from different
flights in one ‘hub’, flights originating from that hub are more likely to be
fully utilised, and it might even be possible to fly larger planes – which
again are subject to economies of scale (it is not twice as expensive to fly
400 passengers as 200). Can you explain why?
7.3.3 Learning economies or the learning curve
Again, we start with a definition of economies of learning:
Economies of learning – definition
MCq > MCq+1
The main difference in comparison with economies of scale and scope is
that the decrease in marginal cost here is in response to an increase in
cumulative output, while the economies of scale and scope refer to
cost decreases in response to production rate. Economies of learning
describe the phenomenon where it is cheaper to produce an additional
unit – that is, the marginal cost of production decreases. This does not
include fixed cost, rather it indicates that having produced q units, it
becomes cheaper to produce an additional one.
Typically, learning economies materialise over a period of time – a firm
cannot decide to produce twice the amount in one go and expect to benefit
from learning economies. Learning takes place for a number of reasons.
Employees learn the tricks of the trade. For example, they find ways of
doing their work more quickly because they leave out unnecessary steps,
do not have to think about the process, they produce less waste and/or
they redesign the production process to make it more convenient and
efficient.
Firms get better at staff selection. When a new product is launched,
a firm may not know how to identify a successful salesperson for this
product or what type of previous experience is helpful. Perhaps most
saliently, think of a firm entering a new geographical market. It might
have a general idea as to what makes a successful product manager (say,
a management degree, an ability to work under pressure and some work
experience in a similar industry), but adapting to the local tradition may
be something the firm has no expertise in, and it will have a hard time
recruiting. Over time, however, these criteria will become more clear, and
staff selection will improve once the firm has had some experience in the
market.
74
3
Source: www.virgin.com
Chapter 7: Strategic asymmetries – persistent dominance over time
A similar argument applies to the selection of suppliers and
investment choices in general. For a new product, it may be unclear
what makes a good supplier, but it will become so over time, so that future
‘batches’ of the product will enjoy lower unit cost simply because the
suppliers have been selected more efficiently – thus either achieving lower
prices or higher reliability.
As much as organisations are able to learn, they can also forget.
Organisational forgetting describes the process of a firm’s marginal
cost increasing over time. This may have a number of reasons,4 much
like the process of individuals forgetting skills. To identify the sources
of organisational forgetting, we have to identify the ‘location’ of
organisational knowledge (is it the individual, the workgroup, the product,
the process, etc.?) and look for reasons why these might not be stable over
time.
Benkard (2000)
analyses the dynamics of
organisational forgetting
in a particular setting,
the wide-body aircraft
industry.
4
Activity 7.3
Which industries would be particularly affected by organisational forgetting and why?
7.4 Dynamic capabilities
Besides economies of scale, scope and learning, dynamic capabilities
are another important explanation of persistent asymmetries of firm
performance. Over the past 20 years strategy scholars have devoted
considerable effort to developing a coherent theory of dynamic capabilities
to complement the rather static perspective of the resource-based view
(discussed in Chapter 6). While resources and operational capabilities can
be used to explain differences in firms’ performances at certain points in
time, they do not explain how firms can cope with changing environments
or even create them. Thus, the aim of dynamic capability research is
to close this gap and understand how firms can sustain a competitive
advantage by responding to and creating environmental change.
While many definitions of dynamic capabilities exist, most scholars in
the field would agree with the one provided by Helfat and colleagues.
They define dynamic capabilities as ‘the capacity of an organisation to
purposefully create, extend, or modify its resource base’. This definition
makes it clear that the dynamic capabilities approach is an extension of,
or complement to, the resource-based view as it answers the question
how the resource base of an organisation is created and changed, where
the resource base comprises the tangible and intangible assets of the
firm. To work on its resource base an organisation must sense, shape and
seize opportunities and match these opportunities with its tangible and
intangible assets to outperform its competitors. Opportunities are not
given (like in a textbook optimisation problem) with given constraints,
but they must be discovered and shaped by searching across different
technologies and markets. Dynamic capabilities are especially important to
organisations working in an international context and firms in industries
characterised by fast technological change. The international context
provides a firm with the opportunity to explore poorly developed markets
and technical change makes it more important to combine different
inventions and complement these inventions with the corresponding
managerial capabilities (e.g. marketing capabilities). However, not every
response of a firm to changes in the environment can be seen as a proof
for the existence of an organisation’s dynamic capabilities. A firm can also
change in a non-deliberate ad hoc problem-solving way, for example, if a
regulation authority forces them to do so.
75
MN3119 Strategy
7.5 Key concepts
•
Sources of firm asymmetries
•
Economies of scale
•
Economies of scope
•
Hub-and-spoke networks
•
Complementary assets
•
Economies of learning
•
Umbrella branding
•
Learning curve
•
Dynamic capabilities.
7.6 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
explain sources of firm asymmetries
•
describe available strategies firms can employ to increase their
industry dominance.
7.7 Sample examination questions
1. Economies of learning are nothing but economies of scale over time.
Do you agree?
2. Most economies of scope have their cause in some form of scale
economies – true?
3. Which of the distinctive capabilities discussed in Chapter 6 is most
similar to dynamic capabilities?
Guidance on answering these Sample examination questions can be found
on the VLE.
Extended activity: competition in the wide-body aircraft
market
Read the following and then answer the questions at the end.
Reading
Cabral, L. and T. Kretschmer ‘Competition in the wide-body aircraft market’.
Available at: http://pages.stern.nyu.edu/~lcabral/teaching/widebody.pdf
Questions
1. Selling aircraft is very intense and often erodes most of the profits
from a sale. What could be reasons for that? Would you expect to find
this in other industries as well?
2. Currently, the wide-body industry is a duopoly of Airbus and Boeing.
Is this a stable long-term market structure? Why or why not? Pay
particular attention to the forces against monopolisation.
76
Chapter 8: Organisation design
Chapter 8: Organisation design
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
explain the interaction of organisation design and strategy at different
levels
•
explain how different technological and structural elements of
organisation design interact
•
discuss how organisation design affects organisational decisionmaking.
Essential reading
Csaszar, F. ‘Organizational structure as a determinant of performance: evidence
from mutual funds’, Strategic Management Journal 33 2012, pp.611–32.
Further reading
Hitt, M., H.I.R. Volberda, R. Morgan, R. Hoskisson and P. Reinmoeller Strategic
Management: Competitiveness and Globalization (Hampshire: Cengage Learning
EMEA, 2011).
Sah R.K. and J.E. Stiglitz ‘The architecture of economic systems: hierarchies and
polyarchies’, American Economic Review 76(4) 1986, pp.716–27.
8.1 Introduction
A further important source of competitive advantage is the way in which
a firm is organised. We could think of this as a form of architecture (as
discussed in Chapter 6), but it is really more than that. Organisation
design refers to the structure of who knows certain things and who makes
decisions about certain things. It determines which of a firm’s actions are
supported by the firm’s structure and which ones are hindered. Therefore,
one of the key lessons from this chapter will be to recognise that
organisation design really does interact with many other elements in the
firm. Of particular interest for this course will be the interaction between
organisational structure and the firm’s strategy.
8.2 Strategy and structure
Strategies are not chosen in a vacuum. Rather, strategies have to fit an
organisational structure as it is this that provides the framework within
which the strategies will be used and implemented. That is, a good
strategy can turn out to be unsuccessful if the structure is not right for
it. Note that not all strategies fit equally well with all kinds of structures
(which should become obvious the moment you look at the world and see
how different firms are, even in the same market). This means that the
organisation design has an interesting role: there is not ‘one best’ structure
that all firms should aspire to, but managing to align a firm’s structure
with strategy can be a source of competitive advantage.
77
MN3119 Strategy
The fit between strategy and structure has to be given at different levels,
as illustrated by Figure 8.1. Within a business unit, it is important to
match business-level strategies and a firm’s functional structure, while the
different divisions have to be arranged so they fit within the corporatelevel strategy of the firm, and the worldwide structure has to be aligned
with the way a firm wants to operate internationally. We will discuss each
of these ‘matches’ in turn later on in this chapter.
Of course, over time a firm will undergo an evolution, which will be
reflected in the typical organisation structure of a firm for its age and stage
in the lifecycle. Most firms start out with a simple structure with little
by way of formal organisation. As a firm grows, it becomes increasingly
difficult for a single person (often the founder) to manage and implement
all aspects of the business. Coordination problems ensue and maintaining
control over all subordinates, all with different functions, becomes
impossible. This typically creates a need for a functional structure in which
a CEO and limited corporate staff lead the firm, but there are functional
line managers in the dominant organisational areas – depending on the
type of firm, this could be production, R&D, accounting, etc. As the firm
continues to grow – and especially continues to diversify into different
geographical markets (more on that later) – many firms make the
transition into a multidivisional structure. Such a multidivisional structure
consists of operating divisions, each of which represent a separate
business, and the corporate top officers delegate authority to division
managers. Of course, this is an idealised evolution, and if you look around,
many firms seem to take this path.
Business-Level
Strategies
Functional
Structure
Corporate-Level
Strategies
Multidivisional
Structure
International
Strategies
Worldwide
Structure
Figure 8.1: The fit between strategy and structure.
8.2.1. Matching business-level strategy and functional structure
Take the two generic strategies proposed by Porter: cost leadership or
differentiation. What structure should an organisation choose if it wants to
follow one of these strategies?
Figure 8.2 shows a structure a firm with a cost leadership strategy might
try and implement. There is a fair number of central staff that coordinate
the different functions; operations is the main function because cost-saving
potential can be realised most efficiently through streamlining operations.
Efforts to improve the firm are focused on process engineering rather than
product R&D, and job roles are highly structured so that a low-cost culture
can be fostered and slack reduced. This mechanistic structure is most
efficient whenever cost leadership is the key aim of the firm.
78
Chapter 8: Organisation design
Office of the
President
Centralised Staff
Engineering
Marketing
Operations
Personnel
Accounting
Figure 8.2: Structure of a firm with a cost leadership strategy.
A typical structure for implementing and supporting a differentiation
strategy is given in Figure 8.3. Here, most functions are decentralised,
with the exception of R&D and marketing, which are most important for
the overall strategy of the firm and work together very intensively. There
is an emphasis of product R&D rather than process engineering, and this is
supported by marketing that does not just sell existing products, but also
engages actively in market research to generate and assess new product
ideas. Generally, the structure is much less formalised to make sure new
ideas are not stifled by formal roles and procedures and change can be
realised quickly. This is true both at the organisational level, but also at the
level of individuals, whose job roles are not highly structured.
Office of the
President
R&D
New Product
R&D
Operations
Marketing
Marketing
Human
Resources
Finance
Figure 8.3: Structure of a firm with a differentiation strategy.
8.2.2 Matching corporate-level strategy and multidivisional
structure
There has to be a match between the different types of multidivisional
structures and corporate-level strategies. Generally, one can distinguish
between the cooperative, the competitive and the strategic business unit
(SBU) form of multidivisional structures, which of course can be linked to
unrelated or related strategies.
We can identify both operational and corporate relatedness: operational
relatedness refers to the sharing of activities among different business
units, while corporate relatedness describes the transfer of corporate
core competencies from the headquarters into different businesses.
This distinction leaves us with a 2×2 matrix of possible multidivisional
strategies in Figure 8.4.
79
MN3119 Strategy
Low
High
High
Related
Constrained
Diversification
Operational and
Corporate
Relatedness
Low
Operational Relatedness
Corporate Relatedness
Unrelated
Diversification
Related Linked
Diversification
Figure 8.4: A 2×2 matrix of possible multidivisional strategies.
What now are the best multidivisional organisational structures for these
strategies?
For a related constrained strategy, i.e. high operational relatedness, but
low corporate relatedness, the preferred organisational form is shown in
Figure 8.5.
President
Government
Affairs
Legal Affairs
Corporate
R&D Lab
Strategic
Planning
Corporate
Human
Resources
Corporate
Marketing
Corporate
Finance
Product
Division
Product
Division
Product
Division
Product
Division
Product
Division
Figure 8.5: The preferred organisational form for a high operational relatedness,
but low corporate relatedness strategy.
The emphasis of this structure is on cooperation and sharing across
units. Many of the key functions are centralised, especially R&D, and
the corporate office emphasises centralised strategic planning, HR and
marketing to foster cooperation between divisions. Structural integration
devices create tight links among all divisions so that all units are aware
of the overall corporation’s goals, and overall contribution as well as
divisional performance are rewarded.
At the other extreme, there is very little integration in the competitive
form shown in Figure 8.6.
80
Chapter 8: Organisation design
Headquarter Office
President
Legal Affairs
Division
Finance
Division
Division
Auditing
Division
Division
Figure 8.6: The competitive form.
Here, the corporate headquarters are very lean, with few staff and almost
no input into business-level strategy. Therefore, finance and auditing are
the most important functions at headquarters as these are mainly used to
manage cash flow and allocate resources to finance, whatever the competing
divisions require money for. It is clear that any synergies across divisions
are likely to be minimal, which is why this structure is most efficient for an
unrelated diversification strategy where the expected synergies across units
are fairly low to begin with.
Somewhere in between the two extreme is the multidivisional form we
see most often – the strategic business unit (SBU) form. Here, SBUs are
fairly independent from each other, but integration across functions within
each SBU is important and resources may be provided by the corporate
headquarters for these functions – which does more than simply allocate
funds, however. They develop and implement a strategic plan, and most
central functions are strategic in nature. Central staff serve as a consulting
unit for the SBUs rather than having direct input in the strategies of the SBUs.
They might serve as consultants for general issues, but most importantly they
may also advise on how best to implement the centrally developed strategic
plan. The SBU form is summarised in Figure 8.7.
Headquarter Office
President
Corporate
R&D Lab
Corporate
Finance
Division
Corporate
Marketing
Strategic
Business Unit
Strategic
Business Unit
Division
Strategic
Planning
Division
Division
Division
Corporate
Human
Resources
Strategic
Business Unit
Division
Division
Division
Division
Figure 8.7: The strategic business unit (SBU) form.
8.2.3 Matching international strategies and worldwide structure
Finally, corporations have to make sure that the structure they give their
international operations matches their international strategy. The key
determinant here is the degree of autonomy given to the local operations
(through organisational structure) and the extent to which a firm’s products
in each country are supposed to ‘feel’ like local products.
81
MN3119 Strategy
A multidomestic strategy aims at mimicking a local business in each
country (or economic area) a firm operates in. That is, products and
product mixes will be adjusted and will cater to local tastes, management
will most likely be local, and local branches should ideally almost be
indistinguishable from a domestic firm. As far as organisation design
is concerned, operations are almost completely decentralised, and the
corporate headquarters coordinates financial resources among independent
subsidiaries. The organisation runs more or less like a decentralised
federation. Of course, synergies across geographical areas are low in this
case, but in turn there is a great deal of differentiation by local demand and
fit to the area and local culture. This is shown in Figure 8.8 below.
United
States
Asia
Latin
America
Multinational
Headquarters
Europe
Middle
East/
Africa
Australia
Figure 8.8: A multidomestic strategy.
The polar opposite is a global strategy in which the product offering
and product design are identical for all the geographical markets in
which a firm operates. In this case, the different product divisions
operate worldwide, and there is a fair bit of centralisation to coordinate
the information flow among worldwide products. To facilitate global
economies of scale and scope, the corporate headquarters uses a series
of intercoordination devices and allocates resources in a cooperative way
(i.e. it does not encourage competition among business units and/or
geographical markets). The organisation effectively runs as a centralised
federation, as shown in Figure 8.9.
Worldwide
Products
Division
Worldwide
Products
Division
Worldwide
Products
Division
Worldwide
Products
Division
Figure 8.9: A global strategy.
82
Worldwide
Products
Division
Global
Corporate
Headquarters
Worldwide
Products
Division
Europe
Chapter 8: Organisation design
Finally, a transnational strategy reflects a mixture of the two polar cases
above. In such a strategy, the corporation attempts to balance the benefits
from the economies of scale and scope accruing from operating on a global
scale while granting sufficient autonomy to its local subsidiaries. For
example, local subsidiaries might be able to select from a global product
portfolio the products they expect to be most profitable in their market,
an assessment they may have gathered through extensive (local) market
research. Accordingly, performance and success for the local subsidiaries
are judged by the profit generated for the overall corporation. There may
also be hybrid forms where some business units are structured around a
product (or product group), while others reflect geographical markets.
Deutsche Telekom may reflect such a hybrid strategy, where T-Systems,
T-Home, T-Mobile represent product-oriented organisational units, while
Telekom US covers the activities in a particular geographical market. This
type of structure is reflected in Figure 8.10. below.
Headquarters
Product Division
A
Geographic Area
Division 1
Area 1
Product
A
Area 2
Product
B
Product Division
B
Geographic Area
Division 2
Area 1
Product
A
Area 2
Product
A
Figure 8.10: A hybrid transnational strategy.
8.3 Organisation design and competitive advantage
Organisation design is not a means in and of itself. Much rather, it is a tool
that can help support and ensure strategies at all levels of the organisation
is used correctly. While related to architecture, it is not the same and
encompasses more elements. Most importantly, there is no one ‘optimal’
structure, but rather a fitting structure for each strategy a firm might want
to follow. As such, competitive advantage can result from using this tool to
reinforce successful strategies.
8.4 Key concepts
•
Strategy and structure
•
Organisational fit
•
Functional organisation
•
Multidivisional structure
•
Worldwide structure.
8.5 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
explain the interaction of organisation design and strategy at different
levels.
83
MN3119 Strategy
•
explain how different technological and structural elements of
organisation design interact.
•
discuss how organisation design affects organisational decisionmaking.
8.6 Sample examination questions
1. Give an example of a firm in which units have low operational
relatedness but high corporate relatedness.
2. Which international organisation would you pick for a restaurant chain
originating from Asia but operating across the globe?
3. How would a firm following a differentiation strategy in the fastmoving consumer goods industry organise their operations and why?
Guidance on answering these Sample examination questions can be found
on the VLE.
84
Chapter 9: Value chain analysis: vertical relations
Chapter 9: Value chain analysis: vertical
relations
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
explain the concept of double marginalisation and why a vertically
integrated monopolist would have made higher profits and charged a
lower price
•
describe and apply to real life/hypothetical situations the concepts of
‘market cornering’, ‘investment externalities’ and ‘market splitting’.
Essential reading
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) Chapter 11.
Saloner, G., A. Shepard and J. Podolny Strategic Management. (New Jersey: Wiley,
2006) Chapter 10.
Further reading
Monteverde, K. ‘Technical dialog as an incentive for vertical integration in the
semiconductor industry’, Management Science 41(10) 1995,
pp.1624–38.
9.1 Introduction
Many of the transactions in the business world are not in the endconsumer market (or B2C, as the dot-com hype has taught us), but are
in fact transactions between businesses. A firm selling raw material
to a manufacturer, a producer selling goods on to a retailer, or a parts
producer delivering products for final assembly are all examples of socalled ‘vertical’ transactions between businesses that do not involve an
end consumer. This has a number of important consequences for the
analysis of such transactions. First of all, both players often have some
degree of market power, that is, firms at both stages of the vertical chain
are aware that changes in the price they charge will have an impact on the
quantity demanded by the end consumer. That said, both parties also try
to maximise their profits in the course of the transaction.
In this chapter, we will look at some specific issues arising from the
vertical relation between firms and how firms may try and improve their
competitive position through them, not only in the vertical transaction, but
also in comparison to their direct competitors in the product market.
9.2 Double marginalisation
Consider a simple situation with two players: an upstream firm (U) selling
a product to a downstream firm (D), which then sells the product to end
consumers with a demand function D(p). We assume that both firms are
monopolists in their respective market stages. Figure 9.1 illustrates this
scenario and the notation.
85
MN3119 Strategy
Upstream
firm (U)
Downstream
firm (U)
End
consumers
Figure 9.1: Value chain.
We assume that the marginal cost of producing the intermediate good is
constant at MCU = c, and the marginal cost of selling the product to end
consumers is zero. It may help if you think of U as a manufacturer and D
as the retailer – a retailer has no extra incremental cost on top of what he
pays the manufacturer.1
What price will the manufacturer set? The retailer? Of course, both will try
to set as high a price as possible, but they have to consider the impact this
will have on final-good prices as well.
As both firms are monopolists in their markets, they will maximise profits
by setting MR = MC. We start with the retailer: as its marginal costs are
assumed to be zero, he should sell a quantity such that MRD = 0, right?
Not quite – we have to take into account that the retailer still needs to
purchase the product from the manufacturer. As we can see from Figure
9.2, the transfer price for that is pt. In fact then, the marginal cost the
retailer faces is then pt + MCD = pt. Therefore, the quantity sold by an
upstream firm with monopoly power will be such that MRD = pt.
Double marginalisation
MC
MR
D
DD
Figure 9.2: Double marginalisation I.
What does this imply for the upstream firm? The upstream firm knows
that if it sets a transfer price pt, it will sell a quantity (to the downstream
firm) determined by the marginal revenue curve for the final goods
market. In other words, the downstream marginal revenue curve becomes
the demand curve of the upstream firm! This then enables us to analyse
the upstream firm’s problem – with a demand curve MRD, the marginal
revenue curve for the upstream firm will shift further in (i.e. will be even
steeper). The manufacturer will then choose a transfer price that sets
MRU = MCU.
86
The assumption of
zero marginal cost for
the downstream firm is
not important for the
intuition of our model,
but it makes life easier.
One of the exercises
will ask you to analyse
the problem, but with
positive marginal sales
cost.
1
Double marginalisation
Chapter 9: Value chain analysis: vertical relations
MC
MR
U
MR D = D U
Figure 9.3: Double marginalisation II.
The final goods price is then based on a transfer price by the manufacturer
that exceeds marginal cost, and the corresponding final goods price further
exceeds the price paid by the retailer – both players add their own profit
margin; double marginalisation.
Activity 9.1
How would the results change if:
a. The downstream company had positive marginal cost?
b. Marginal cost for the upstream firm was increasing (and MCD = 0)?
c. The market for the upstream good was competitive?
Illustrate your answers graphically.
Guidance on this activity can be found in the VLE.
A single (i.e. vertically integrated) monopolist would have made higher
profits and would have charged a lower price. To see why, consider the
monopolist’s decision rule, which is setting marginal revenues in the
final goods market equal to total marginal cost c. In fact, this is often
considered one of the main motivations for vertical integration. There
are, however, other techniques of getting to the joint-profit-maximising
outcome, as illustrated in the following activity.
Activity 9.2
Suppose that the upstream firm can charge a transfer price pt and a fixed fee F. Show
that this can generate the joint profit-maximising outcome. (Hint: Assuming that the
downstream firm has no additional marginal cost, what transfer price would give the best
(i.e. profit-maximising) outcome? Given this transfer price what conditions would the
fixed fee have to fulfil?)
Guidance on this activity can be found in the VLE.
Assuming that vertical integration or sophisticated contracting is not
possible (or not desired), another way of lowering prices is to encourage
competition in the related market. In other words, if a manufacturer
manages to sell products via two retailers, they will compete some of the
margins in the downstream market away, which is unambiguously good
news for the manufacturer, since lower margins imply higher volumes at
the same transfer price – and therefore higher profits (the same applies to
87
MN3119 Strategy
a retailer purchasing from several competing manufacturers). As a general
rule, we can say that the more intense competition in the related market,
the higher the profits for the firm on the other side of the market. In the
extreme, this brings us back to the situation where a firm is faced by a
large number of identical players which do not really have any market
power – in other words, the more players there are in the related market,
the more the situation resembles a consumer market.
9.3 Vertical foreclosure
In the previous section, we considered only a single upstream and
downstream firm. In this section, we extend this stylised situation and look
at some phenomena that might occur if one stage of the value chain has
several firms operating in it.
9.3.1 Market cornering
As we have seen in the previous chapter, firms without any obvious
capabilities or advantages over their rival can still manage to secure an
advantageous position in the market by controlling essential resources
or inputs, similar to owning a strategic asset. This is ‘cornering the
market’: a product consisting entirely of generic components may still
end up being very profitable if one of the components is controlled by
a single firm – which can then either sell it profitably to its rivals or
monopolise the market altogether.
Example: Minnetonka Corporation.
One firm that has been very successful with cornering the market in the face of major
(potential) competition was Minnetonka Corporation. In 1980 it was the first firm to
introduce liquid soap in the US market, but knew that if it were to take off, it would be
smothered by Procter & Gamble and Unilever which could spend more resources on
marketing and enjoyed a bigger brand reputation. Recognising this, Minnetonka also
realised that in order to sell liquid soap, small, plastic pumps would be necessary – a
product that had hardly any other use at the time. Consequently, Minnetonka bought up
two years’ worth of supply of plastic pumps – much more than it would typically commit
to – and therefore cornered the market and forced the major players to build up their
own capacity. Minnetonka’s privileged position did not last forever of course, but it was
able to enjoy two years undisturbed by major competitors. Ultimately, the liquid soup
business was bought by Colgate Company.2
9.3.2 Investment externalities
To market a product successfully, investment on different stages of the
value chain is necessary. A shop selling luxury products will typically
be decorated carefully – which in turn increases the returns to any
investment in quality by the luxury goods producers. This is not a problem
per se (as both players benefit from investments in quality), but it may
get more complicated if there are several competing retailers of the
same luxury good. Suppose they are located close to each other, so that
an advertisement by either of the two retailers increases demand for
the product, but that consumers split up more or less equally between
the two retailers. This would imply that whoever bears the cost of the
advertisement does not get the full benefit because the other retailer
receives a positive externality. The private benefits to advertising therefore
are lower than the benefits to the manufacturer, which leads to an
inefficient advertising level from the manufacturer’s point of view.
88
Source: http://inventors.
about.com/library/
inventors/blsoap.htm
2
Chapter 9: Value chain analysis: vertical relations
There are several ways to solve or at least reduce this problem. First, the
producer can pick up the bill, which it may be reluctant to do because the
investment benefits the retailers as well (in other words, the producer’s
private benefits do not include the retailers’ profits). Second, the producers
can exert resale price maintenance (RPM) so that retailers have no
incentive to compete on price – the worst scenario from a retailer’s point
of view would be that it spends money on advertising and the rival retailer
lowers its price to pick up the higher demand. Third, a manufacturer can
grant exclusive geographical territories to retailers, which would help to
minimise the spillovers from one retailer to the other – this will then help
align the incentives of retailer and producer.
Activity 9.3
Why do Pepsi and Coca-Cola insist on exclusive dealing in most major venues?
Guidance on this activity can be found in the VLE.
9.3.3 Market splitting
Vertical relations are not always a move to weaken direct competitors. In
fact, sometimes restricting vertically related markets can help maintain
collusion. For example, a firm that has significant switching costs by selling
complementary products (i.e. vertically related products like services,
software, etc.) may end up creating two monopolies covering half the
market rather than with one – a genuinely competitive market. In this
case, restricting your competitor’s access (and vice versa) to a vertically
related product will create such a market split; neither you nor the
competitor has an incentive to steal customers from each other and two
cosy monopolies can co-exist side by side (see example below).
Example: Fire pumps
Hale Products and Waterous jointly hold approximately 90 per cent of the market
for fire pumps. Their customers, fire stations in the US, typically have to sign an
exclusivity agreement with one of them to use their spare parts and services. While
this could potentially lead to a competitive duopoly in which firms compete for new
consumers vigorously, it turned out that the ‘duopoly’ was in fact a happy couple of two
geographically divided monopolies – Hale Products controlled one half of the market,
Waterous the other half, and vertical restrictions (through exclusive dealing agreements)
meant that there was no effective competition.3
Source: www.ftc.gov.
opa/1996/07/ hale.htm.
3
9.4 Key concepts
•
Double marginalisation
•
Intermediate goods
•
Upstream and downstream firms
•
Vertical foreclosure
•
Market cornering
•
Investment externalities
•
Market splitting.
89
MN3119 Strategy
9.5 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
explain the concept of double marginalisation and why a vertically
integrated monopolist would have made higher profits and charged a
lower price
•
describe and apply to real life/hypothetical situations the concepts of
‘market cornering’, ‘investment externalities’ and ‘market splitting’.
9.5 Sample examination questions
1. Exclusive dealing benefits not only the manufacturers (i.e. the
upstream firm), but also the retailer. Why is that?
2. Multiple sourcing is the practice of purchasing an input or raw
material from several different sources. How does this affect the
negative consequences of double marginalisation?
3. Under which circumstances can cornering the market work? When is it
likely to be unsuccessful?
Guidance on answering these Sample examination questions can be found
in the VLE.
Extended activity: outsourcing at Eriksson and Sony and
Loews
Read the following and answer the questions at the end.
Reading
Baccara, M. et al. ‘Sony and Loews’. NYU, Stern School of Business. Available at:
http://pages.stern.nyu.edu/~lcabral/teaching/sony.pdf
Backus, D. and L. Cabral ‘Outsourcing at Ericsson’. NYU, Stern School of Business.
Available at: http://pages.stern.nyu.edu/~lcabral/teaching/ericsson.pdf
Questions
1. Read the minicase on outsourcing at Ericsson and answer the
questions at the end of the case.
2. Read the minicase on Sony and Loews and answer questions (a) and
(b).
90
Chapter 10: Vertical integration and transaction cost
Chapter 10: Vertical integration and
transaction cost
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
comment upon the links between ‘core competencies’ and ‘distinctive
capabilities’ to the vertical relations and the value-based strategy
literature
•
describe and explain the sources and types of transaction costs.
Essential reading
Besanko, D., D. Dranove, M. Shanley and S. Shaefer Economics of Strategy. (New
Jersey: Wiley, 2009) Chapters 5 and 6.
Further reading
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) Chapter 3.
10.1 Introduction
Building on the previous chapter, the incentives for foreclosure are often
only one part of the decision to make or buy a particular input. In fact,
many firms are not even in the position to foreclose entry strategically
or exploit their market power vis-à-vis competitors and/or suppliers or
buyers. Most frequently, the decision to make or buy is governed by cost
considerations, although assessing these costs is often more intricate than
simply comparing the price of the input on the free market with the cost
of producing in-house. In this chapter, we will briefly discuss the different
costs and how firms can influence or avoid them.
10.2 Purchasing versus production costs
Firms often feel they are paying too high a price for inputs bought
externally. However, if the alternative is producing the input at a higher
cost than the market price, then buying in still makes economic sense. In
other words, if the market price (including a mark-up for the supplier)
is still lower than the internal production cost, a firm should ‘buy’ rather
than ‘make’. This could be the case if there are significant economies of
scale of production that do not apply at the quantities demanded by the
firm – as can be seen in Figure 10.1. Firm i would only produce a small
amount itself, while the supplier could take advantage of economies of
scale, collect a mark-up and still offer the product cheaper than firm i
could produce it internally.
91
MN3119 Strategy
Ci
PM
Qi
QM
Figure 10.1: Cost and demand functions for purchasing versus own production.
Activity 10.1
Consider an input with a marginal cost function of MC = 1 + 1/(1 + Q), i.e. marginal
costs are decreasing with output. Market demand (including firm i) for the product is
Q = 10 – P.
a. What is the price a monopolistic supplier would set?
b. How much would firm i need to produce for its own use (i.e. not selling to or buying
from the external market) for internal production to be more profitable than buying
from a monopolistic supplier?
c. (Optional:) If supply on the outside market were competitive, how much would firm i
need to produce to prefer in-house production?
Guidance on this activity can be found in the VLE.
These economies of scale could come from ‘true’ benefits to scale, such
as production economies, but they could also originate from greater
specialisation by the input suppliers. For example, outsourcing of
business processes such as payroll services has the advantage that a firm
specialising purely in payroll systems will be better at it than a firm with a
different focus.
As we have established in the previous chapter, the effect will be weakened
by the degree of competitiveness of the input market. If it is relatively
competitive, market price will be closer to marginal cost, which means that
a firm would have to produce even larger quantities to ‘beat the market’.
(This can also work the other way if the market is a natural monopoly
and the economies of scale are so strong that prices are higher with two
smaller operators than with a single monopolist.)
10.3 Coordination costs
This cost is often ignored in the calculation of total purchasing cost –
informally, we could call it the loss of ‘peace of mind’ when purchasing
from an outside supplier. It basically stems from the fact that suppliers will
have priorities other than you if the contract does not allow for the cost of
delay, or a faulty product, would have for you. Consider a simple example:
a restaurant requires its produce and meat to be delivered by 2pm, and
the contract it signed with its suppliers states that every hour’s delay will
92
Chapter 10: Vertical integration and transaction cost
incur a contractual penalty of $x. Suppose now that the cost of a one-hour
delay is $y < $x, but a two-hour delay would mean that the chef has to
rush preparations or even purchase meat from another source. The cost of
delay therefore increases dramatically, and two hours’ delay means a cost to
the chef of $z > $2x. Can this be a problem? If every restaurant requested
delivery at 2pm, it might make sense for the supplier to deliver orders along
the most convenient (and cost-efficient) route – which could mean that our
chef receives the delivery an hour late, which is not a problem as such since
the penalty of $x covers the extra cost of $y and, assuming that the supplier
obtains a cost saving of at least $x, everybody gains from this rescheduling
of deliveries. But suppose that an additional hour’s delay would give the
supplier another cost saving (or additional job he can take on) of equal size
(> $x), he would be willing to pay another hour’s penalty, which would send
the chef into a (costly) panic which would not be covered by the additional
penalty payment.
In the example above, the problem is that the penalties do not increase
according to the real cost to the restaurant – the key insight is that by
purchasing externally, it is important to realise that the supplier’s incentives
will be different to that of any one individual firm.
Of course, a similar argument would also be true for the quality of an input.
If the incentives to cut corners for a supplier are not adequately offset by
severe enough controls or contractual penalties, it might lead to a situation
where a supplier produces with a higher tolerance for small imperfections
and willingly takes the penalties into account because they are outweighed
by the cost savings achieved. In the extreme, if the cost of delay, faults and
so on are prohibitively high, a supplier might not even be willing to enter a
contract that would adequately reflect these costs. If the risk is too high, a
supplier might prefer not to take the risk, and compensating the supplier for
it would be prohibitively costly.
10.4 Proprietary knowledge
Frequently, making or buying a particular input is not governed by a
calculation of the cost, even if the costs of producing in-house are very high
and there is a ready and competitive supply on the market. For example, one
might argue that there are plenty of small research laboratories (or at least a
number of entrepreneurial scientists) that would be able to offer outsourced
research services to larger organisations. (In fact, this is already happening
to a certain extent, but the majority of research-intensive firms will still do
their research in-house.) The key reason is that the closer an activity is to
the core competencies (or is a distinctive capability) of a firm, the more
concerned the firm will be to protect this intellectual property. While it is
still possible for employees to leave and take company secrets with them,
there is a much larger degree of control over in-house activities than
outsourced activities, so if the firm wants to protect proprietary knowledge,
it is likely to keep activities in-house. A particular case of knowledge kept
in-house is the control over the selection of staff – again, companies
performing outsourced business functions are often motivated by different
performance targets (their own!) than the organisation that outsources.
However, keeping knowledge in-house need not be a distinctive capability
or crucial to the firm’s success. For example, information about demand for
one firm’s product might also be transmitted via supply prices and delivery
schedules to rivals. The danger of having rivals learn about the success of
a particular market or product then might prompt a firm to produce inputs
in-house.
93
MN3119 Strategy
10.5 Transaction costs
The term transaction costs defines those costs arising from an exchange
relationship – which could be governed by a written contract or an implicit
or relational contract. Again, there are direct costs and indirect costs –
indirect costs may often affect the value of a transaction more significantly
than direct ones.
Direct costs are the time spent on specifying a contract, lawyers’ fees
for drawing up the contract, writing job descriptions and so on. Most
often, these will be much lower within the firm than on the market –
predominantly because a contract need not be written. Another direct cost
would arise from monitoring the quality of the exchanged product and
from the cost of faulty items – if someone has to spend time and effort
removing faulty products that the other party has delivered and if this
screening process is not perfect, costs arise from completing this specific
transaction.
Indirect costs are more difficult to quantify, but they often have more
bearing on the decision to make or buy – if indirect transaction costs
are too high ‘it’s just not worth the effort’. Indirect costs mainly arise
from the fact that contracts are incomplete in the sense that they do not
specify precisely what should happen (or what both parties are obliged
to do) for any possible contingency in the market. If a situation arises
that has not been specified in the initial contract this is then subject to
haggling and negotiations, and the expected losses from the partners
behaving opportunistically. Both parties will have to take into account
that incentives are not perfectly aligned and that the other party will
try to maximise its own pay-offs if the finer points of the contract
allows. Insurance for many types of natural disasters are subject to large
uncertainties in the determination of damages and payouts. For example,
Hurricane Katrina had (and still has) insurance companies and insured
parties in bitter fights because many policies exclude water damage, but
include wind damage – it is now a matter of interpretation if the damage
to property has been caused by the severe winds or tidal waves – and
naturally both parties claim the outcome most favourable to them, which
is surely going to generate a significant revenue stream for the legal
profession.1 Generally, indirect transaction costs are higher in market
transactions because the incentives are less aligned than within the firm.
This implies that if a contract cannot be specified well so that a large (and
costly) number of contingencies cannot be catered for, relationships are
likely to be within one firm.
Both elements of transaction costs are important – for example, it is easy
to imagine that a homeowner expecting to incur huge legal costs to secure
a payout following a natural disaster will decide not to take out insurance
despite the direct transaction cost being relatively small, with a standard
contract being offered over the phone, for example.
Activity 10.2
Using transaction cost arguments, explain why research and development is often kept
in-house, while firms purchase options to hedge against currency risk.
Guidance on this activity can be found in the VLE.
94
For a discussion of
the legal conundrum
surrounding Hurricane
Katrina, see www.
insurancejournal.
com/news/southeast/
2005/09/27/60219.htm
1
Chapter 10: Vertical integration and transaction cost
10.6 Asset specificity
Closely related to the discussion on vertical relations is the concept of asset
specificity. An asset is said to be specific if it has significantly less value
outside one particular relationship than it has within it.
Activity 10.3
Rank the following four investments in terms of their asset specificity and explain why.
Explain who makes the investment and who benefits from it.
a. A wind energy turbine is designed and built according to a terrain’s specific features.
b. A McDonald’s franchisee builds a ‘Ronald McDonald Playground’ in the back of the
restaurant.
c. McKinsey offers a mini-MBA to its new employees.
d. A firm invests in a high-performance server to improve the productivity of its logistics
and distribution.
Guidance on this activity can be found in the VLE.
A specific investment changes the dynamics of a relationship or transaction
quite significantly: the moment the investment has been made, the
investor is now vulnerable to opportunistic behaviour by its contractual
partner. Basically, before investing, anybody could have entered the same
transaction on either side. On the other hand, after the investment, the
two parties are bound together – this is what is called the fundamental
transformation from a ‘large-numbers bidding (or market) situation’ to
a Fundamental
‘small-numbers
(bilateral) bargaining scenario’.
transformation
Firm
A
B
C
Supplier(s)
1
2
3
Once the parties invest in relationship-specific assets,
the relationship changes from a “large numbers”
bidding situation to a “small numbers” bargaining
situation.
Figure 10.2: Fundamental transformation.
Why is this problematic? In a market situation, everybody can walk
away from a deal that does not seem favourable. In a situation where
an investment has already taken place, this is a lot harder. Contractual
parties will take this into account and, if possible, will try to exploit this by
holding up their partner:
Hold up – definition
If a contract is incomplete, party A can hold up B by trying to renegotiate the initial
agreement in its favour if leaving the contact unfulfilled would involve significant costs
for B and little for A.
The important concepts in this definition are the contractual
incompleteness and the difference in costs – if the contract is incomplete,
there is space for renegotiation, if it is complete, any attempt to change
the terms could simply be stopped by insisting on the initial terms of the
contract. However, not many contracts are complete. The reasons that we
do not fear being held up in all incomplete contracts is that the costs of not
fulfilling the contract are often relatively symmetric – a producer stands
to gain revenues from selling a product and a consumer stands to enjoy
use of a product, which makes trade between the two mutually beneficial.
On the other hand, if one party depends on this particular contract while
95
MN3119 Strategy
the other has plenty of alternative options (or simply is not that bothered
about this transaction) there is some scope for hold up. The term to
describe the degree of dependency on this particular transaction is the
concept of quasi-rent.
Quasi-rent – definition
Quasi-rent is the difference between the revenue the seller expects to receive under the
initial terms of the contract and the minimum revenues the seller must receive in order
not to exit the relationship.
Therefore, if one player has high quasi-rent and one has not, the former is
typically in danger of being held up.
10.7 Alternatives to ‘make’ or ‘buy’
We increasingly see firms choosing alternatives to pure vertical integration
or exclusive market contracting. Described below are a number of
alternatives.
10.7.1 Joint ventures and strategic alliances
As we will see in Chapter 12, strategic alliances and joint ventures are
often initiated over several stages of the value chain – as an alternative
to vertical integration. The main difference lies in the nature of contracts
between the partners – while vertically integrated firms will typically get
by without formal contracts between units, the most important contractual
element is the employment contract, which will often include an element
of monitoring so that the firm’s interests are secured, and a degree of
performance pay so that employees will have their incentives aligned
with the firm. A market transaction on the other hand is governed by an
exchange contract that specifies the terms of delivery and so on. Strategic
alliances and joint ventures are in-between in that they sometimes have an
element of joint asset ownership (in the case of joint ventures) and always
a degree of (formal or informal) profit sharing, which is used to align
incentives. This makes joint ventures and strategic alliances a hybrid form
since there is no authority in the same way as in an employment contract,
but no fully specified contract as in a market transaction.
10.7.2 Sub-contracting networks and franchising
Another alternative is to build and maintain sub-contracting networks
and engage in franchising. In some senses, this is similar to a market
transaction as the contracts try to specify the terms of sub-contracting
and so on, as much as possible with the main difference being that in
both cases (sub-contracting network and franchising) there is a relational
element to the contract. A franchisee would not enter the relationship
and invest accordingly if he did not expect the relationship to continue
for some time, very often beyond the end date of the initial franchising
contract. The long-term outlook and relational element will make it easier
and more attractive to invest in the relationship for both parties. Another
reason why such networks do not seem to trigger fears of being held up by
the franchiser is that news of opportunistic behaviour would very quickly
spread throughout the entire network of franchisees, which would lead
to a relatively severe backlash (in the sense of lowered investment by the
franchisees), not only by one franchisee, but by many of them.
96
Chapter 10: Vertical integration and transaction cost
10.7.3 Tapered integration
Tapered integration involves making some input in-house and buying
some on the market. Very often, the rationale given for this is that external
demand can be used to ‘buffer’ demand fluctuations of the input, while
the stable component of demand is produced in-house. It is important to
note, however, that the main difference between tapered integration and
the other two alternatives is that there is no attempt to align the incentives
of supplier and buyer – if anything, incentives are now even more opposed
as the internal input supplier and the external supplier are now in indirect
(or direct) competition for business from the downstream unit. Another
motivation for tapered integration is to lessen the danger of being held
up. If a firm can produce a significant proportion of input demand itself,
suppliers will not be able to hold up the firm by threatening to withhold
supplies.
10.8 Key concepts
•
Purchasing versus production cost
•
Make or buy
•
Coordination cost
•
Transaction cost
•
Asset specificity
•
Hold-up
•
Quasi-rent.
10.9 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
comment upon the links between ‘core competencies’ and ‘distinctive
capabilities’ to the vertical relations and the value-based strategy
literature
•
describe and explain the sources and types of transaction costs.
10.10 Questions for discussion
1. Discuss the relative advantages and disadvantages of the three
alternatives to make or buy compared with vertical integration and/or
market transactions.
2. Why do you think outsourcing is more frequent in mature industries
than in the early stages of an industry’s life cycle?
Extended activity: athenahealth
Read the following and answer the questions at the end.
Case study: athenahealth, US Healthcare industry
Based in Waltham, Massachusetts, athenahealth is an IT provider in the
healthcare industry that offers internet-based revenue-cycle management
services to physician enterprises – hospitals and doctors in private practice.
Founded in 1997 by Jonathan Bush, Chairman and CEO, and Todd Park,
Chief Development Officer, athenahealth achieved sales of nearly $40
million in 2005. The company’s mission is to get health care to work the
97
MN3119 Strategy
way it should – so that information is conveyed efficiently and accurately
to the benefit of both patients and doctors. athenahealth is backed by bluechip venture capital firms in the US, including Venrock Associates, Oak
Investment Partners and Draper Jurvetson.
athenahealth began not as a technology company but as a women’s
health practice focused on improving maternity outcomes for women and
babies. The idea for the company took hold when Bush, a recent graduate
of Harvard Business School, met Park at the consultancy Booz-Allen &
Hamilton where both worked as associates focused on the healthcare
sector. With entrepreneurial spirits high, they set out to build this business,
but soon found their goals thwarted by something seemingly very basic:
getting paid.
athenahealth was born from our struggle of actually running
a group practice, when the Herculian effort and complexity of
getting paid started taking time away from patient care. The
practice delivered high-quality patient care, but the mounting
red tape from payers meant cash flow was volatile at best. We
couldn’t get paid and we didn’t know why.
When we started researching solutions, we only found
companies selling hardware and software but nothing that
could help us get the results we hoped were possible. These
companies couldn’t give us back control of our practices.
We decided to tackle the problem ourselves and build a true
revenue cycle management solution—one that is used by
thousands of US physicians today.
(athenahealth website)
To date, athenahealth has been focused on developing and rolling out
an electronic billing solution that, by running back-office processes more
efficiently, boosts doctors’ profitability and power, particularly vis-a-vis
insurance companies. With this foothold firmly in place, the company is
now applying their skills in process control to the clinical side of a medical
practice – tracking lab orders, results and prescriptions, and streamlining
the medical group’s interface with these players.
Talking with Mr Bush reveals his vision of transforming the entire supply
chain in the healthcare industry from a flabby and fragmented consortium
into a system that ultimately delivers enormous cost savings and vastly
improved health care to patients. The company’s strategy remains focused
on medical groups as its primary customers, helping them become
strategic entities that, through gathering momentum, influence the way
every other player in the supply chain acts and interacts.
Excerpts from a conversation with Jonathan Bush about
athenahealth and the future structure of the supply chain in the US
healthcare industry:
Everything in health care comes out of an order from a
doctor. The order-making engine is the doctor. Associations of
physicians create an artificial shortage of their expertise so that
they can charge higher prices – and it works. The people who
get to decide how many people in the guild there should be are
the specialists in that specialty. So the doctor gets to control the
original supply of the order-making entity. It’s called great work
if you can get it. If you were a taxi driver, you would really like
to control how many taxis are competing for fares, right?
98
Chapter 10: Vertical integration and transaction cost
When you look at a supply chain, you want to look at where’s
the fat, where’s the place where there is extra value created
because everything is supposedly getting pushed toward
commodity prices all the time. Well the doctor gets to control
supply so he doesn’t get pushed so he isn’t forced to merge with
other doctors or to get economies of scale in his office, because
he can actually make a pretty tidy sum with a lady and a little
one-shingle office.
When you look at organisational scale in the supply chain, you
see that in the first link in the supply chain – the average office
has three doctors in it! That’s averaging in the Mayo clinic and
all the mega clinics – 55 per cent of medical doctors in the
United States are practising in groups of one or two doctors.
So there is not a lot of efficiency and scale in the entity that
is creating the orders that then flow through the rest of the
system. And so our first step is to help doctors in private medical
practices go from a guild to an operating entity, helping them to
achieve basic efficiencies in the information infrastructure.
Our second step is to move medical groups from an operating
entity to a strategic entity. So once you – the medical practice,
the order-making entity – have basic internal integrity where
you know how many orders you sent out, you know how many
results came back, you know how many claims you sent out, you
know how many got paid, how many didn’t, whether they got
paid enough – that basic stuff – then you can use that integrity
to move from being an operating to a strategic organisation. You
are choosing when to fight and where. A strategic organisation
is in the business of looking through the things they do and
picking and choosing – I only want to see patients with this or
that insurance company versus this other one, or I want to do
my own pap smears with my own laboratory because it seems
like there’s a lot of money to be made there. So they are making
decisions around what in the supply chain around me do I want
to control, and what in the supply chain around me do I want to
avoid. So that’s the second thing that we want to help medical
practices do.
And the third thing that medical practices don’t do – but will do
some day – is go from being a strategic entity to a system where
they actually use their market power to influence what happens
elsewhere, in part by coordinating with other medical practices.
An example of different entities becoming operating entities, and
then strategic entities and then a system is the Visa network. All
the different merchandisers in the visa network have agreed to
be a system – they have agreed to create this not-for-profit called
the Visa association that clears the transaction. Anybody who
wants to issue credit on a retail basis can use this really efficient
network, where, for pennies a customer can stick her card into
any slot in the world and somebody’ll figure it out. Sometimes
it’s debit and sometimes it’s credit, all you know is you shove the
thing in and the money comes out.
So that is the last step in the journey of that order-making entity.
Fundamentally we are still stuck at the guild level. We are in the
business of moving physicians from guilds to operating entities
to strategic entities, and we believe that someone will come
along and build a system using athena as an ingredient. And
99
MN3119 Strategy
the reason is because we have integrated all of the technology,
knowledge and work that go into the mechanics of an order
into a solution that doesn’t have any upfront costs. Just like the
visa network. When you’re a store and you want to take Visa
you don’t have to go buy some software into a computer and
install it to capture the Visa network. Visa says, “ok it’s going
to cost you 3.5 per cent of all the transactions you do and we’ll
show up, we’ll show you how it works, we’ll help you open your
account, and we’ll be watching, and as we deliver value to you,
we’ll take our piece.” [Indeed, what differentiates athenahealth
from competitors is its decision to share risk with customers by
charging a percentage fee of a doctor’s collections, rather than a
flat fee.]
So what athena is doing is for the first time providing a valuebased service like that, which integrates all of the software,
hardware, the subject matter knowledge, what type of insurance
does this person have, what drug are you ordering, is that
something they are allergic to, or is it consistent with other
drugs they are already on. We are responsible for what software
needs to be written, what hardware it needs to sit on, what
knowledge, what resources, insurance or laboratories that are
available for the doctor to interact with. We aim to formalise
and disaggregate the supply chain so that there is essentially a
reliable grid through which orders and results are trafficked. To
make a fiduciary platform, a national backbone that makes sure
those things are cleared properly. We believe that doing this will
create a lot of potential.
To begin with, it will create a lot of savings for everybody. Right
now we do everything twice. We send out every lab twice, we
send out every claim two and a half times, we fill out every
prescription 1.8 times. It is all manual and it is all coming out
on paper charts in the offices of teeny tiny doctors who don’t
have any really strong need to integrate. When a doctor orders
something, he or she expects a result to come back, and 30
per cent of the time it doesn’t. That affects patient care pretty
dramatically. In the US we killed 96,000 people in the last year
by accident by misreading or mis-transcribing information. If the
airline industry had the same fatal error rate as the healthcare
industry, there would be two jumbo jet airliners crashing into
each other every third day of the year killing everyone aboard
both planes – by accident!
Once we’ve established an information infrastructure, it then
becomes possible for doctors to decide their strategic direction.
Maybe they want to be their own lab, maybe they don’t, they
can figure it out. Once doctors become strategic entities, once
all of the complexity is stripped away and it’s reliable, then they
can make those decisions; and so that’s our goal. To get all of
this tough work to go away and then you’re actually looking
at information and you make informed decisions and we’re
thinking when that happens, doctors will integrate, will merge,
will see the opportunities of doing their own surgeries in their
centres, and the supply chain will dramatically reaggregate. It
will go from the general store model of a hospital that has a
total mishmash of whatever it is that the doctor wants done with
no particular idea of whether it works for them or not – to places
100
Chapter 10: Vertical integration and transaction cost
that take different parts of the supply chain and do them really
well in competition with others.
Everybody in this country has always been interested in whether
we can cure this or that disease, a new stint, or new device, or
a new operation but the truth is we can already do some pretty
absurdly elaborate things, we just can’t reliably give each other
flu shots. It turns out that the number of lives lost, angst caused,
humiliation caused by just doing the things that we already
know how to do badly vastly outweighs the lives saved by
learning how to do more things. And so there’s this gaping hole.
We are going medical group by medical group to grow our
business. Remember when you take on the supply chain, the
whole supply chain can be included, but if we don’t have this
or that insurance company or lab wired up, then we have to
replicate the manual work that the lady in the smock used to do
in the back office with the fax machine. At athena, we receive
two metric tons of paper every week which is the stuff that
insurance companies don’t give us electronically, and we run it
through scanners to verify the amounts and we run it through
the data centre in India and they punch in all the little numbers
on all the little pages so that they turn it into the electronic
information that we should have gotten in the first place. We
have to replicate the experience of everybody operating in a
paperless world by actually just handling all the paper. We go to
an insurance company and say, dude, you give us 65,000 pages
of paper a month! You have to print it, you have to mail it. If you
just build an interface, you don’t have to do anything!
You got to get there somehow. Imagine trying to start FedEx,
when you got a million planes flying every day, the incremental
cost of another package is very low…but how do you get to the
point where you’ve got a million planes flying? We at athena
have been lucky in that with the Internet and India and the
evolution of optical character recognition – all those things have
converged to allow us to burrow into these very cumbersome
pieces of the supply chain without enjoying the luxury of having
everybody on the same standard. It has been a kind of solar
eclipse that has made us lean enough, and the industry as we
walk into it is fat enough, that we can wedge ourselves into the
supply chain and still make money.
Source
This extended activity is based on an in-depth interview with Jonathan
Bush conducted by Brooke Russell.
Questions
1. Mr Bush talks about how the lack of scale and efficiency in medical
practices, the first link in the supply chain, has been a major factor in
creating the healthcare supply chain as we know it today. He argues
that, with access to information, doctors will become strategic entities
that see the benefits of merging and integrating. Analyse this projection.
What are some of the likely effects such consolidation would have?
Where might the projected sources of economies be? What might some
impediments be? Why do you think this has not happened before?
101
MN3119 Strategy
2. Mr Bush speaks about the ‘guild’ culture among medical practitioners.
Assuming the restraints on supply of physicians remain in place, what
besides empowering through information, could athenahealth be
doing to foster a business culture among practitioners in this unique
profession?
3. Mr Bush uses the Visa network as analogy of what athenahealth is
trying to iniate in the healthcare industry. Do you think this analogy is
apt? What are some of the differences between the two systems that
might lead to different outcomes?
4. athenahealth is focused on medical practices as its primary customer.
Analyse this strategy in light of the company’s vision of building an
information infrastructure that includes all players on the supply
chain. If you were Todd Park, responsible for business development
at athenahealth, what strategies would you propose to fasttrack this
process?
5. What other businesses started or could be created as a result of
problems or inefficiencies in the supply chain of this or another
industry? Brainstorm to find at least one example and discuss.
102
Chapter 11: Collusion
Chapter 11: Collusion
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
explain the factors determining the stability of cartels and other
cooperative agreements
•
discriminate between and explain different motivations for cheating in
cooperative agreements.
Essential reading
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) Chapter 8.
Further reading
Porter, R. ‘A study of cartel stability: the Joint Executive Committee, 1880–1886’,
Bell Journal of Economics 14(2) 1983, pp.301–14.
11.1 Introduction
Apart from choosing which strategic variables to compete in, firms can
also choose not to compete. By behaving to maximise joint profits rather
than in a purely self-interested fashion, firms might create a bigger ‘pie’
that they can distribute among themselves and each end up with higher
profits than if they were to compete vigorously. So far, so good. From basic
microeconomics you know that a monopolist will try to extract as much
surplus from consumers as possible, and increase profits accordingly – this
is also the reason why antitrust authorities will try to stop collusion, as
consumers are commonly assumed to suffer from a collusive cartel. This
chapter will present a very simple model of the stability of cooperative
agreements and outline some of the factors that will determine the
stability of such arrangements – note that we will need this model of
cooperation stability in later chapters, since the basic set-up is quite
generic. We then discuss some of the factors that will affect the likelihood
that cooperation is maintained.
11.2 Stability of cooperation
Suppose that two firms compete in a price-setting industry. As we know
from Chapter 3, equilibrium prices are at marginal cost, and in equilibrium
firms make no profits. Clearly, since the two firms are the only ones in
the industry, they feel that they can do better than this. One obvious way
of ‘doing better’ would be to agree to set monopoly prices and split the
market – as equal prices mean that the players split up the market. If both
firms agree to set monopoly prices, therefore, each of them makes half the
monopoly profits – quite a step up from zero profits. Suppose now that
one firm expects the other one to stick to the agreement. In that case, it
has an incentive to undercut the cooperative firm by a small amount, say
e, and take the entire market. (Remember that in Bertrand competition
whichever firm sets the lowest price takes the entire market.) Taking this
further, this means that neither firm would stick to the initial agreement –
both will want to undercut, and the situation will resemble a price war.
103
MN3119 Strategy
This intuition, however, only holds in a one-shot game. If firms play this
game repeatedly, being nice to each other may hold significant pay-offs
in the future. Consider this somewhat stylised situation: firms set prices
at the monopoly level and share the market at the start. They continue to
do so until the other firm has set any price other than the monopoly price,
and following that they set competitive prices forever, i.e. P=MC.
Activity 11.1
Can you show that this is a sub-game – perfect Nash equilibrium?
Guidance on this activity can be found in the VLE.
This gives us the following pay-offs:
A
B
Both cooperate:
pM/2
pM/2
A cooperates, B cheats
0
pM
A cheats, B cooperates
pM
0
Both compete
pC = 0
pC = 0
Activity 11.2
Draw this as a one-shot game.
Guidance on this activity can be found in the VLE.
Now if firms cooperate forever, they will make half the monopoly profits
forever. Recalling the discount factor, we know that tomorrow’s payoffs are worth d (<1) times today’s pay-offs, pay-offs the following day
are worth d2 today’s, etc. As this is a geometric progression, this can be
simplified to:
π
π t + δπt + δ 2π t + +δ 3π t ... = t
1− δ
You should look up a basic maths textbook to confirm this. What will
happen if a firm cheats? Given our strategies, cheating works once in
the sense that the cheating firm will take all the market, but it will suffer
in the future as a consequence, as from then on firms will compete a la
Bertrand and make zero profits. Cooperation is profitable if the profits
from cooperating are higher than the profits from cheating. This gives
πM
> π M , which is shorthand for 1  π M  > π M + δ π C .
2(1 − δ )
(1 − δ )  2 
(1 − δ )
Solving this for d, the discount factor, gives the following condition:
1
δ >δ =
2
d denotes the critical value of the discount factor such that both sides are
exactly equal. As such, it gives a critical value for the time value of money
above which cooperation is sustainable and below which it is not. What
does this mean? If the future is sufficiently important (i.e. d is close to 1),
the two firms will want to cooperate, while if it is not, they will prefer to
cheat and cooperation breaks down. The intuition for this is quite clear:
cheating gives a one-time pay-off since
π
πM > M .
2
104
Chapter 11: Collusion
On the other hand, the cheating party will expect cooperation to break
down from then on, which results in lower profits than under cooperation
– basically cheating means sacrificing long-term profits for short-term
gains.
11.3 Extending the model
The model we considered above was very stylised: Two players, Bertrand
competition, eternal cooperation or competition, symmetric firms, etc. We
now extend the model in a number of directions – we allow for different
types of competition, we increase the number of firms, and we discuss the
expected changes if punishment strategies would not be quite so severe.
11.3.1 Different competition models
Our assumption of Bertrand competition made life easy for us since
competitive profits were zero. Cournot competition, or any other type
of competition, is not as nice analytically, but can be solved in the same
manner. Basically, think of pM as monopoly profits, pC as non-cooperative
profits (whichever competitive model you assume in your specific setting),
and pCH as the profit arising from playing the best response to your rival’s
cooperative strategy. In matrix form, we can structure any situation in
which cooperation would be profitable if both kept to it, but there is an
individual incentive to deviate (i.e. a prisoner’s dilemma) as follows:
Figure 11.1: A prisoner’s dilemma.
The parameters have to be in the following order to constitute a prisoner’s
dilemma:
pChC > pCC > pChCh > pCCh
In words, cheating is better than cooperating if the rival cooperates (and if
the rival does not cooperate), but it is better for both players to cooperate
than for both to cheat.
Activity 11.3
Suppose that two firms with undifferentiated goods competitive in an industry with
demand function Q = 5 – P and Marginal Cost MC = 2. Find the threshold value of d, the
discount factor, for collusion to be stable if the mode of competition in each round (the
stage game) is as follows:
a. quantity (i.e. Cournot) competition
b. price (i.e. Bertrand) competition.
Guidance on this activity can be found in the VLE.
105
MN3119 Strategy
11.3.2 More players
What happens if we extend our model to allow for n > 2 firms? Although
this would also be possible with Cournot competition, we will use Bertrand
competition to keep things simple. Our cooperation stability condition
πM
changes to
> π M , since in an industry with n identical firms each
n(1 − δ )
firm will get 1/n of overall monopoly profits. Rearranging this and solving
n −1
for d, we find δ > δ =
.
n
This expression for d is increasing in n, which implies that the discount
factor must be higher for cooperation among more firms to be stable. In
the extreme, if we tried to achieve cooperation between n → ∞, all firms
would need to have a discount factor of almost 1, that is, all firms would
have to value tomorrow’s pay-offs almost as highly as today’s! Therefore
the stability of cooperation will decrease with the number of firms
included in the cartel.
11.3.3 ‘Forgiving’ punishments
We assumed that once cooperation breaks down, it can never be
resurrected. This is clearly an extreme assumption and not necessarily
a realistic one. If punishment were not eternal but only went on for a
limited number of periods, we would expect cheating to become more
likely! After all, a cheater does not suffer as much if, say, the other firm(s)
only compete non-cooperatively for two periods and then revert to the
cooperative agreement. To see how this changes things, consider the
following inequality:
3
πM
π
> π M + δπ C + δ 2π C + M δ
2 (1 − δ )
2 (1 − δ )
Cooperative profits on the left-hand side are unchanged, but cheating
profits are high in the cheating period (as before), zero in the following
two periods, and they revert back to cooperation profits in the third period
after having cheated. We can rewrite the left hand side as
and subtract the last term on both sides. After some algebra, we find the
new stability condition to be
δ >δ =
5 1
− ≈ .618
2 2
With eternal punishment the minimum value of d was 0.5, so that
cooperation has now become less likely (or more difficult to sustain).
Of course, if the stability condition is not met at one point in time, this
would mean that a cheater who (in our simplest model) found it profitable
to cheat one period will want to cheat again two periods from now when
the rivals revert back to collusion. So, if punishment is assumed to be of
limited length, cooperation may break down completely! This implies that
it is also in the best interest of firms to credibly commit to a draconian
punishment regime as this makes cooperation more likely – in other
words, the range of discount factors that will support cooperation is larger
for more draconian punishments.
106
Chapter 11: Collusion
11.4 Key concepts
•
Stability of cooperation
•
Cooperation
•
Finite v single games
•
Price wars.
11.5 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
explain the factors determining the stability of cartels and other
cooperative agreements
•
discriminate between and explain different motivations for cheating in
cooperative agreements.
11.6 Sample examination questions
1. Why would firms in an industry have periods in which they compete
very hard and periods in which they collude? Why does competition
not break down once and for all?
2. Which factors will influence the discount factor in an industry? Would
you expect collusion to be more likely in a fast-growing industry or in
a stable industry?
3. We often see collusion breaking down in the declining stages of an
industry lifecycle. Why is that?
4. Would asymmetry between firms make collusion more or less likely?
Explain by identifying the incentives to cheat for small and large firms
separately.
Guidance on answering these sample examination questions can be found
in the VLE.
Extended activity: the DeBeers diamond cartel
Read the following and answer the questions at the end.
Reading
Kretschmer, T. and L. Cabral ‘De Beers and beyond: the history of the international
diamond cartel’. LBS case study.
http://pages.stern.nyu.edu/~lcabral/teaching/debeers3.pdf
De Beers and beyond: an update
The need for change
[The radical and far reaching changes represent] De Beers’
response to the challenges of a competitive environment and a
change from a supply to a demand-driven business.1 –
Nicky Oppenheimer.
1
Chairman’s Statement,
De Beers Annual Report
2000.
De Beers’ traditional role in the diamond market has been to take on the
position of the custodian of the entire industry, protecting producers,
dealers and cutters through its cartel from the vices of free markets. De
Beers’ self-implied obligation to the stakeholders has been to shield the
industry from volatility in prices and demand, and to the consumers
to secure diamonds as a prestigious and exclusive luxury item. In the
107
MN3119 Strategy
last century De Beers managed to do so, by safeguarding its cartel from
competitors, by buying off excess supply, storing it in vast stockpiles to
protect the prices, and by launching advertisement campaigns on behalf of
the entire industry (e.g. A Diamond is Forever).
In the late 1990s, however, De Beers realised that the diamond market
was getting increasingly competitive, especially with the opening of
new diamond mines in Canada, resulting in a decline in market share of
DeBeers from nearly 80 per cent in the old days to 65 per cent in 1999
and a general underperformance of its stock.2 The diamond market at that
time was characterised by flat demand and excess supply, which caused
De Beers to increase its diamond stockpiles from US$2.5 billion in 1990 to
US$5 billion in 1998. It became obvious that in an industry where De Beers
was no longer the sole supplier of diamonds, the traditional custodian
role that De Beers had adopted over the past, could not be sustained.
Total percentage of
diamonds distributed
through the De Beers’
Diamond Trading
Company – De Beers
share of worldwide
production was 44 per
cent in 2002.
2
Formal strategic review
De Beers decided that it had to change its conventional strategy of an
all-embracing stronghold of the diamond industry to one of a highly
aggressive competitor. With the announcement of its new chairman,
Nicky Oppenheimer, and its new managing director, Gary Ralfe, De Beers
seriously considered changing its overall strategy and decided to launch
a formal strategic review of the industry, its operations and their future
outlook in an increasingly competitive market. With the help of the
Strategic Review Report, De Beers identified three main areas that had to
be transformed, in order to assure De Beers’ success in the years to come:
•
First, current operations had to be optimised. This included increasing
efficiency and cutting costs by 15 per cent in its operations (As is
plus strategy), as well as the restructuring of the organisation into
transparent, customer-orientated business units.
•
Second, De Beers realised the protection of its market leader position
demanded long-term measures. However, through it being a publicly
listed company on the JSE and LSE, De Beers felt that it had to endure
the quick-fix profit demand of its shareholders, instead of focusing
on necessary long-term projects. It thus decided to go private in 2001
after its stock had been listed for over 100 years.
•
Lastly, De Beers claimed there was unrealised potential in decreasing
advertisement costs and increasing customer demand by promoting
a more competitive environment at the downstream portion of the
industry pipeline while binding them closer to their own upstream
business (Supplier of Choice strategy).
The implementation of the first measure, to decrease costs and to restructure
the company, was a fairly standard procedure in the light of the new
competitive nature of the global economy. The other two aims, however,
express a fundamental change within De Beers as a corporation and the
diamond industry as a whole, intended to enable ‘De Beers [to enter]
the new century, fitter, leaner and streamlined for success’.3
Going private
De Beers realised that a company such as itself could not solely be
influenced by the short-termism of the stock market. Throughout history,
De Beers had to decrease its profits to balance the market, to prevent
excess supply and to sanction competitors. Trying to sustain the market
position of the stronghold of the industry, De Beers recognised that it
needed to maximise profits, not on an exclusive year-to-year basis, but
108
3
Chairman’s Statement,
De Beers Annual Report
1999.
Chapter 11: Collusion
rather on a long-term view. The only way to implement this was to go
private, contradictory to the general trend in the global economy.
Preceding the organisational restructuring of the De Beers Group
instigated in 1999, De Beers separated its management ties from Anglo
American Corporation in 1998, initiating an exclusive refocus on its
diamond products. In 1999 De Beers acquired the minority interests in
the DTC from Anglo American Corporation, sold most of its non-diamond
businesses and two years later announced its proposal to acquire all of the
outstanding shares of its shareholders.4
Within less than half a year, 93 per cent of the shareholders accepted the
offer by De Beers Investments, comprised of the Oppenheimer Family
through Central Holdings Limited (45 per cent), Anglo American plc (45
per cent) and Debswana, the joint-venture between the Government of
Botswana and De Beers (10 per cent).5 In June 2001 the trading of De
Beers’ shares on the JSE and LSE was thus terminated and De Beers went
private.
Supplier of choice
The rough diamond market is becoming a competitive one and
individual producers have to build up a loyal supportive base
if they want to be able to sell their output under less buoyant
market conditions.6
De Beers’ Supplier of Choice (SoC) marketing strategy, which was finally
cleared by the European Commission in 2002, is the attempt to closely
bind downstream customers to the upstream business and to foster
advertising and branding efforts of the downstream firms, thus growing
and encouraging consumer demand and total sales.
As part of the SoC strategy, De Beers (re)-evaluates current and potential
sightholders through a set of rigid criteria, such as financial standing,
market position, distributional and marketing strengths, etc., inviting
the most promising of them to benefit from the formalised and closer
relationship with De Beers. De Beers guarantees a steady supply of
diamonds and the support of marketing activities of its sightholders
through its Added Value Services. Of its 120 sightholders, many of them
with long-established relationships with De Beers, 25 per cent were
disposed of at the end of 2003, creating significant uncertainty in the
diamond industry.7
Because of the criteria applied by De Beers, which nobody really
seems to understand, [the Antwerp manufacturing] companies
have seen their viability come under real threat.8
Closure of the opportunity gap
The primary goal of SoC is to initiate growth in the diamond market
through downstream advertisement campaigns, driving the consumer
demand while allowing De Beers to transfer its advertisement efforts.
De Beers aimed to instigate a market where retail firms advertise and
establish their own brands, creating more effective, customer-near
marketing campaigns. Ultimately, De Beers is seeking to increase total
diamond jewellery sales by 4 per cent per year, compounding to a total
growth of 50 per cent over 10 years.
4
Previously, Anglo
owned 32.2 per cent of
De Beers, while De Beers
owned 35 per cent of
Anglo.
See Appendix 1:
De Beers Ownership
Structure.
5
Chasm Even-Zohor,
Tacy Diamond
Intelligence, 01.09.2000.
6
See Appendix 2: Official
DTC Sightholder List
2004.
7
Peter Meeus, HRD’s
Managing Director,
Antwerp Focus,
September 2003.
8
In the 1990s the diamond market was not growing proportional to
worldwide GDPs; compared with sales of similar luxury goods, it was
underperforming. The growing divergence between the rate of increase of
luxury goods sales and the rather flat rate of increase of the diamond sales
is depicted as the Opportunity Gap below:
109
MN3119 Strategy
500
CAGR 1980=100
400
Luxury Goods
300
GDP
200
100
Opportunity
Gap
Diamonds
80 82 84 86 88 90 92 94 96 98 2000
Figure 11.1: Supplier of Choice, Opportunity Gap (the difference between overall
luxury goods sales and diamond sales).
Source: www.debeersgroup.com
Comparing the amount of money that was spent on advertisement in
similar industries, the diamond industry was spending only 1 to 2 per
cent of total sales on advertisement, whereas the watch industry was
spending 6 per cent and the general luxury industry was spending even
up to 10 per cent of total sales. De Beers itself spent the industry’s lion
share, allocating about 4 per cent of total sales to advertising. It appeared
that the downstream diamond jewellery market and even other upstream
competitors were mainly relying (free-riding) on the industry-conducive
marketing campaigns launched by De Beers (A Diamond is forever).
Determined to end this inbalance, De Beers decided to introduce a new
marketing strategy Supplier of Choice, choosing only the best retailers as
its sightholders and training them in marketing and advertising.
At the official launch of its Supplier of Choice strategy in July 2000,
De Beers renamed its Central Selling Organization as Diamond Trading
Company (DTC) and introduced the DTC’s new logo the Forevermark,
which is inscribed into its stones to ensure its quality. In addition to the
Supplier of Choice, De Beers also introduced the Miner of Choice and
the Employer of Choice schemes, promoting company culture, equal
employment opportunities and such.
Vertical integration
Many diamond-producing companies have started to build vertically
integrated structures, a trend that will strengthen within the next
three to five years when we will see quite a different structure of
the world diamond market.9
The Supplier of Choice strategy would undoubtedly result in a highly
competitive, growth-orientated retailer market, as the removal of De Beers’
industry-wide cloak of protection would force retailers to establish their
own differentiated brands. As other diamond suppliers, such as BHP-Billiton
and Alrosa, had announced before, De Beers saw a chance to gain from the
downstream branding development. It decided to collaborate with the luxury
producer LVMH to set up the joint venture Rapids World Ltd. The resulting
new retail stores, named De Beers LV, target the very high-end element of
the jewellery market, attempting to obtain a 25 to 30 per cent premium over
unbranded jewellery. Its first store opened in London at the end of 2002,
three more in Tokyo in September 2003. The US launch will take place at the
end of 2004 with locations in both New York and Los Angeles.
Very recently, Aber Diamond Corporation announced that it would acquire
the luxury jeweller Harry Winston Inc., following the trend of vertical
110
Alrosa Vice President
Serguei Oulin, Interview
with Rapaport Diamond
Report, 09.04.2003.
9
Chapter 11: Collusion
integration set out by De Beers’ partnership with LVMH. Aber, however, felt
that it did not have the name recognition and thus was bound to link up
with a well-know brand, such as Harry Winston. Matthew Manson, Aber’s
vice president of marketing, explained to the New York Times:
We wanted to link up with a well-known brand. The very best brands
like Tiffany and Cartier are already out there. You can’t create those
overnight. We want to be a more vertically integrated company, not
just a producer and a seller. We are looking to attract investors.10
Antwerp Facets News
Service, 02.12.2003
10
De Beers had the advantage that it already had established its own name,
and thus could rely on sufficient consumer confidence to make De Beers
LV a successful brand name. Interestingly, the cartel is effectively profiting
from an identical strategy to Intel’s Intel Inside campaign. In each case the
companies had already established sufficient brand awareness with the
consumer through its upstream activities. Both used their existing reputation
to expand downstream and thus to benefit from its own name on various
levels of the diamond pipeline. This of course, is intended not simply as a
means of obtaining a greater overall return but should be seen as a stark
diversification within De Beers’ usual operations. De Beers further expects
that the partnership with LVMH will allow the cartel to access the US retail
market, a market so far closed to them due to US antitrust regulations.
It should be noted that Rapids World is set up as a completely independent
company from the rest of De Beers’ operations. De Beers has repeatedly
asserted that it would not directly supply Rapids World with stones.
However, this would mean that Rapids will have to buy polished stones
from the market at regular prices, while De Beers would sell them
beforehand to polishers. Cynics argue that this is going to lead inevitably
to a conflict of interest. De Beers needs to make certain that the new
Rapids World is lucrative, as it would otherwise severely harm De Beers’
reputation as a whole. Thus, one has to explore what options, besides
providing cheaper diamonds, De Beers has to ensure the success of its joint
venture over other downstream competitors.11
Developments in operations and exploration
Canada
Following the increase in competition through the split of Argyle from the
cartel, De Beers faced a new challenge to its market power. So far three
diamond deposits have been discovered in Canada: Snap Lake, Diavik and
Etaki. In 2000, De Beers was able to acquire Winspear and is expected to
start mining at Snap Lake by 2005, expecting to produce 1.5 million carats
per year over 22 years. De Beer’s competitors Rio Tinto and Aber Diamond
mines secured ownership of the Diavik Mines, producing 6 million carats
per year over 20 years. Another mine, Etaki, is owned by Dia Met along
with BHP-Billiton. De Beers has signed a contract to distribute 35 per cent
of Diaviks diamonds from 2000 to 2003. It is expected that Canada will be
the third largest diamond producer, worth 17 per cent of the total market.
To De Beers’ misfortune, it seems that it is having increasing difficulties in
establishing itself in this new market.
See ‘De Beers and
beyond: an update
(1997-2003) or
‘Conjecturing about
“incredibly hidden”
strategies behind
dropping of so many
US sightholders’ and
‘DTC strategies: a view
from the top’ on www.
diamondintelligence.
com, for a thorough and
extremely interesting
discussion of De Beers’
potential strategic
motives behind the
Supplier of Choice
implementation.
09.04.2003.
11
Botswana and Namibia
De Beers has continued to strengthen ties with Botswana and Namibia,
through the joint ventures Debswana and Nambed, held equally between
De Beers and the respective governments of both countries. A new mine in
each of the countries was opened in 2002; Debswana will have increased
its production to 30 million carats per year by the end of 2003 and
111
MN3119 Strategy
Namdeb is aspiring to double its value from 2000 by 2005 (as set out in
‘Big Hairy Audacious Goal’12 announced in 2001).
Congo and Angola
Problem areas such as Angola and the Congo, still torn apart by the
consequences of civil war, pose another challenge to the cartel. Because
of past civil unrests in Angola, De Beers had cut all operations there. At
the moment it is engaged in discussions aiming to organise reconstruction
and development in Angola. De Beers plans to return to the Democratic
Republic of Congo, as it has regained some stability. The United Nations,
however, commenced an investigation against De Beers for allegedly
exploiting the natural resources of Congo during the civil war.
12
The goal of growing
the value of the
diamond business
owned by De Beers to
US$10 billion by the
year 2004 was indeed
titled BHAG. The reasons
for this curious naming
are unknown.
Russia
Russia has, since the fall of communism, been a concern for De Beers.
Only part of Russia’s diamonds are traded through the DTC, while the
rest are sold directly to cutters. De Beers had traditionally bought its
Russian diamonds through the state-owned supplier Alrosa; however, at
the end of 2001 the official contract between Alrosa and De Beers ended.
The European Commission has since been reviewing the renewal of the
contract, an agreement to sell up to US$4 billion worth of rough diamonds
over five years, as it fears that ‘by entering into the agreement, De Beers
has abused its dominant position in the rough diamonds market’.13 At the
moment, De Beers and Alrosa are trading diamonds on a willing buyer/
willing seller basis, selling US$634 in 2003 – 25 per cent lower than the
laid-out contract would provide for. The question is: does Russia gain from
a fixed contract with De Beers or is it better off on a willing-buyer/willingseller basis allowing it to explore other opportunities simultaneously?
European Commission,
January 2003.
13
Challenges to the diamond market
Using De Beers’ forecasts, the value of diamonds will grow by 4 per cent
per year to US$90 billion in 10 years. A greater demand for diamonds,
through the prospect of an accessible US market and expanding
developing markets, notably India and China, is likely to result in a
shortage of available rough diamonds in the years to come. Industry
experts believe that the 50 per cent value increase in the diamond industry
will be achieved through vigorous price increases, instead of higher
diamonds sales.
Despite sluggish economies and uncertain stock markets there
is already a shortage of ongoing rough supplies which have only
been met from stock.14
James Picton for
UK stockbrokers W.H.
Ireland, ‘A review of the
world’s rough diamond
market’, www.minesite.
com, 18.07.2003.
14
In the light of this, De Beers’ shift in strategy, targeting a demand-driven
rather than a supply-driven diamond market, might have severe consequences
for the diamond trade. Industry experts predict that the combination of rising
prices and the novel costs of intensified advertisement and branding efforts
of the downstream market will lead to a complete re-organisation of the
industry. Many retail downstream firms will be forced to consolidate with
competitors or at least form joint ventures to avoid bankruptcy.
First, there is no guarantee that sufficient rough will be
available to support all of these initiatives. Second, the market
cannot support 130 different diamond brands. So we are likely
to see many companies bankrupted by high costs.15
At present, the retailer market has incurred an extremely high debt of
US$8.66 billion in 2002, from US$6.88 the year before. This was a result
112
Moshe Leviev, Leviev
International Diamond
Group, Gemological
Institute of America
Newsletter.
15
Chapter 11: Collusion
of the general de-stocking activities of the rough suppliers, specifically De
Beers, in the last few years. Much of the extra supply was taken up by the
extra demand in 2002. The anticipation of a steep rise in prices and the
low interest rates caused many downstream firms to buy off a great deal
of diamonds, stockpiling them for the future. In 2003 the prices of rough
diamonds increased by ca. 10 per cent, while the price of polished stones
continued diluting. Retailers are speculating that the inflationary return
on their inventory will exceed the cost of holding the diamond stockpiles.
However, the transfer of non-debt-financed stockpiles of De Beers to highly
debt-financed stockpiles of the retailers is clearly increasing the risk of
firms in the downstream market and promoting the expected consolidation
trend.
These factors will lead inevitably to a shakedown in the market, leaving
only the strongest, or better, the ones with the best strategy in the market.
Experts predict a major and increasing trend to vertical integration of
companies, similar to oil firms, where a small number of firms control the
whole of the supply chain (industry pipeline) from mining to cutting to
retail.
Conclusion
It is certain that the diamond market has changed from a tightly-controlled
market, in the hands of the market-custodian De Beers, to a increasingly
competitive industry, directing De Beers to fundamentally reshape its
business proposition.
Once upon a time, the key to success in the diamond business
was skill…In this new era, skill is still important, but strategy is
more important.16
Excerpt from
discussion: ‘The past,
present and future’ at
Rapaport International
Diamond Conference
2003, 20.10.2003.
16
The question is whether the radical change of strategy will allow De Beers
to hold on to its market share, and whether the Supplier of Choice strategy
can defend the De Beers Group from further challenges in the future. Also
is there another underlying plan hidden in the Supplier of Choice strategy
or is De Beers only optimising its distribution system?
Diamonds are forever, but is the De Beers cartel?
Source
Reichel, M.C. ‘De Beers and beyond: an update’. LSE case study.
113
MN3119 Strategy
Appendix 1 – De Beers ownership structure (abridged)17
(following January 2002 restructuring)
11%
Anglo American
Group
Central Holdings
Group
Debswana Group
89%
Central Investments
DBI (Lux)
10%
45%
45%
100%
DB Investments (Lux)
(new company)
De Beers sa (formerly
DB Investments)
100%
De Beers Consolidated
Mines (DBCM)
100%
De Beers Centenray
(DBCAG))
Questions
1. Cabral (2000) gives an overview of factors facilitating or hindering
collusion. Which ones are relevant in the diamond industry? Explain
how they affect the parameters of the model he outlines on pp.128–30.
2. Why do you think the cartel grew increasingly unstable in recent
years? Is there a pattern in the type of attacks on the cartel, or in the
type of attacker? If there is, explain what happened on a broad level,
and if not, pick two cases and analyse the attackers’ motives for their
behaviour.
3. Analyse De Beers’ recent changes in their strategy. Will the ‘Supplier
of Choice’ strategy be sufficient to shield De Beers from changes in the
diamond industry? Which complementary measures (or alternatives)
would you use as a De Beers executive?
114
Group ownership
structure, www.
debeersgroup.com
17
Chapter 12: Strategic partnerships
Chapter 12: Strategic partnerships
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
discuss types of capabilities and motivation for investing or not
investing in them
•
evaluate the benefits and pitfalls of using strategic alliances.
Essential reading
Dierickx, I. and K. Cool ‘Asset stock accumulation and sustainability of competitive
advantage’, Management Science 35(12) 1989, pp.1504–511.
Dyer, J. and H. Singh ‘The relational view: cooperative strategy and sources of
interorganisational competitive advantage’, Academy of Management Review
23(4) 1998, pp.660–79.
Further reading
Arora, Alfonso Gambardella ‘Complementarities and external linkages: the
strategies of large firms in biotechnology’, Journal of Industrial Economics 38(4)
1990, pp.361–79.
Kretschmer, T. and P. Puranam ‘Integration through incentives within differentiated
organisations’, Organization Science 19(6) 2008, pp.860–75.
12.1 Introduction
Strategic alliances combine many of the pros and cons of cooperation
between firms as discussed in the previous chapter. Both parties benefit
if they collaborate or behave in the manner that they had previously
specified. However, each party has an incentive to hold back efforts or
resources in the hope that the other one will deliver or there might be
policies which restrict collusions but which give firms incentives to enter
into alliances; a classic prisoner’s dilemma. This chapter will not dwell on
detail regarding the motives of firms to enter strategic alliances, but will
look at the mechanics of how such alliances work and at the trade-offs in
this context. We will focus on one kind of strategic alliance, namely an
alliance to build up capabilities. This type of strategic alliance is a hybrid
solution between internal development and external sourcing.
12.2 Building capabilities
There are two distinct types of strategic alliances: result-driven and
capability-building. Result-driven strategic alliances come about because
a specific project, or joint output, needs to be completed, a result needs
to be generated, etc. Imagine it as a group of students getting together
to prepare a class presentation or a group assignment. This has all the
features of a strategic alliance where there are benefits to working hard,
but also incentives to let others do the work. Capability-building strategic
alliances, on the other hand, may not generate a tangible outcome, but
they are aimed at improving the efficiency of the alliance partners in
general. Imagine a study group consisting of students taking the course
– studying together may not generate a tangible joint output, but it
115
MN3119 Strategy
will improve the ability of the students to take their exams (which are
individual, so we cannot talk of a joint output). Again each student would
like all the others to take charge of topics and be well prepared, but there
is still an incentive to listen to others’ explanations and save oneself some
work. In this chapter, we will focus on the second type of strategic alliance,
capability-building alliances.
Building firm capabilities is traditionally considered an ‘inside job’. After
all, firm capabilities are what set firms apart from their rivals, and therefore
maintaining as much control as possible is preferable (we discussed this
in Chapter 6). So how do firms build capabilities? Figure 12.1 gives an
overview of the methods of building capabilities.
Figure 12.1: Building capabilities.
12.3 Business and strategic partnerships1
One way of building capabilities are strategic partnerships. First, we
distinguish between strategic partnerships and business partnerships – not
all collaborations, even long-term ones, between firms qualify as strategic
partnerships.
A strategic partnership can typically be identified by the following.
Firms will usually make relation-specific investments, that is,
investments that carry most use within a particular partnership. If an
investment is generic and not specific to the partnership, it is easy for each
party to walk away without losing much money.
Strategic partnerships often rely heavily on the voluntary exchange
of information among the parties. In a business partnership, all the
information that has to be exchanged will either be exchanged up front or
it will be written in a formal contract. Strategic partnerships on the other
hand will be governed by relatively incomplete contracts (contracts that
do not specify what to do in each and every possible situation – but more
on that in Chapters 7 and 10), so that information exchange becomes
important as the relationship unfolds.
Linked to the previous point, strategic partnerships will be much more
integrated and interdependent than regular business partnerships.
This implies that firms will be interdependent both in their actions (so
that one firm’s actions will be more effective if they are coordinated with
the other firm’s) and in their pay-offs (so that one firm’s pay-offs from the
partnership will depend on the other firm’s pay-offs as well).
Finally, the way in which strategic partnerships and business partnerships
deal with transaction costs also differs. In a business partnership, the
goal is to minimise transaction costs by specifying as much as possible in a
contract before launching the partnership. On the other hand, a strategic
116
This section is based
on Dyer, J. and H.
Singh ‘The relational
view: cooperative
strategy and sources
of interorganisational
competitive advantage’,
The Academy of
Management Review
23(4) 1998 pp.660–79.
1
Chapter 12: Strategic partnerships
partnership will accept that contracts are incomplete and attempt to build
in mechanisms to cope with transaction costs – for example, by joint asset
ownership which aligns incentives of the different parties of a strategic
partnership.
We can therefore finally give a definition of strategic partnerships:
Strategic partnership – definition
Strategic partnerships are relationships between organisations that involve significant
amounts of shared decision making and that may involve equity ownership (economic
integration), organisational linkages and coordination mechanisms (organisational
integration).
Business partnerships score lower on most or all of these dimensions.
In particular, the details of investments is typically lower in business
partnerships, and business partnerships will often specify as many
terms and conditions as possible in the initial contract, thus minimising
transaction cost and formalising the exchange of information. In other
words, business partnerships are often drawn up as formal contractual
propositions, while strategic partnerships will rely on the willingness of the
parties or incentives to ‘make a relationship work’.
Equity ownership
(Economic integration)
X%
A
B
Y%
Coordination mechanisms
(Organisational integration)
Figure 12.2: Elements of a strategic partnership.
Figure 12.2 illustrates the two elements that typically define a strategic
partnership. First, strategic partnerships will require some form of
organisational integration – coordination between the two strategic
partners, either through communication and coordination efforts or even
by forming an entirely new entity (as illustrated in Figure 12.3 – a joint
venture). The other element typically found in a strategic partnership is a
degree of economic integration, that is joint ownership of assets or crossasset holding. This makes coordinating any actions easier simply because
incentives by both partners are aligned – if you do well, so do I (because I
have a stake in your own) and vice versa.
A particular form of strategic partnership is a joint venture, where a new
organisation is set up, and therefore organisational integration takes place
within this new unit. This type of strategic partnership is special in the
sense that a new entity is created in which the partnership ‘takes place’.
The firms can be entirely separate except for the newly-formed joint
venture. The advantage of this is that coordinating mechanisms can be
realised with less friction since the main activities of the respective partners
are not affected. On the other hand of course, the potential for generating
synergies originating from the main activities of the partners is lower.
117
MN3119 Strategy
Joint Venture
Equity ownership
(Economic integration)
50+x %
50- x %
A
B
Coordinating mechanisms
(Organisational integration)
Figure 12.3: Joint venture.
Activity 12.1
How would you organise the following and why?
a. Two firms, a hardware and a software firm, want to develop a revolutionary new
audio technology.
b. Two pharmaceutical firms want to enter a new medical sub-class and want to build
capabilities in this treatment area.
c. The municipality of a large city in an emerging economy has awarded a contract to
build a large leisure centre to a local entrepreneur, who will require input from a large
number of different parties.
Guidance on this activity can be found in the VLE.
It is important to note here that strategic partnerships might become
useful for two (not mutually exclusive) different reasons. The actions
of the strategic partners might be complementary or the actions might
require a certain degree of coordination to confer their full benefits.
12.4 Equity ownership
Having established that there is some need for organisational integration,
where does financial integration come in? In other words, why is equity
ownership important in such a situation? As in many cases, the purpose
of equity ownership is to enable decision making in the interest of the
partnership and not the individual party. Shared decision making is
typically difficult since contracts are incomplete (as mentioned before) and
there is considerable scope, particularly in open-ended partnerships, for
opportunistic behaviour by both parties. Equity ownership, either directly
or by setting up a new legal entity through a joint venture, will help align
incentives and minimise opportunistic behaviour.
As with most things, however, equity ownership does not come without
risks and costs, which we will discuss in more detail.
The main benefits of equity ownership are the alignment of
interests, the retention of control and exclusivity of the
partnership, and the feasibility of inter-organisational coordination.
There is less risk of one of the strategic partners being exploited by an
opportunistic player with no equity stake in the partnership, and it is more
likely that the partner will take actions that are inline with the overall
success of the undertaking.
118
Chapter 12: Strategic partnerships
The key costs are a loss of flexibility and a loss of motivation.
Committing to an equity stake in a project lowers the room for manoeuvre
if the technology or the market develops in unexpected ways. Locking into
a partnership with equity ownership is therefore more risky if there are
elements outside the relationship that cannot be taken into account. Also,
once equity ownership has been established, motivation to exert maximum
effort may decrease since the threat of not continuing a relationship is
much lower if equity is involved – if the relationship was governed by a
series of spot contracts, anything lower than the optimal outcome (or the
agreed level of effort) would lead to the contract not being renewed.
Equity ownership and organisational integration are interlinked in that
there is a relatively strong positive relationship between the two, as
illustrated by Figure 12.4. It is, however, important to note that there
can also be cases with high equity ownership but low organisational
integration such as pure financial investment or a set of independent
ownership
and inter-organisational
integration
business units Equity
under the
same umbrella
(e.g. a holding company).
inter-organisational integration
High
R&D JV’s
Collaborative R&D
Acquisitions in which target
firms are “absorbed” into
acquirer
Acquisitions in which target
firms enjoy intermediate levels
of organisational autonomy
Contract R&D
low
Licensing agreements
low
Minority Equity
Equity ownership level
Acquisitions in which target
firms are managed as autonomous units
High
Figure 12.4: Equity ownership and inter-organisational integration.
Two questions should be asked before engaging in a strategic partnership:
first, will a strategic partnership actually create value or should firms
extend their firm boundary? Will it actually be beneficial to the two firms
to join forces, and is it preferable to integrate to some extent rather than
just engage in cross-shareholding without a controlling stake? Is there
more value created by a strategic partnership than by a clever investor?
Assuming that this is so, the second question should be about the level of
organisational and financial integration. Considering the costs of equity
ownership, it is therefore advisable to limit ownership so as to sustain the
requisite level of shared decision making.
119
MN3119 Strategy
12.5 Key concepts
•
Result driven
•
Capability driven
•
Business and strategic partnership
•
Strategic partnership
•
Internal and external development
•
Asset stock accumulation
•
Stocks and flows
•
Equity ownership.
12.6 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
discuss types of capabilities and motivation for investing or not
investing to them
•
evaluate the benefits and pitfalls of using strategic alliances.
12.7 Sample examination questions
1. Business partnerships are relatively loose forms of cooperation among
firms. Is this form of cooperation particularly common in specific
stages of the industry life cycle and, if so, why?
2. Why would a firm voluntarily choose not to build up capabilities to
match those of the market leader?
3. Is equity ownership more useful and important for capability-building
alliances than for results-driven alliances?
Guidance on answering these Sample examination questions can be found
in the VLE.
Extended activity: the EU aviation industry
Read the following and answer the questions at the end.
Strategic alliances – case study: The EU aviation industry
The civil aviation industry is defined by a fundamental paradox: it is
one of the key instruments of globalisation and is truly international in
the scope of its operations; yet it is restricted by arcane regulations and
national politics – perhaps more than any other business in the world.
Historically, the industry took shape through the creation of national
airlines or ‘flag carriers’ by governments eager to propel the economic and
social development of their countries. This legacy has created a myriad of
vested state interests that result in significant barriers to exit or a political
inability to let failing flag carriers go bankrupt.
The economic realities of the airline industry are tough. Burdened with
high fixed costs and a cyclical problem of overcapacity, airlines also face
a declining rate of growth in air travel and stiff competition –particularly
from low-cost carriers. In recent years, the industry has seen a marked
decline in yields, the average revenue per passenger, due to the persistent
pressure on airlines to reduce costs and pass on the savings to customers
demanding lower prices in a fiercely competitive market.
120
Chapter 12: Strategic partnerships
European civil aviation has undergone significant liberalisation and
privatisation over the past 20 years through liberalisation ‘packages’ in
1987, 1990 and 1993. Today the member states of the European Union
act as one aviation market, with unlimited traffic rights and rights of
establishment, a liberal fare regime, and healthy competition.
Airline consolidation, widely believed to be the natural and beneficial
follow-on to liberalisation, has been thwarted to date, largely by the ‘open
skies’ bilateral treaties governing air traffic between EU states and the US.
By imposing ownership and landing restrictions stipulating that designated
airlines must be majority owned and controlled by a national of the given
state, these treaties place EU airlines at a disadvantage. European airlines
are deterred from making ‘cross-border’ acquisitions – even within Europe
– because the nationality clauses in the prevailing treaties put into question
whether the acquired company would have to forfeit its right to serve
foreign (particularly the US) markets. Unable to merge, these carriers cannot
fully exploit either the US or European markets. Thus over 20 European
carriers divide transatlantic traffic between them, each operating from their
individual home base. Many in the EU lament that the ‘open skies’ bilaterals
create market asymmetries that give the US an unfair advantage.
Unless and until the necessary regulatory changes open the way for
significant consolidation of the global market, airlines will continue to
forge the next best thing – alliances, which help them achieve some of the
economic gains that genuine consolidation would offer: greater economies
of scale and the seamless travel services across regions that large
multinational corporations – the most lucrative customers – demand.
Alliance
Members
Star Alliance Fourteen, including:
United, Lufthansa,
SAS, Air Canada, Thai
Airways, Varig
One World Eight: Aerlingus,
American, British
Airways, Cathay
Pacific, Finn Air, Iberia,
LanChile, Qantas
SkyTeam
Six: Air France, Alitalia,
Delta, CSA Czech
Airlines, AeroMexico,
Korean Air
Northwest Northwest –KLM
–KLM
Founded Countries Destinations
Annual
served
served
passengers
(m)
% share
of world
traffic
Anti-trust
immunity?
1997
124
729
289.3 in Year
1999
18.5
Yes for United,
Lufthansa, SAS
1999
135
571
238 in Year
2000
11.8
No
2000
120
500
228 in Year
2000
11.1
No
1993
N/A
N/A
72.9 in Year
1999
4.7
Yes for NWAKLM, and
NWA- Malaysia
Source: company websites and AEA 2001 Yearbook
Without antitrust immunity from the relevant regulatory bodies –
for example, the European Commission and the US Department of
Transportation – alliances can do little more than achieve cost savings
through joint marketing campaigns and possibly joint purchasing. Those
with immunity like Lufthansa and United can align fares and flight
schedules, offering customers the much-touted seamless travel experience
of one-stop check-in and easy flight connections. Combining operations in
this way gives alliance partners access to significant scale economies and
allows them to jointly tighten or expand capacity to optimal levels.
121
MN3119 Strategy
European airlines fall into two broad categories: the traditional flag carrier
and the low cost carriers. The latter have taken the industry by storm
following its liberalisation, and continue to gain market share with their
winning formula of fleet standardisation, quick turnaround on point–point
short-haul routes, and direct sales and marketing practices.
Lufthansa, British Airways (BA), and Air France are the three dominant
airlines in Europe today. Among these market leaders, however, one
sees divergent business strategies. Lufthansa exemplifies the ‘aviation’
or vertically integrated business model in which its passenger and cargo
airline businesses sit alongside five independently competitive business
groups including catering, ground handling, maintenance and overhaul,
IT services, and leisure travel. In contrast, BA pursues a ‘lean’ strategy,
focusing mainly on its airline services and outsourcing as many of the
related support services as possible. Still this strategy leaves BA competing
head-to-head with the low-cost carriers on short-haul routes, and so makes
continued cost-cutting a necessity.
Airlines that can reduce costs through efficient corporate structures and
external alliances will have an advantage in the coming years, where data
show that passenger yields, or the average revenue that airlines earn per
passenger kilometre, is declining. If airlines acted in concert to lower
capacity, they could boost passenger load factors and yields by keeping
fares robust. Acting alone, however, airlines that cut costs by cutting
capacity might never see concomitant gains in yields for they risk losing
significant market share to competitors who might seize the opportunity to
expand their own capacity.
Industry consolidation would indeed make it easier for airlines to correct
any capacity inequalities, and would also make it easier for airlines to
deal with the natural cycles in supply and demand, which are an integral
characteristic of the industry. But, as we have seen, true consolidation
will not happen in Europe until states remove the nationality clauses in
their air transport agreements – an issue that governments, not firms,
can resolve. For now and the foreseeable future, European airlines must
operate within the realities of the market, shaped as it is by regulations.
Source
Russell, Brooke under the supervision of Dr T. Kretschmer ‘The EU aviation
industry’, LSE case study.
Questions
1. Outline the costs and benefits of airline consolidation.
2. Compare the relative benefits of airlines consolidating versus forming
strategic alliances.
3. If you were CEO of an airline, and there were no nationality clauses,
how would you grow your business?
122
Chapter 13: Competitive dynamics
Chapter 13: Competitive dynamics
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
explain the core ideas of competitive dynamics by using the actoraction-response-performance framework
•
describe how the awareness, motivation and capability of a firm
influence its responses towards rivals’ actions.
Essential reading
Chen, M.-J. ‘Competitor analysis and interfirm rivalry: towards a theoretical
integration’, Academy of Management Review 21(1) 1996, pp.100–34.
Ferrier, W.J., K.G. Smith and C.M. Grimm ‘The role of competitive action in
market share erosion and industry dynamics: a study of industry leaders and
challengers’, Academy of Management Journal 42(4) 1999, pp.372–88.
Smith, K.G., W.J. Ferrier and H. Ndofor ‘Competitive dynamics research: critique
and future directions’ in M.A. Hitt, R.E. Freeman and J.S. Harrison (eds), The
Blackwell Handbook of Strategic Management. (Oxford: Blackwell Publishing,
2001) pp.314–61.
Further reading
Chen, M.-J. and D. Miller ‘Competitive dynamics: themes, trends, and a prospective
research platform’, The Academy of Management Annals 6(1) 2012, pp.135–210.
13.1 Introduction
Below are three major headlines from The Wall Street Journal:
•
‘Alex Trotman’s goal: To make Ford No. 1 in world auto sales.’
•
‘Kellogg’s cutting prices... to check loss of market share.’
•
‘Amoco scrambles to remain king of the polyester hill.’
As we can see from these quotes, at the heart of many a firm’s strategy
is competing against its rivals. This is done by carrying out specific
actions to become the market leader or to avoid becoming dethroned
as the market leader. As already mentioned in Chapter 6, firms seek to
gain a competitive advantage in order to consistently outperform other
players in the industry. Because firms compete for limited customers and
resources, one firm outperforming the other often comes at the cost of
the failure of its rival. To compete with each other and to attain market
leadership, firms can choose from a wide range of competitive actions like
introducing new products, cutting prices or engaging in R&D. Studying
which actions specific firms take when they compete with specific rivals
is at the core of competitive dynamics, which we discuss in this chapter.
Competitive dynamics studies the process of interfirm rivalry in which
firms jockey for competitive advantage by exchanging competitive actions.
We study competitive dynamics for two main reasons. On the one hand,
it complements the traditional perspective on strategy and the wellknown structure-conduct-performance paradigm and five forces analysis
in several important dimensions. First, instead of studying competition
and strategic decisions from a macro industry level, competitive dynamics
123
MN3119 Strategy
looks at an individual firm (or pair of firms) and studies specific actions.
Second, competitive dynamics integrates internal (e.g. resources and
capabilities) and external (e.g. rivals) concerns instead of just focusing
on the external ones. Third, contrary to the rather static perspective often
taken, competitive dynamics takes a dynamic perspective on competitive
behavioural patterns. Competition is seen as a dynamic market process
rather than a static market condition. The focus is on the process by which
a market moves away towards the equilibrium but not on the equilibrium
per se. On the other hand, unlike game theory models, competitive
dynamics is empirically driven. A well-known example of such a set of
empirical insights is a case study on the competitive interactions between
Kodak and Polaroid to become the leader of the camera industry shown in
Figure 13.1.
Polaroid
1976
Announces
Australian
partner for
movie camera
system
Files patent
suit against
Kodak (instant
camera)
Reports
consumer tests
“Kodak has
‘fading’
problem”
01/1976
04/1976
04/1976
11/1976
03/1976
Kodak
Introduces
tele-instamatic
camera
(new features)
Introduces
instant-picture
camera and
announces
attack strategy
12/1976
Announces
exchange of
patent licenses
with Xerox
Sets price for
motion picture
camera at
$699
03/1976
09/1977
09/1977
1978
Announces
new low-priced
instant camera
ane other
products
Announces
they’re adding
4 pictures to
20-picture roll
for 35mm film
...
Figure 13.1: Competitive interactions between Kodak and Polaroid.
13.2 A framework to analyse competitive dynamics
Most of the research on competitive dynamics can be integrated into
the framework shown in Figure 13.2. The framework is useful because
it can help us think about the process of competitive dynamics in a
structured way. The first element of the framework is the actors. Here, we
are interested in the characteristics of firms initiating actions and their
motivation to do so. Second, when studying actions, we look for answers
to the question regarding which actions firms use. The next building
block are the responses – when and how do rivals respond? And, finally,
research on competitive dynamics looks at the performance implications
of the actions and responses. In the next paragraphs we will discuss each
building block in more detail.
Actor
(focal firm)
Action
Response
(by competitors)
Performance
(competitive
advantage)
Figure 13.2: Competitive dynamics framework.
13.2.1 Actor
Firms ‘act’ by initiating competitive actions like price cuts or new product
introductions. The timing of the action is crucial for the performance of
124
...
Chapter 13: Competitive dynamics
an action. Should firms try to act first or wait and see whether something
works? Research shows that late followers usually have a disadvantage,
but it is not clear whether it is better to be a first mover or an early
follower. On the one hand, a first mover can enjoy monopoly rents from
an action (like an increase in sales after a price cut) for a short period
until rival firms follow. Second, being first helps the firm to build customer
loyalty and branding advantages. Third, early firms can build up market
share to benefit from economies of scale. On the other hand, early
followers can often still capture a large part of the market while incurring
fewer development costs and bearing less technical and market risk. In
addition, they have more time to improve the technology to avoid mistakes
that can destroy the reputation of the firm.
Besides timing, a second important question is who moves first: incumbents
who have a long history in the industry or new entrants? Entrants
have strong incentives to disrupt the status quo by introducing radical
innovations or undercutting prices as they have little to lose and a lot to
gain. Their only fear is that the large incumbent might retaliate and drive
them out of the market. For incumbents, on the other hand, incentives
are mixed. There are some advantages to proactively undertaking actions.
First, by engaging in some action, incumbents can use existing (slack)
resources to improve their market position. Second, by initiating an
action, incumbents can deter market entry or drive out small entrants.
Third, they can use actions to fight other incumbents. However, there
are also some downsides. First, the introduction of new products can
lead to the cannibalisation of the old products. Second, new products
might require new capabilities that the firm currently does not possess.
Moreover, developing new capabilities might cannibalise resources that
might otherwise be used for old products. While there is a lot of research
on the question of whether entrants or incumbents initiate more actions,
the question is still not fully answered. Incumbents usually face a trade-off
between continually engaging in new actions and cannibalising some of
their own profits or not engaging and getting some of their profits stolen
by rival firms. This idea goes back to the work of Joseph Schumpeter
who predicted that industry leaders who do not continually engage
in competitive actions would have their market position destroyed by
incumbents. Market leaders can only retain their market shares and are
‘dethroned’ less often if they initiate and undertake more actions more
quickly than challengers and if they have a broader range of actions.
13.2.2 Action
Competitive actions can be defined as ‘externally directed, specific and
observable competitive moves initiated by a firm to enhance its relative
competitive position’. Broadly speaking there are two types of actions:
strategic and tactical actions. Strategic actions are all actions that
have a long-term impact on the firm, require a significant commitment
of organisational and managerial resources, and are hard to reverse.
Examples of strategic actions include mergers and acquisitions, new
product development or the appointment of a new CEO. By contrast,
tactical actions involve fewer resources, are easier to reverse and have only
a short-term effect on the organisation. Typical examples of tactical actions
are price cuts or an advertising campaign in a local newspaper. Strategic
and tactical actions not only differ in their importance for the firm carrying
them out, they also have different impacts on competitors.
Besides carrying out individual actions, firms can also engage in action
sequences (i.e. combining different single actions like changes in quality
125
MN3119 Strategy
and price combined with an increase in the production capacity).
However, managers need to be careful when combining different actions.
Research has shown that the length and the volume of attack duration has
a positive impact on the performance of the focal firm, but more complex
attacks reduce the positive effect, as a complex series of actions is harder
to conceive and to implement. This means that complex actions are often
delayed.
13.2.3 Reactor
As already mentioned in the introduction, the competitive dynamics
literature does not assume that the relationship between every firm in an
industry has to be symmetric. Put differently, not every firm in an industry
is fighting with all other firms to the same extent and not every firm is
threatened by the action of a rival in the same way. Hence, not every firm
might react to an action in the same way and some firms might not even
react at all. The question then is which firms react and why do they react?
Two distinct features are said to have an influence on a firm to respond
to a rival’s action: resource similarity and market commonality. Resource
similarity is the degree to which two firms are similar in their resource
endowments in terms of both type and time. This idea is rooted in the
resource-based view of the firm and the idea of strategic similarity. The
notion of market commonality goes back to the literature on multimarket
contact. Firms engage in multimarket contact if they compete with a rival
in several markets at the same time. When that happens, they have to take
all markets into consideration when initiating an action in a focal market.
The reason is that an action in one market might also cause retaliation in
the other markets. As a result, multimarket competitors may hesitate to
attack in one market for the fear of retaliation in other markets. This effect
is called ‘mutual forbearance’ and there is some empirical evidence that
firms competing in several markets face less fierce competition in these
markets. Empirical studies in the telecommunication industry, for example,
show that multimarket competitors have higher margins than singlemarket rivals.
Firms are more likely to react to a rival’s competitive move if they have a
high market commonality, but also if competitors are similar in terms of
resources as they have the ability to retaliate.
13.2.4 Response
When a firm takes action to gain competitive advantage, rivals often
quickly respond so as not to fall behind. Responses can differ in terms of
their volume, speed and type. The way rival firms respond to a competitive
attack depends on three different factors: awareness, motivation and
capability. Awareness means that a firm must perceive and comprehend
an action of a rival. In addition, the firm must be motivated and given
incentives to react to a rival’s move. For example, if an opponent
introduces a new product in a market that is not important enough to
the focal firm, it might not be motivated enough to react as losses are
limited. The third criterion that must be met is that a firm must be capable
to react. An organisation might not possess the required resources or
know how to react, even if it is aware of the rival’s action and even if it is
motivated enough. Only if all three conditions are met together will a firm
react to the competitive actions of a rival.
The magnitude and speed of responses not only depend on the
characteristics of the respondent in terms of motivation, awareness and
126
Chapter 13: Competitive dynamics
capability but also on the kind of action (i.e. strategic or tactical action)
that triggered the response. Strategic actions tend to elicit fewer and
slower responses by rivals as they require more time and more resources
to be effective. On the other hand, tactical actions tend to elicit more and
faster responses.
13.2.5 Performance implications
Engaging in competitive dynamics can have severe implications for both
incumbents and challengers. Research has shown that market leaders can
only retain their strong position if they initiate more actions than their
challengers, if they use a broader range of actions and if they undertake
their actions quicker than rival firms. Maintaining market leadership
therefore appears to follow a ‘strategic substitutes’ logic – more actions by
an early mover leads to fewer actions and therefore lower performance by
rivals.
13.3 Key concepts
•
Competitive dynamics
•
Competitive action
•
Resource similarity
•
Market commonality
•
Awareness, motivation and capability.
13.4 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
explain the core ideas of competitive dynamics by using the simple
actor-action-response-performance framework
•
describe how the awareness motivation and capabilities of a firm
influences its responses towards rivals’ actions.
13.5 Sample examination questions
1. Give two examples each of a strategic and tactical action in the
alcoholic beverages industry.
2. Most experts will argue that responding quickly to a rival’s action leads
to higher performance. Explain why and give a counterexample to this
expectation.
Guidance on answering these Sample examination questions can be found
in the VLE.
127
MN3119 Strategy
Notes
128
Chapter 14: Entry and entry deterrence
Chapter 14: Entry and entry deterrence1
Learning outcome
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
define structural and strategic entry barriers and compare the two.
Essential reading
Swaminathan, A. ‘Entry into new market segments in mature industries:
endogenous and exogenous segmentation in the US brewing industry’, Strategic
Management Journal 19 1998, pp.389–404.
Further reading
This chapter is based
on a guest lecture given
by my former supervisor,
Paul Geroski, who
died much too young.
Hearing Paul talk about
entry, and indeed any
economic or managerial
phenomenon, makes it
difficult to think about
entry in any different
way or structure. My
intellectual debt to him
goes far beyond this
chapter, but I hope to
do him and his thoughts
justice with this attempt
to condense his
thoughts on entry.
1
Besanko, D., D. Dranove, M. Shanley and S. Shaefer Economics of Strategy. (New
Jersey: Wiley, 2009) Chapter 9.
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) Chapter 15.
Geroski, P. ‘Early warning of new rivals’, Sloan Management Review 40(3) 1999,
pp.107–16.
McAfee, P., H. Mialon and M. Williams ‘What is a barrier to entry?’, American
Economic Review Papers and Proceedings 94 2004, pp.461–65.
Saloner, G., A. Shepard and J. Podolny Strategic Management. (New Jersey: Wiley,
2006) Chapter 9.
14.1 Introduction
One of the most ‘strategic’ decisions in the existence of a firm is whether
to enter into a particular market or not. It is also obvious that the
(anticipated) reaction of firms currently in the market plays an important
role in the entry decision. Suppose that an industry currently yields very
high profits for the firms operating in it – for arguments sake, operating
profits in the pharmaceutical and biotech industry which run close to
20 per cent. Does this mean that a firm should enter it? Not necessarily.
First, an additional competitor may disturb a delicate balance achieved
by established players after interacting over a long period to reach these
profit margins (recall Chapter 11). This need not even be a problem for
antitrust authorities, it could simply mean that firms are specialised in
particular product classes and as such don’t compete very hard against
each other because they produce distant substitutes. A new firm taking
on one of these firms in their ‘home market’ would then encounter fierce
reprisals and witness a significant drop in these initially enticing profit
margins. Plus, it may not even be possible to enter these industries.
Pharmaceuticals are again a good example: even if an entrepreneur could
raise all the resources to establish a lab and had a good idea for a product,
scientists might not be willing to join a start-up, might have invested in
other firms or there could simply not be the expertise available to develop
and produce a revolutionary new product. Finally, even if there were
scientists who could do the job and were available, an entrepreneur might
still find it difficult to enter because the incumbent has patented every
conceivable innovation even loosely related to its current product, so that
an innovation by an entrant would almost inevitably violate some patent
or other by the incumbent.
129
MN3119 Strategy
MacAfee et al. (2004, p.463) give a list of entry barriers following a review
of the history of the concept. In particular, they distinguish between
economic and anti-trust barriers to entry:
Entry barriers – definitions
An economic barrier to entry is a cost that must be incurred by a new entrant but which
incumbents do not or have not had to incur.
An antitrust barrier to entry is a cost that delays entry and thereby reduces social
welfare relative to immediate but equally costly entry.
Entry barriers can be said to have two effects, a direct and an indirect one:
Entry barriers:
i. Make it difficult for firms to enter into a market and earn profits.
ii. Enable incumbents in the market to sustain higher levels of profits
than under perfect competition.
The first effect is obvious and basically defines entry barriers. The
second one is more subtle (but also more interesting): if a market has
entry barriers, profit levels need not be very low in order to make entry
unattractive. To see why, consider the following example. Suppose a
newcomer has to purchase a licence at a fixed price to start operating in
that market. The entrant will take this into account when making an entry
decision – only if the expected profits from entering the market exceed
the entry fee will entry take place. The higher the fee (the entry barrier),
the higher post-entry profits would have to be to trigger entry – putting it
the other way around, the entrant will find entry unattractive for higher
current profit levels, because it would still not recoup the entry fee. Of
course, this will be different if the licences are freely traded and the price
of the licence will settle on the level that lets entrants make a regular
rate of return, but for some markets there are fixed costs of entering that
cannot be traded. We distinguish between two types of entry barriers:
structural barriers or strategic barriers. Structural barriers originate from
the structure of the market, the product or the production technology
– they arise from external sources. Strategic entry barriers only exist
because incumbent firms change their behaviour in the face of (actual
or threatened) entry. Some entry barriers will have elements of both
structural and strategic entry barriers, but we will keep them separate for
clarity.
14.2 Structural entry barriers
Cost advantages. If a firm has been operating in a market for a long time,
it will have gathered experience about the product and the production
process and is likely to realise economies of learning (see Chapter 7).
Entrants in the industry would have the same cost structure if they
operated in the market for the same length of time, but at the time of
entry they are at a disadvantage – the incumbent will be able to produce
more cheaply than an entrant, which intrinsically makes entry less
attractive. Imagine a Bertrand-type competition with asymmetric cost. A
firm with a cost disadvantage will not make any sales, and if future cost
in turn depends on past sales, the incumbent will become even more
dominant.2
Economies of scale, scope and network effects. Incumbents are
typically much larger than entrants. This will give rise to economies of
scale (and scope, if they operate in adjacent product classes or markets
130
2
Cabral and Riordan
(1990) model such a
situation and show
that aggressive pricing
policies that would
seem predatory in other
situations are simply
attempts to proceed
down the learning curve
in such a market.
Chapter 14: Entry and entry deterrence
as well) which is difficult for entrants to compete with in the early stages
of their existence. Similarly, in a market with strong network effects,
entrants offer a smaller network and therefore a less enticing alternative
to the dominant firm. This problem becomes even more severe if we think
about goods with a degree of switching costs – the cost a consumer has to
incur if he were to switch from the brand he is currently using to another.
If a market is relatively mature and consumers are contractually bound
(as for example for mobile phone contracts) or have skills specific to one
version of the product (as is the case with end-user software), then at any
point in time, an entrant will only be able to compete for the unattached
consumers (whose contract is running out, say, or who are about to
upgrade from their current software), which makes the process of building
up an installed base a long and costly one. Clearly then, entrants will think
twice about entering into such a market.
Control of essential resources. This was already mentioned in the
chapter on vertical relations – if you control an essential input, other
firms will find it difficult to compete. This is an important entry barrier
and can partly be structural and strategic. For example, Taxi Badges in
New York City are limited, as are General Practitioners’ permits in most
countries – this has nothing to do with strategic behaviour on the part of
the incumbents, it’s the way the market has been structured. On the other
hand, DeBeers’ attempt to control most of the world’s diamond mines was
(among other things, as we have seen in Chapter 11) a strategic attempt to
shut out competition. The main message, however, is that if an entrant has
to settle for a less profitable resource (for example, a less central location
for a retail shop, a less rich ore mine, or less skilful staff), entry is less
attractive, creating a barrier to entry.
Marketing and reputation advantages. The first firm in the market
will often establish a brand reputation and brand recognition before
others enter. Sometimes the first brand on the market even becomes the
generic name for the product class. This need not make entry impossible,
however – what matters is that consumers make their purchase decisions
accordingly (see example below).
Example: Kleenex3
Kimberly-Clark manufactures the dominant brand of facial tissue sold to consumers
in the United States, a brand so familiar that it is almost synonymous with facial
tissues: Kleenex. Nationally, Kleenex brands have a market share (in 1994 revenues)
of approximately 48.5 per cent. Scott, which manufactures Scotties facial tissue, is the
third-largest producer of branded facial tissue in the United States, with a market share
(in 1994 revenues) of approximately 7 per cent. Only one other manufacturer, Procter &
Gamble, sells a significant amount of retail facial tissue in the United States. Kleenex and
Scott’s combined sales of facial tissue would be nearly twice those of Procter & Gamble,
which holds a market share (measured in 1994 revenues) of approximately 30 per cent.
Source: Antitrust case
United States of America
and State of Texas
versus Kimberley-Clark
Corporation and Scott
Company, www.usdoj.
gov/atr/cases/f0400/
0482.htm
3
It would seem that facial tissues are neither a particularly complex product that
Kimberley-Clark has a significant advantage in producing, nor would one expect huge
differences in quality. The fact that even a major competitor like Procter & Gamble has not
been able to challenge the Kleenex brand would seem evidence that ‘old habits die hard’
– consumers choose Kleenex because it is the first brand that comes to mind.
Barriers to exit/sunk costs. Entrants might also speculate that their
entry will trigger exit by the incumbent – in fact, most entrepreneurs will
probably feel they have something to offer that the market currently does
not offer, and if the new product is preferred by most (or all) consumers of
the existing product, the incumbent might well go out of business. There
131
MN3119 Strategy
are, however, reasons why an incumbent might find it profitable to remain
in a market even with an inferior technology. One is a classic option
value argument – if the entrant turns out to be less efficient than initially
expected or if the market could grow to allow for healthy profits for
both firms, an incumbent would not exit even in the expectancy of shortterm losses. The other set of exit barriers originates from the real costs
of leaving the market – this could be due to asset write-offs, redundancy
payments, contractual penalties for long-term contracts with suppliers, a
loss in goodwill from existing consumers who will find it difficult to obtain
spare parts, for example, or more generally a loss in credibility of being
able to compete effectively in a market.
14.3 Strategic entry barriers
Strategic entry barriers typically involve some cost on the part of the
incumbent. After all, the definition of a strategic entry barrier is that an
incumbent takes an action that would not have been taken without the
threat of entry. This of course means that is an entry-deterring action
does not involve any cost (i.e. if profits are higher with the action than
without), it would be part of the equilibrium set of actions anyway – entry
or no entry. A threatened monopolist trying to deter potential entrants by
choosing specific actions to deter entry will have lower profits. But bluntly,
it is always preferable not to have competition, even potential competition.
Strategic entry barriers then describe the class of actions that will decrease
profits compared to the pure (unthreatened) monopoly or oligopoly case
but have the goal of deterring entry.
Entry deterrence is profitable if:
a. The incumbent earns higher profits (despite deterring) as monopolist
than as duopolist (i.e. post-entry).
b. The strategy changes the potential entrants’ expectations about the
nature of post-entry competition.
The first condition states that entry deterrence must not be destructive
or too costly – in other words, letting an entrant into the market would
decrease profits even further than trying to keep the entrant out. The
second condition states that entry deterrence must be successful in the
sense that the entrant believes that post-entry profitability would be lower
under normal competition (i.e. without entry deterrence). Figure 14.1
illustrates the trade-off.
Block entry if NPV(A) > NPV(B)
Profits
A (block entry)
B (allow entry)
Time
Figure 14.1: The basic trade-off of early deterence.
132
Chapter 14: Entry and entry deterrence
In the early phase – that is, prior to entry, the monopolist’s profits are
lower if she deters entry (dotted line) than if she does not, If this succeeds
in keeping a potential entrant out of the market after point t* (which is
when the entrant would enter the market, if at all), profits will be higher.
This illustrates the basic trade-off very clearly: Sacrificing early profits to
make higher profits in the future. This trade-off also allows us to make
some educated guesses about the factors that will affect the likelihood of
entry deterrence taking place. If t* is far into the future (say, because the
technology an entrant would enter with is not yet fully developed), the
losses from lowering profits are large relative to the higher profits in the
future (remember that to get the Net Present Value (NPV), we have to
discount future profits as they are worth less than present profits), which
makes entry deterrence less likely. Also, the life cycle of a technology has
to be taken into account – is it likely that the technology will be replaced
relatively soon? If so, entry deterrence becomes less likely. Finally, we also
have to take into account the profit differences – how much lower are the
incumbent’s profits if it tries to deter entry, and how much lower would it
be if it permitted entry?
Activity 14.1
Show analytically in a model that the Net Profit Value of both strategies depends on the
discount factor.
Guidance on this activity can be found in the VLE.
Activity 14.2
Taking into account the trade-off outlined above, rank these situations in terms of their
likelihood and profitability of entry deterrence.
a. Old technology about to be replaced
b. Stable product with Bertrand competition and new, more efficient technology
c. Unrelated products
d. Restaurant with long planning permission process
Guidance on this activity can be found in the VLE.
On a conceptual level, this seems sensible – sacrificing some profits to
avoid the worst case of someone else entering. What now determines this
trade-off in practice and what do firms do to deter entry?
Limit/predatory pricing. Sometimes even before entry takes place,
prices in an industry drop (limit pricing). Incumbents will try to signal
to entrants that entry would not be profitable in a market with such low
prices. Failing that, incumbents frequently lower their prices after entry in
order to drive competitors out of the market (predatory pricing). Both
these strategies certainly fulfil the first condition of profitable predatory
pricing – profits are lower with competition than with entry deterrence.
The problem lies in the second condition – why would a potential (or
actual) entrant believe that this is a long-term sustainable price? It could
be argued that once entry takes place, the incumbent has an incentive to
raise prices again and share the market. This then makes it a non-credible
commitment to lower prices in the first place, as the entrant knows that
the incumbent knows that keeping prices low forever is not rational (if this
does not trigger the concept of ‘sub-game perfection’, now is a good time
to reacquaint yourself with it – see Chapter 4). Limit and predatory pricing
are therefore promising ways of deterring entry only if the incumbent can
somehow convince the entrant that the new, lower price can be sustained
indefinitely, for example, by signalling low cost.
133
MN3119 Strategy
Example: Ferries v Eurotunnel
For several decades, P&O and Stena Line were a stable duopoly offering cross-channel
sailings between Calais (France) and Dover (England). There were, however, persistent
rumours that a tunnel was going to be built by an international consortium and with
significant support by the two governments. As these rumours gathered momentum and
a consortium of interested parties started to emerge, P&O and Stena Line took a number
of coordinated actions. First, they replaced their current fleet with new, more fuel-efficient
ferries. Second, they lowered prices from approximately £120 to £80 per crossing.
Third, they announced that the new and upgraded fleet enabled them to lower prices
significantly and permanently. All three actions in conjunction suggested that this was an
attempt at limiting pricing designed to keep the Channel Tunnel from being built.
Sadly, the attempt was unsuccessful, the tunnel was finally built and opened in 1994,
and current prices for a regular crossing hover around the £90 mark. Still, this was an
illuminating attempt at creating uncertainty about the cost structure of a firm (or a set of
firms in this case) to signal the long-term sustainability of low prices.
Capacity expansion. Firms typically choose a capacity level
approximately equal to the amount they expect to sell. An optimal
choice of capacity will of course also include some allowance for
unexpected rapid upturns in demand, for expected future growth and
so on. Frequently, however, we witness firms in an industry that carry a
surprisingly large amount of overcapacity that cannot be explained by any
of these factors. Indeed, overcapacity can also serve as an entry-deterring
mechanism. Effectively, a firm will signal to potential entrants that they
will face a fierce fight with the incumbent who is in a position to flood the
market with output. Why would the entrant believe this? This is a valid
question which can best be answered by asking what credibility means in a
game-theoretic framework. A strategy in the future, as you will remember
from Chapter 2, is only credible if it is a sub-game perfect strategy. In other
words, if the firm has no incentive to carry through a threat once its bluff
is called, there’s no reason to believe the commitment.
In this case (as illustrated in Figure 14.2), the threat of utilising one’s
overcapacity comes basically from the act itself. In other words, if a firm
did not plan to use its overcapacity on a rival entering the market, it
would not have built the capacity in the first place. Of course, if a firm’s
cost structure is well known and it is obvious to an entrant that increasing
output would clearly be sub-optimal even if overcapacity had been built,
the entrant should call the incumbent’s bluff. With sufficient uncertainty
about the incumbent’s cost, however, overcapacity will successfully serve
as an entry deterrent.
Enter
Normal capacity
[2, 1]
E
not Enter
[4, 0]
I
Enter
Excess capacity
[1, -1]
E
not Enter
Figure 14.2: Pre-emption with capacity.
134
[3, 0]
Chapter 14: Entry and entry deterrence
As we have stated earlier, this might pose some problems because the
threat to utilise the overcapacity might not be credible, or sub-game
perfect. One argument would run like this: if I was not planning on
carrying out my threat, why would I build it in the first place, so it should
be credible once I have built it. Another, slightly cleaner argument would
involve using the overcapacity even before entry takes place – in other
words, producing more than the optimal (monopoly) quantity serves as an
even more credible signal to entrants than simply having the capabilities to
overproduce in response to entry. This is where theories of overcapacity
and limit pricing converge – prices will be lower in both cases, thus
making the market less attractive to entrants and creating a belief that
these prices or quantities would be upheld even after entry.
Endogenous sunk costs.4 We also mentioned the sunk costs of entry in
the section on structural entry barriers. That it reappears here illustrates
the dual nature of entry costs. On the one hand, some expenses are
inherent in the market – building up a brand, say. On the other hand,
how much an entrant would have to spend is also to some extent in the
hands of the incumbent. For example, if Kleenex had been the first firm
to produce, well, Kleenexes (or paper tissues, to be precise), but had not
spent a great deal of advertising, Procter & Gamble (P&G) would have
had to spend a certain amount on getting their product known and onto
the shelves of the major supermarkets. If, on the other hand, Kleenex had
invested greatly in building a national brand with strong recognition, then
P&G might not have sold much, even if it had made consumers aware
of its product and had successfully placed it on supermarket shelves.
The problem clearly is that it is just as important in such a context to
establish the new product to established brand, which turns out to be a
function of the amount the incumbent has invested in branding. (A similar
argument also holds for research and development (R&D) – an incumbent
might invest in the latest technology and continually improve its existing
products not because it expects huge sales growth, but because this
makes it almost impossible for a rival to enter this market and match the
incumbent’s product in quality, etc.) Thus, incumbents can keep entrants
out by ‘overinvesting’ in a brand or technology compared with the level
they would invest if there was no threat of entry.
Consumer/asset lock-in. Locking-in consumers or assets is also partly
external (i.e. structural) and partly internal (i.e. strategic). The important
distinction is the way in which lock-in occurs. In other words, if lock-in
is part of an equilibrium strategy with or without the threat of entry, it
is likely to be exogenous. On the other hand, if consumers are locked-in
long-term contracts by aggressive pricing, lock-in is strategic in nature.
For example, Nutrasweet succeeded in staving off the threat of Holland
Sweetener’s entry by signing Coca-Cola and Pepsi, the world’s largest
buyers of artificial sweeteners, on long-term contracts, leaving Holland
Sweetener with the leftovers. Nutrasweet, as a consequence of this
strategy, was able to maintain a much higher market share than if it had
not locked in Coca-Cola and Pepsi.5
Product proliferation. Finally, the number of products that an
incumbent sells in a market is also related to the attractiveness of entry.
If consumers are differentiated in their preferences and an incumbent
chooses to locate products in product space (by choosing a certain set of
characteristics appealing to some, but not all consumers), an entrant will
choose to locate its products in ‘holes’ left by the incumbent. If these
holes are too small for an entrant to profitably sustain a brand (taking
into account the cost of launching a new variety, etc.), entry will be
This section is based
on the work of John
Sutton (Sunk costs
and market structure,
MIT Press 1991 and
Technology and market
structure, MIT Press
1998), who shows that
some sunk costs are
endogenous and can
have a significant effect
on market structure.
4
Source: Brandenburger,
A. and B. Nalebuff,
Competition. Harper
Collins Publishers, 1996.
5
135
MN3119 Strategy
deterred. For methodological and data-related reasons, the market for
breakfast cereals has been the leading example for academics for this
kind of strategy and it is now commonly accepted that Kellogg’s (as the
dominant supplier) produces more varieties than it would as a monopoly
in order to make it more difficult for entrants to target market niches left
by it.
Activity 14.3
Which of the above strategies of entry deterrence (if any) are likely to work best in the
following industries?
a. Airlines
b. Pharmaceuticals
c. Office furniture
d. Breakfast cereals.
Guidance on this activity can be found in the VLE.
14.4 Summary
Entry by new rivals will almost always have a negative effect on the
profitability of incumbent firms. Similarly, the prospect of profits in an
industry is typically what attracts entrants into a market in the first place.
In this chapter, we have discussed and analysed two types of entry barriers
that allow incumbents to maintain market share even if newcomers
threaten to enter the market. This can either mean that newcomers enter,
but are only able to do so on a smaller scale (so that the incumbent
maintains a higher market share) or that the entrant succeeds in keeping
all entrants out of the market and remains a monopolist. We distinguished
between exogenous, or structural, and endogenous, or strategic, entry
barriers – structural barriers are a godsend for incumbents as they do
not have to change their behaviour at all and entry is still unattractive.
Strategic entry barriers are the ones erected by incumbents, sometimes at
a considerable cost to them. The reason they do this is frequently to signal
to entrants that entry would not be profitable, or to change the structure
of the market or the product that entry indeed would either be very
expensive or not very profitable.
14.5 Key concepts
136
•
Structural entry barriers
•
Control of essential resources
•
Marketing and reputation advantages
•
Barriers to exit
•
Strategic entry barriers
•
Entry deterrence
•
Limit/predatory pricing
•
Capacity expansion
•
Endogenous sunk costs
•
Consumer/asset lock-in
•
Cost advantages
•
Product proliferation.
Chapter 14: Entry and entry deterrence
14.6 A reminder of your learning outcome
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
define structural and strategic entry barriers and compare the two.
14.7 Sample examination questions
1. Entry deterrence. How will fixed (entry) cost affect market outcomes
in a symmetric Cournot game and in a quantity-setting game with a
Stackelberg leader? Discuss separately the cases in which fixed cost are
low and high, respectively.
2. Explain how the likelihood and profitability of entry deterrence
depends on the degree of substitutability of your product with the
product of the potential entrant.
3. What are the major barriers to entry into the brewing industry and
how did the microbrewers get around them?
Guidance on answering these Sample examination questions can be found
in the VLE.
Extended activity: Dubai flowers and internet banking
Read the following and answer the questions at the end.
Dubai threatens Dutch hegemony in global floriculture industry
In 2005 the Dubai government opened the Dubai Flower Centre (DFC)
with the goal of becoming a global floriculture logistics hub, and grabbing
a significant share of the $4.7 billion global floriculture industry, which
involves the growth and distribution of cut flowers. The move threatens
the incumbent position of the Dutch whose players have for the past 50
years dominated this industry, and now enjoy a 60 per cent share of the
global flower trade.
Within Dubai’s own business landscape, the DFC supports the growth of
Dubai’s flag-carrier Emirates airline by increasing cargo flows, and the
state’s business and political leader Sheik Ahmed bin Saeed Al Maktoum
has accordingly invested $70 million in the DFC facilities to date, with
further expansion planned.
The core driver of this industry is the maintenance of a cost-effective and
efficient ‘cool logistics chain’ or refrigerated distribution system of these
highly perishable products.
From a global market perspective, the DFC enjoys several key competitive
advantages. Dubai’s tax-free policies and geographic position allow
shippers routing through the DFC to save money and shorten delivery
times, particularly in linking the producing nations in the Southern
Hemisphere – where for example African flower production is growing
at a rate of 20 per cent — to the top importing Asian and Middle Eastern
consumers.
Indeed, the entry of the DFC into the market could change the current
distribution pattern whereby an orchid grown in Sri Lanka and sold in
Japan must take the intermediary step of travelling north to Amsterdam
to be auctioned, packaged and redistributed. In an article published by
The Times of India, the managing director of Karuturi Floritech predicts
a radical shake-up of the floriculture export business. He cites the 30
137
MN3119 Strategy
per cent transaction cost (including import tax and handling fees)
associated with trading through Amsterdam as being the golden ‘arbitrage’
opportunity of DFC, who can potentially bring these costs down to below
10 per cent and thereby attract a big share of the market.
The impact of Dubai’s entrance in the market remains to be seen.
‘The DFC insists that it is not seeking to take business from the
Netherlands but that it will open new trade routes for perishable goods
between Africa and Asia,’ reported the Financial Times, which also
quoted the CEO of the Flower Council of Holland as saying, ‘It took the
Netherlands half a century to build its horticulture export industry…so I
don’t believe that can be copied overnight’.
Still, Dutch brokers are already investing in overseas facilities to protect
their market share.
Internet banking – a conversation with Betsy Z. Cohen
Betsy Z. Cohen is the Chairman and Chief Executive Officer of The
Bancorp, Inc. (NASDQ: TBBK) and its subsidiary, The Bancorp Bank,
which provides internet banking and financial services to small and midsized businesses and their principals nationwide in the US. The Bancorp
Inc. has a market capitalisation of $264 million. Since 1997 Mrs Cohen
has also served as Chairman and Chief Executive Officer of the real
estate investment trust, RAIT Investment Trust. A lawyer by training, she
has been a leader in the US banking industry for over 30 years, having
founded the Jefferson Bank in Pennsylvania in 1974. In 2005 US Banker
magazine named Mrs Cohen one of ‘25 Women to Watch’.
What is an internet bank?
The internet has evolved, and internet financial products and banking have
evolved, over the past five years as banks have grown and understood what
the medium is all about. The term ‘internet banking’ is not a unitary kind
of description. For most banks, it really refers to a distribution mechanism
which is very powerful but which is not an end in itself.
Internet banking also refers to banks without branches. It is the
positioning of an institution in contra-distinction to the recent 10-year
trend in the positioning of bricks-and-mortar banks in the US, in which
they are creating branches primarily as collection points for deposits, and
in order to reinforce their brand.
To some extent the branchless internet banks are taking today a different
position – that the banking that gets done on the web is supportive of
another business or a link to another business. So if you’re taking e-trade,
they’ve put together an online brokerage business and now own a range
of financial services all of which are internet-based primarily but which
support one another.
Some ‘Net’ banks, however, simply went on the web and held out a shingle
for retail clients. ING does this. ING uses their internet bank, which is
very aggressive in granting higher rates of return on deposits in order to
pull the customers in. What they’re using those customers for, though, is
as a base with which to sell their profitable products, which are their
insurance products. ING is another non-bricks-and-mortar-bank, except
that they’ve started internet cafes, which are physical spaces.
What does Bancorp do?
We chose a different path, in part because I would rather be doing
wholesale business, which was to provide private-label banking products
138
Chapter 14: Entry and entry deterrence
for companies that are not banks, but that want to provide banking
products to their customers. The only way we could really do that
effectively is through a web site. We couldn’t do it by physical branches,
because our partners are too spread out across the country.
So is this like a department store credit card?
It is in the nature of that, yes. Bancorp leverages the customers that our
partners have, and, from our perspective and from their perspective,
it puts a fence around those customers. We’ve chosen lines of business
that are in competition with a bank, and have helped our partners avoid
exposing their customers to a banking institution in order to get them
banking products. So we’ve chosen asset managers that are not banks and
not connected with banks – either non-bank trust companies or companies
that just manage your assets without regard to other parts of your
financial life.
For example, one of our partners is a company called SEI which is
traded on the New York Stock Exchange. Part of their business is that
they administer about $450 million worth of mutual funds, and they
have about $30–35 million worth of pure asset management for wealthy
families. What they used to do was say to Mrs Smith who was building a
second home, ‘We’ll take your portfolio to Smith Barney or we’ll take you
to Wachovia and get you a construction loan or a securities-based loan for
this project,’ rather than have her sell her stock.
Wachovia and Smith Barney each got a look at Mrs Smith’s financial
statement and said, ‘Wouldn’t it be nice if we managed her assets. Mrs
Smith, if we managed your assets, you wouldn’t have to come to us
separately and have all the heartache, we’d just do it!’ So SEI said, ‘This
is not working out.’ But they knew it was a product that they had to offer.
So we have set up the SEI financial network for them, which offers deposit
accounts and the basic banking products their customers need. They do
business by having business referred by 9,000 financial advisers, so the
financial adviser says to the customer, SEI can provide that through its
financial network, and then it’s done. So we have 150 of those partners in
three lines of business.
Why did you start an internet bank? What opportunities did you seek to
exploit?
We saw an opportunity to have a different business model. Instead of
holding out our shingle and waiting for retail customers to come in,
we looked to find an overarching customer who had a need to provide
banking products to their customers, and needed a bank involved, but
didn’t want to be a bank. So we would become that bank for them.
One of the things regulators are so concerned about is
maintaining competitive local markets where there are many
smaller players, and not just one big bank monopolising
things. Let’s say your cohort of 20 Internet banks grows
and becomes the dominant model, wouldn’t that change the
way regulators would be able to define and determine local
market concentration?
Yes, it’s counter to the whole community-based model. We applied for an
internet charter based in Delaware. But really since we don’t have any
branches and we don’t have any branch restrictions, there is no restriction
on where our emails can go.
139
MN3119 Strategy
So in effect it’s a national charter?
Exactly. That’s why they no longer issue something called an internet
charter – it would now have to be a national or international charter
– because they realised that their whole banking system was based on
community need and this is just something very different.
Will internet banks change the structure of the banking industry, and the
way that regulators look at the industry?
Again, let’s differentiate. There are really only 20 charters like ours out
there in the country. And the reason is because regulators realised after a
bit – and we are fortunate that it was after rather than before – that what
they were granting was a national and indeed an international charter,
which is not how the US banking system works. So we can talk about
those 20 banks, or we can talk about how internet access as a distribution
mechanism will change the way in which people do banking. We are just
now at the beginning of being able to see that because we are at the end
of this 10-year trend where banks were branding.
Let’s talk about the 20 banks. They are a pioneering group, doing
something different, no?
In fact they are. It is a different business model, but it’s a model that
began just as the internet boom was fading and about to become a bust
and, except for these 15 or 20 banks, has not expanded since that time. In
part it’s because people today are still using the standard banking model,
which is the branding model. In part it’s because people believe that
successful internet businesses must be like JC Penney, which generates the
most traffic on the internet of any URL in the world, and which depends
on people being able to go into a JC Penney store, touch something and
order it over the internet. That model of dual distribution, where you can
touch and see and build trust, and then go on and do transactions over the
Internet, they believe is essential. So it’s really this model that they haven’t
broken away from.
It seems that some face-to-face contact would be important in order to
build trust.
In fact the reason we chose our particular business model is that we
wanted someone to know the customer. So SEI has already vetted their
customers. They all have to keep $2.5 million or more in assets to be
managed by SEI, so that’s easy.
Let’s assume that, as time goes on, the internet as a mode of distribution
will become more widely accepted. Do you think that this will diminish
the need for branches?
As banks find their branches less effective gathering points, they will invest
more in beefing up their internet access. It’s a combination of things that
will occur.
We at Bancorp are one of maybe two or three business models that
are really different. I don’t know what’s going to happen. I don’t know
whether people will come up with different business models or not. So far,
they don’t seem to have done that!
Sources
Lane Fox, H. ‘From catwalk to conveyor belt Amsterdam is to the flower trade what
Paris is to the fashion industry’, The Daily Telegraph (London), 14 February 2004.
McSheehy, W. ‘Flower power in Dubai could be a thorn in the side of the Dutch’,
The Financial Times, 20 October 2005.
140
Chapter 14: Entry and entry deterrence
Saifur, R. ‘Dubai flower centre to build complete cool logistics chain’, Gulf News, 27
September 2004.
Sujit, J. ‘Floriculture set to bloom’, The Times of India, 7 December 2003.
Questions
1. Mrs Cohen makes reference to a number of different banking business
models that employ the internet. Enumerate these examples and
explore their relative strengths and weaknesses.
2. What are the barriers to market entry a branchless internet bank faces?
Are they different from the barriers faced by a traditional bricks-andmortar bank? How did Bancorp navigate these barriers and develop a
successful business model in spite of them?
3. The Bancorp Inc. has found a niche in helping wealth management
companies retain their customers. What other avenues for growth
could Bancorp pursue?
141
MN3119 Strategy
Notes
142
Chapter 15: Research and development competition
Chapter 15: Research and development
competition
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
discuss various types of innovation
•
explain the concept of replacement and efficiency effect
•
assess the impact of market structures on R&D activities.
Essential reading
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) Chapter 16.
Further reading
Angelmar, R. ‘Market structure and research intensity in high-technologicalopportunity industries’, Journal of Industrial Economics 34(1) 1985, pp.69–79.
Cabral, L. ‘R&D competition when firms choose variance’, Journal of Economics and
Management Strategy 12(1) 2003, pp.139–50.
Gilbert, R. and D. Newbery ‘Preemptive patenting and the persistence of
monopoly’, American Economic Review 72(3) 1982, pp.514–26.
15.1 Introduction
Research and development (R&D) is often heralded as one of the key
functions within a firm. Even in industries not traditionally thought
of as ‘high-tech industries’, development of new products takes centre
stage in the considerations of managers and strategists. In the context of
R&D, a number of important strategic questions come up consistently.
In particular, how do firms choose the intensity and riskiness of their
research activities, and how do R&D activities depend on the nature of
the innovation and their position in the technological ‘race’ towards an
innovation. Of course, a great many innovations come about through a
stroke of luck or serendipity,1 but it would seem sensible to assume that
the majority of innovations come about because their inventors actively
strove towards this particular innovation.
15.2 Terminology
What do we mean when we talk about innovation and R&D? There are
several distinctions we should make.
Closeness to the end user. The closer to the end user an innovation
is, the more ‘applied’ it is. That is, research that is applicable to a wide
range of potential uses and requires extensive work and market research
to make it palatable to the end consumer, is considered basic research.
Applied research is defined as ‘the application of scientific knowledge
to the solution of a specific, defined problem.’2 This basically means that
research is done with a specific application and potentially even end
user in mind. Product development is even closer to the end user: an
The now-ubiquitous
Post-It notes by 3M
were discovered in the
process of developing
a much stickier glue.
Similarly, Viagra was
only found to have its
well-known effects
when it was tested as
a medication for heart
diseases.
1
Source: http://research.
uiowa.edu/dsp/main
2
143
MN3119 Strategy
end product exists and consumers have been readily identified and the
process of product development is about tailoring the product to the target
consumers’ needs.
Activity 15.1
Categorise the following activities?
a. Second-stage drug testing to fulfil FDA (Food and Drug Administration) requirements.
b. A celebrity chef combining lemongrass with different cuts of meat to find out about
the interaction of flavours and textures.
c. Apple developing a hard disk that would double the capacity of its iPod.
Guidance on this activity can be found in the VLE.
Of course, the stage of research also impacts on the way in which firms
do research. The outcomes of basic research are often difficult to protect
and are subject to knowledge spillovers between firms. That is, one firm’s
results are likely to trickle to its competitors, either through personnel
movement or because basic research is published in scientific journals
(for everyone to read). This creates a free-rider problem: given that we
are bound to benefit from other firms’ basic research efforts, we have an
incentive to let others do it (and save the costs of setting up a large R&D
lab, hiring scientists, etc.). Since of course other firms will think similarly,
we expect that commercial firms do less basic research than we would
wish. And indeed, we can see that most basic research is done either by
universities, public laboratories or even networks of firms. The latter is an
effective way of internalising the externalities (or spillovers) from basic
research – if everybody contributes and all firms that would benefit from
spillovers join such a research network, participating firms will contribute
according to their marginal benefit.3
How does this compare to applied research or product development?
First of all, new products and processes can often be protect by patent
or copyright. This makes it more feasible to stop others from making
use of this innovation, meaning that spillovers will be lower. Copyrights
and patents also require that an innovation is distinguishable from ‘prior
art’. This implies that if two firms are developing very similar products,
whoever comes second might not be granted the patent – this has
encouraged scholars to compare the process of looking for a patentable
innovation to a ‘race’ – only the first to pass the finish line gets the prize.
This opens up a new type of competition: what my rival does matters to
me – if Bayer Pharmaceuticals has just announced a significant increase in
its research budget, should BASF try and match this increase to stay in the
race or is it better to just give up? These and similar questions of strategic
interaction are what we will focus on most in this chapter.
Product versus process innovation. We can think of innovations
as an improvement of the existing product – all (or most) existing or
potential buyers experience an increase in their willingness to pay.
Graphically, this has the effect of shifting out the demand curve for a
particular product. Second, an innovation could also be a novel way of
producing an existing product – say, a new production process that makes
producing the existing product cheaper. Graphically, this implies a lower
marginal cost curve. Of course, in real life most innovations will have
elements of both, but let’s keep them separate to keep things simple. For
most of our analyses we will use process innovations, but this is purely
because it is simpler to draw – our results and conclusions would hold for
product innovation as well.
144
3
One could argue,
however, that this simply
pushes the free-rider
problem one step back:
what are my incentives
to join the network if
I would benefit from
spillovers even I did not
join?
Chapter 15: Research and development competition
Product vs. process innovation
DNEW
MC
MCNEW
D
D
Figure 15.1. Product versus process innovation.
Drastic versus incremental innovation. Most people will have an intuitive
idea of what drastic and what incremental innovations are. For example, a
drastic innovation is often considered one that changes the business model
of the industry (whatever that may mean). Without knowing exactly how
to define ‘business model’, however, it is difficult to define a threshold of
innovativeness that warrants calling an innovation radical. Most will agree
that the compact disc was a radical innovation because it revolutionised
the audio market, but was Dolby surround sound? The digital audio tape
(DAT)? We will give a more precise definition of radical and gradual
innovations.
Types of innovation – definition
An innovation is drastic if an innovator could behave as a monopolist in the market
despite the fact that there are substitute products (without access to the innovation)
on the market. Conversely, an innovation is incremental if it increases profits for the
innovating firm but its strategy is still restricted by the other non-innovating firms.
Let us illustrate this with a process innovation. In Figure 15.2 both
innovations lower the marginal cost of the innovator somewhat. It is
relatively easy to see that the innovation to the left is more radical on the
right since marginal cost goes down more significantly. But how will an
innovator behave in each of these markets?
• Product vs. process innovations
• Drastic vs. gradual/incremental innovation
p M (i)
p M (d)
p c = MC OLD
MC NEW(i)
MC NEW(d)
MR
D
MR
D
Figure 15.2: Types of innovation.
145
MN3119 Strategy
In fact, an innovator in the left panel (i.e. lowering his marginal cost to
MCNEW(d)) can price as if he were alone in the market – his monopoly price
pM(d) is lower than any price one of the other firms (who still produce at
MCOLD) would want to set. On the other hand, if an innovation lowers
marginal costs to MCNEW(i) as in the left panel, monopoly pricing would
mean that the innovator is pricing himself out of the market – pM(i) is higher
than the lowest price his rivals will want to go – MCOLD.
Activity 15.2
Given the above definition of radical and incremental innovations, would you still agree
that the CD was a radical innovation? What test would you apply to establish if it was or
was not?
Guidance on this activity can be found in the VLE.
So armed with the definitions given above, we will now try to establish
some predictions about the effect of market structure and a firm’s relative
position on the intensity and character (i.e. the riskiness of a research
strategy) of innovation.
15.3 Innovation and market structure
In which markets will firms innovate most? On the one hand, firms in
competitive markets could be said to have strong incentives to innovate:
after all, they are not making huge profits from their existing operations
and have every incentive to innovate and differentiate themselves from
their competitors. However, firms that have a strong position in their
industry will claim that their dominant (or even monopolistic) position
is under threat and that they innovate because they have to. Of course,
which side of the argument you believe will depend on where you stand –
Microsoft’s lawyers will point out that Microsoft’s R&D/sales ratio is well
above the industry average and that their dominance stems from their
superior ability to come up with innovations that consumers want. In
contract, entrepreneurs will stress their role as an ‘engine of innovation’4
and point out that they should be receiving tax breaks, subsidies and other
help for their service to society. So which argument is right? Are they both
correct?
Let’s look at the static efficiency of different market structures for help: any
economics textbook will tell you that the most efficient market structure (if
there are no notable economies of scale) is a perfectly competitive market:
prices will be at marginal cost, and every consumer who is willing to buy
above that marginal cost will be served. Monopoly power only introduces
deadweight losses. However, consider investing in R&D knowing that the
market will remain competitive after the innovation: how much would
you invest in such an innovation? Nothing, since you are not expecting
any profits from innovating. Scholars therefore agree now that in order to
promote innovation, there needs to be some degree of monopoly power
(as a carrot), but a sufficient threat of entry (as a stick). This notion is
represented in the patent system, where a patent is ‘a set of exclusive rights
granted by a government to an inventor or applicant for a limited amount
of time’, thus granting the rights to extract the benefits from an innovation,
but only for a limited period of time and a restricted range of application so
that other innovative firms can ‘invent around’ the patent or freely use the
technology after the patent runs out or is not renewed.5
Let us now analyse this conundrum in detail by looking at a very simple
situation: a process innovation would lower a firm’s marginal cost from c
146
Kate Barker,
Confederation of British
Industry.
4
5
A patent holder must
choose to renew a
patent periodically
at a small cost. If a
firm decides that the
potential profits of
keeping the patent are
not worth the cost, it
goes into the public
domain and can be used
by other firms (who
might have a better idea
of what to do with the
innovation).
Chapter 15: Research and development competition
to c. Looking at different market structures, what is a firm’s willingness to
pay for this innovation?
We first look at a competitive market with no patent protection. Lowering
marginal cost is good for consumers, but since the innovating firm would
not get any profits from developing the innovation, its willingness to pay is
zero (Figure 15.3).
• Competitive firm, non-appropriable innovation
c
∆ CS
c
Figure 15.3: Who benefits most from R&D?
What if a competitive firm can protect its innovation? Assuming (as we do
in the graph) that it is an incremental innovation, they will set a price of
(just below) c and make profits of DP.
• Competitive firm, proprietary innovation
c
∆Π
c
Figure 15.4: Who benefits most from R&D?
What incentive will a monopolist have? To start with, profits from
innovating are higher if you are guaranteed a monopoly position
afterwards – for an incremental innovation, not being limited by any
competitors (even less efficient ones) is always preferable. On the other
hand, a monopolist also makes significant profits in the first place, that is,
before innovating. Figure 15.5 illustrates this.
• Monopolist replacement effect
p0
p1
Π1
Π1
c
c
Figure 15.5: Who benefits most from R&D?
147
MN3119 Strategy
So if we wanted to compare the incentives to innovate of a competitive firm
and a monopolist, we have to take into account two things: profits with the
innovation and profits without it – the difference between the two being the
incentive, or willingness to pay, to innovate. This might seem trivial, but we
will show in the following numerical example that it matters.
Example
Let us look at the incentives to innovate for a firm in a competitive industry and a
monopolist. Suppose that the demand function is P = 5 – Q in both markets, and
that marginal cost without the innovation is MC(old) = 3 and MC(new) = 2 after the
innovation. The following calculations show that the incentives for a monopolist to
innovate are lower than for a competitive firm. (Make sure you know how we derived
these numbers!)
Comp Industry:
P(old) = 3,Q(old) = 2,P(old) = 0
P(new) = 3 – e,Q(new) = 2,P(new) = 2
Monopoly:
P(old) = 4,Q(old) = 1,P(old) = 1
P(new) = 3.5,Q(new) = 1.5,P(new) = 2.25
From the example above we can see that a monopolist is less inclined to
invest in an innovation because he is replacing old profits with new ones. A
monopolist’s lower incentive to innovate because of the higher level of preinnovation profits is summarised in the replacement effect.
Replacement effect – definition
For a given market structure, a monopolist has less incentive to innovate because of the
higher level of pre-innovation profits.
So does this mean that monopolists always invest less in R&D? Not
necessarily. After all, most of the significant R&D in pharmaceuticals
is done by large firms, Microsoft is still considered a pretty innovative
company, and firms like IBM, AT&T, Nokia and others are constantly at
the top of the pile of patent applications and other measures of innovative
activity. Why do they do it, given the replacement effect? They innovate
‘because they have to’ for fear of losing their dominant position.
Consider now the following simple scenario: a monopolist and a potential
entrant are competing for an innovation that would lower the marginal
cost from c0 to c1. It is an ‘innovation race’ in the sense that whoever
spends more on the innovation gets the innovation – for example, by being
granted a patent or by being first on the market. Again, we are looking
for the willingness to pay for the innovation – the belief that whoever is
willing to spend more will end up getting it. Post-innovation there are two
possible scenarios: either the monopolist gets the innovation and remains
a monopolist (but with higher profits because the new product is more
efficient) or the market becomes an asymmetric duopoly with the entrant
taking more than half of profits because he is using the better technology.
Starting from the initial situation, what are the two players willing to
invest? The entrant makes zero profits if he does not innovate and would
get Pd(c1,c0) if he innovates, which is therefore his willingness to pay. The
monopolist defends his monopoly position and gets Pm(c1) if he innovates
and gets the smaller share of duopoly profits Pd(c0,c1) – the difference
between the two is the monopolist’s willingness to pay.
If we want to find out who invests more in innovation, we need to compare
the two:
148
Chapter 15: Research and development competition
Pm(c1) – Pd(c0,c1) > Pd(c1,c0)
is the condition for the monopolist to invest more. Is this the case?
Rearranging the inequality gives Pm(c1) > Pd(c1,c0) + Pd(c0,c1) – which is
given since monopoly profits are always higher than the sum of duopoly
profits! Therefore we can say that if a monopolist knows that he would lose
his monopoly position unless he gets the innovation, he will invest more in
R&D than the entrant. This has been termed the efficiency effect.
Efficiency effect – definition
Under the threat of entry, a monopolist has higher incentives to innovate than a potential
entrant into the market.
Example
We will use our standard demand function again, this time allowing for two firms in the
market, a monopolist (M) and an entrant (E): P = 5 – Q, where Q = qM + qE. Marginal
cost before the innovation is MC(old) = 3, new marginal cost will be MC(new) = 2. A
monopolist will make profits of 1 before the innovation. If he innovates, profits increase to
9/4 – the potential entrant’s profits are zero.
What happens if the entrant innovates? We have an asymmetric duopoly
with the monopolist facing marginal cost of MC(old) = 3, and the entrant
MC(new) = 2. This results in quantities of qM = 1/3 and qE = 4/3. Profits
will be pM = 1/9 and pE = 16/9. Note that the sum of duopoly profits are
pM + pE = 17/9 < 9/4 – competition erodes profits.
The monopolist will therefore be willing to pay 9/4 – 1/9 = 77/36 for the
innovation, while the entrant only bids 16/9 = 64/36. The monopolist’s
willingness to pay is therefore higher because he wants to protect his
monopoly position, while the entrant would only be a duopolist.
We have looked at two simple, but reasonable models that gave us two
very different results: the first one says that a monopolist will invest less
in R&D, the second one says he will invest more. We will now extend our
models to find one that generates both outcomes.
Consider now a situation in which a monopolist does not know for sure
if there is a firm trying to enter the industry. If research is secret, for
example, or if it is not clear if another firm has the capabilities to develop
a technology of sufficient quality, this seems realistic. It seems reasonable,
however, to assume that the monopolist has at least some idea if there
is a potential rival for an innovation – suppose he thinks that there is a
probability r that there is a rival. This changes the situation to that shown
in Figure 15.6.
•• Model with uncertainty
r = probability that rival is doing R&D in this area
(‘bids fo the innovation)
•• WTP fpr entrant: Πd(c1,c0)
•• WTP fpr monopolist:
Πm(c1) – (rΠd(c1,c0) + (1 –r)Πm(c0))
Πm(c0) > Πd(c0,c1), so dWTP/dr > 0
→
High r: efficiency effect > replacement effect;
Low r: vice versa
Figure 15.6: Market structure and R&D.
149
MN3119 Strategy
We have now integrated both situations: if there is hardly any threat of
entry (i.e. low r), the replacement effect will be more important and a
monopolist has less incentive to innovate. If the threat is relatively high
(high r), the efficiency effect will dominate and the monopolist will have
to innovate to stay ahead of the (as yet non-existent) competition.
So far we have only looked at incremental innovations – what about
drastic ones? If there is at least some chance that there is no rival
innovator, a monopolist will have a lower willingness to pay to innovate
than an entrant.
•• Previous model, but drastic innovation:
Pm(c1) for whoever gets the innovation, 0 for the loser
•• WTP fpr entrant: Pm(c1)
WTP fpr monopolist: Pm(c1) – (1 –r)Pm(c0)
→
Monopolist is less ‘keen’ on drastic innovations than the entrant, so in real life:
•• Incremental innovations: incumbents;
•• Drastic innovations: entrants.
Figure 15.7: Market structure and R&D
Activity 15.3
We use our standard set-up again: Q = 5 – P, MC(old) = 3, MC(new) = 2. Suppose that
firms compete in prices and produce identical goods. There is a probability r that an
entrant will be actively competing for this innovation. Show that:
a. The willingness to pay for an entrant is WTPE = 2
b. The willingness to pay for a monopolist is increasing in r
c. WTPM > WTPE for r > 3/4.
Guidance on this activity can be found in the VLE.
Combining all this now suggests that a monopolist may be more
innovative in some situations where the threat of entry is quite high and
the innovations are incremental, but that an entrant is more likely to
implement drastic innovations and (if he is capable of innovating) if the
perceived threat of entry is low and the innovation is incremental.
Activity 15.4
Interpret the following quote in light of the theories about innovation incentives.
Background: Xerox has an early monopoly in the plain paper copies market, but saw its
patents run out at some stage. Existing consumers were to some extent locked-in due to
service contracts with Xerox, and Xerox had a strong brand name in plain paper copying
(people still talk of ‘Xeroxing’ papers to this day!).
‘The transition period saw a great deal of innovation activity from entrants and Xerox...
In the monopoly period, Xerox was a highly innovative firm. But the innovations were
characterised by being ‘in the copier.’...The innovations by entrants include some of a very
different character, those oriented toward the “user interface”...document feed devices,
two-sided copying, whole new product markets like the “convenience copier”.’6
Guidance on this activity can be found in the VLE.
150
Bresnahan, T. ‘Postentry competition in
the plain paper copier
market’, American
Economic Review 75
1985, pp.15–19.
6
Chapter 15: Research and development competition
15.4 Strategic issues in R&D
The previous model was less strategic in the sense that one player’s
behaviour depended on the other player’s actions – we simply derived
the willingness to pay for entrants and monopolists. In this section, we
will address some of the questions that might arise if multiple firms can
select certain aspects of their R&D strategy: the decision to be active in
a particular field and the riskiness of an R&D strategy. Throughout this
section, we will assume that players are risk-neutral, that is, they only care
about expected profits.
Let us first look at the question to be active in a particular application
field. Consider two firms who have to decide whether to engage in R&D
at a fixed cost K. If they engage in R&D, they have a likelihood of r to
come up with a successful (marketable) innovation. The profitability
of a successful innovation depends on the competition: an innovation
that is alone on the market generates profits of pm, an innovation with a
competitor in the market generates profits of pd.
From this we get the expected profits P for different strategy combinations
(Make sure you know how to derive these expected profits!).
Firm A does R&D, B does not:
P(R,N) = rpm + (1 – r)0 – K (and vice versa)
Both firms engage in R&D:
P(R,R) = r(1 – r)pm + r2pd + (1 – r)0 – K
Firm A does not do R&D:
P(N,R) = P(N,N) = 0
You should note that P(R,N) > P(R,R) – it is better to do R&D on your
own than with a rival competing for the same market (obvious really, but
you should be able to show it using the profit functions).
Putting this in a pay-off matrix, we get Figure 15.8.
Figure 15.8: A patent race.
Let us now derive the Nash equilibrium/–a for different values of r, the
likelihood of innovation. We fix the other parameters to the following
values:
K = 2, pm = 8, pd = 2
What if the likelihood of innovating is very low, say 10 per cent? Generally,
we would expect that innovation is not very profitable (because we have
to pay the fixed cost anyway, and the likelihood of having an innovation to
151
MN3119 Strategy
show for it is quite low) and firms are less inclined to innovate. This also
shows in our pay-off matrix, since pay-offs for (R,N) are –1.2 and pay-offs
for (R,R) are –1.26. Both are negative, which means that it is better not to
do any R&D at all, even if my rival does not do R&D either.
Similarly, if the likelihood is very high, say 90 per cent, we would
expect both firms to engage in R&D – the Nash equilibrium is (R,R). For
intermediate values of r, the situation becomes a little more complicated:
If the likelihood of success is r > 25 per cent, not doing R&D is not a
dominant strategy anymore. On the other hand, if it is r < 33 per cent,
doing R&D is not a dominant strategy. Putting in a number between 25
and 33 per cent will show that there are two Nash equilibria – (R,N)
and (N,R): if my rival is not developing, I will develop, but if my rival is
developing, I will choose not to develop.
Therefore, we now have a pretty good idea of what will happen to R&D
activity in relation to the likelihood of success. Low-probability projects
with high fixed costs will probably not be undertaken, medium probability
projects will be undertaken by whichever firm commits to the project first
(and thus pre-empting the others), or whenever two firms compete in a
medium-probability field, losses are likely (say both firms play a mixed
strategy and end up being active in the same field). High-probability
projects will be undertaken by many firms at once, which may turn out
to destroy parts of the profits to be made from the innovation in the first
place (it is likely that high-r projects will have two innovations and thus
competition in the product market, which of course is less profitable than
monopoly).
Activity 15.5
Using the pay-off functions and the pay-off matrix above, show what happens to the
equilibria as the following changes:
a. Market size increases by 10 per cent
b. Fixed cost decreases by 50 per cent
c. Competition becomes more fierce, lowering duopoly profits by 50 per cent
Guidance on this activity can be found in the VLE.
Now, frequently firms do not decide only if they want to be active in R&D,
but also how much risk to take. Taking risk typically comes with a tradeoff: on the one hand, there is a likelihood that I will come up with a better
innovation (or innovate faster) if I am successful, but on the other hand,
there is a lower probability of being successful in the first place. A stylised
situation might look like this:
Consider two competing firms having to choose between a safe and a risky
strategy. The safe strategy generates an innovation with certainty, the risky
one is successful with a probability of x per cent (< 100 per cent). Safe
and risky innovations both compete for a patent of value 1, and the risky
strategy is faster so that if successful, a risky strategy gets awarded the
patent. If both firms innovate at the same time, they have a 50 per cent
chance of being awarded the patent.
152
Chapter 15: Research and development competition
Figure 15.9: Risk taking in R&D.
The pay-off matrix shown in Figure 15.9 gives expected pay-offs for each
combination of strategies. Again, we can see that the equilibrium strategies
depend on the value of x. In particular, if x is high (>58 per cent), we find
that taking risks is a dominant strategy. From a societal standpoint, this
may not be the best outcome since there is now a risk of not seeing any
innovation as both players try and outpace the other, which includes a risk
of not having any successful innovation at all.
Consider two competing firms having to choose between a safe and a
risky strategy. The safe strategy generates an innovation with x per cent
likelihood, the risky one is successful with a probability of y per cent
(y < x). The willingness to pay by consumers for a safe innovation is A,
and B (B > A) for a risky innovation. If two innovations are of the same
quality, they compete against each other on prices (marginal cost is zero).
If they are of different quality, the higher-quality firm will be able to
charge a price equal to the quality differential A – B (why?).
Activity 15.6
Draw a pay-off matrix with the general parameters. For which parameter values will the
equilibrium be (R,R)?
Guidance on this activity can be found in the VLE.
One of the potential outcomes of an R&D race with firms choosing risky
strategies is that some firms turn out to be successful and some not. This
means that at any one point in time firms will be differently placed in
their position in the R&D race. Cabral’s (2003) paper on dynamic R&D
competition illustrates under which circumstances firms will choose the
risky or safe R&D strategies. We present a very simplified version here.
Suppose there are two firms competing for a pharmaceutical innovation.
The innovation process can be divided up into four steps, from finding the
compound to having it tested by the FDA (Food and Drug Administration).
Similar to our previous example, all that matters is getting to the finish
line (i.e. coming up with a patentable innovation) first, and if both
players arrive simultaneously, they both have a 50 per cent chance of
succeeding. One firm has already made some progress and is at stage 2
of the process, while the other is at stage 1. Both have two strategies in
each period: Safe R&D will ensure progressing one step further with 100
per cent probability, Risky R&D would propel the firm’s research two steps
ahead, but only with 50 per cent probability. Note that no one strategy is
inherently ‘better’ – they both have the same expected progress per period.
To keep things simple, suppose that the race only has two more periods
of play (this is actually not important and the game will end after two
153
MN3119 Strategy
periods anyway, but assuming it from the beginning makes things easier to
illustrate).
A player’s strategy for the entire game now consists of an action in period
1 and an action in period 2, so there are four options for each player: Safe
in 1 and Risky in 2 (sr), Risky in 1 and 2 (rr) and so on. This leaves us
with the rather daunting-looking pay-off matrix below.7
Figure 15.10: Risk taking in R&D.
This matrix looks relatively complicated to solve mechanically, that is,
by checking every single cell for a Nash equilibrium. In fact iterated
elimination of dominated strategies comes in extremely helpful here –
first we will be able to solve the game more quickly and second we will
discover some of the intuition behind the firms’ incentives to take R&D
risks. Below is an intuitive account of which strategies we can rule out, but
you should make sure that you are comfortable with the mechanics of the
process as well:
•
Follower playing (safe, safe): we can immediately rule this out because
F will never reach the finishing line with this strategy.
•
Follower playing (risky, risky): if the first-period gamble was
successful, F is one step away from the finish. All that matters is
getting there, not getting further than the finishing line. Taking a risk
in the second stage is taking an unnecessary risk.
•
Leader playing (safe, risky): same argument as the previous one – one
step away from the finishing line so you don’t want to take any more
risks.
•
Leader playing (risky, safe): if L was successful in the first period, the
game is over. If L gambled and failed, taking a safe step would mean
that L never gets to stage 4.
By eliminating these strategies, we obtain a much simpler matrix
(Figure 15.11).
154
The probabilities of
winning are derived
from the probabilities
of successful R&D and
getting there first and
winning or getting
there simultaneously
and winning with 50
per cent likelihood.
If you feel you need
some practice on
combinatorial statistics,
try retracing these
probabilities.
7
Chapter 15: Research and development competition
Figure 15.11: Risk taking in R&D.
In fact, from inspecting this simplified matrix, we can even see that (risky,
risky) is dominated for L, meaning that L’s dominant strategy is to play it
safe in both periods, while F will have to try and catch up at some point by
taking a risk. This carries an important implication:
A firm lagging behind in an innovation race will choose riskier research
strategies than a leader.
Of course, this may also lead to an outcome where the lagging firm fails to
catch up, ends up taking higher and higher risks and finally falls behind too
far and gives up – leading to monopolised high-tech markets. Therefore, it
could be argued that even though some high-tech firms do not even have
potential rivals (which, as we have seen in the earlier part of this chapter,
can generate an incentive to engage in R&D), they are only in this position
because they have beaten a number of rivals in the R&D stage and are now
reaping their reward for their superior research performance.
15.5 Some further thoughts on R&D
Sleeping patents. Another strategy that firms often use is creating a thicket
of ‘sleeping patents’. These are patents that are relatively minor and
that the firm has no intention of ever commercialising. Why would firms
do that? After all, patents are preceded by R&D efforts and have to go
through an administrative procedure, lawyers are involved and so on. It
turns out the incentive is strategic: by patenting an innovation sufficiently
‘close’ to one’s main product, a firm is foreclosing would-be competitors
from inventing a close rival to one’s own product. In fact, Xerox has been
found guilty of this practice and was subsequently forced to license its
sleeping patents to competitors because this practice was found to be
hindering technological progress in the copier market. It is not surprising,
however, that most firms will try and create a blanket of patents around
their main product, which means that again new entrants would be more
likely to come up with drastic innovations simply because they are not
protected by the incumbent’s intellectual property.
R&D spillovers. Spillovers can originate from several sources:
employees move from one firm to the other,8 knowledge becomes public
because insiders divulge their knowledge, either legally (through academic
publications for example) or illegally (through industrial espionage), or
firms reverse engineer each others’ products, meaning that they can at
least retrace some of the steps that went into an innovation.
Formally, spillovers are often expressed as follows:
(R&D) = f(X1, bX2),
P1
8
Microsoft has hired
a large number of
Nintendo developers to
write games for its nextgeneration Xbox. Part
of the motive may be to
gain some knowledge of
the development process
at Nintendo.
where b = strength of spillovers and X = R&D effort.
155
MN3119 Strategy
That is, the benefits from doing R&D depend partly on your own R&D
effort (X1), but also on your rivals’ efforts (X2). Since spillovers are rarely
complete (in that everything you learn will be available to others), we use
a parameter b to capture the spillover effect – high b means high spillovers
and R&D is an almost public good, low b means that knowledge generated
stays more or less in-house and is a private good. Studies have shown
that typically this parameter is in the order of xx, meaning that 1 hour of
someone else’s research time is as productive to your firm as xx minutes of
your own time (and you don’t pay for it!).
R&D cooperation. Antitrust authorities have often been more lenient
about allowing collaboration by rivals in research and development than
in product market competition. Why is that? Implicitly, it is recognising
the existence and detrimental effect of spillovers: if spillovers are high,
we are facing a free-rider problem (as mentioned early on in the chapter).
Letting firms cooperate at this stage means that overall more R&D may
be done, which is generally considered a good thing. On the other hand,
if spillovers are very low, R&D competition may be mutually destructive
since double (or even more than double) the resources are spent if two
firms compete for the same innovation. Letting firms cooperate may be
a way of reducing this duplication of R&D efforts. Of course, one of the
challenges is to make sure that firms that collaborate in the R&D stage do
not collude in the product market stage.
15.6 Key concepts
•
Basic research
•
Applied research
•
Product versus process innovation
•
Drastic versus incremental innovation
•
Replacement effect
•
Efficiency effect
•
R&D risk choice
•
Patenting
•
Sleeping patents
•
Spillovers.
•
Cooperation
15.7 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
156
•
discuss various types of innovation
•
explain the concept of replacement and efficiency effect
•
assess the impact of market structures on R&D activities.
Chapter 15: Research and development competition
15.8 Sample examination questions
1. It has been said that internal and external R&D are complements, that
is, firms are better off doing both simultaneously. Why?
2. The wide-body aircraft industry was a triopoly for a while and
eventually evolved into a monopoly (see graph).
100
B747
DC10
L1011
Market share (%)
80
60
40
20
0
65
70
75
Year
80
85
90
Figure 1: Market share development in the wide-body aircraft market.
Show and discuss how the willingness to pay for an innovation (i.e.
the incentives to innovate) changed through the earlier to later stages
in the industry’s evolution.
3. Since the 1990s, the market has developed into a duopoly between
Airbus and Boeing and innovation opportunities in the classical
wide-body market have been limited. Consequently, Boeing started
developing the ‘Sonic Cruiser’, a 250-seater plane that could fly
close to the speed of light, and Airbus launched the ‘Superjumbo’,
a 550+-seater. Demand for both innovations was uncertain and
development costs high, and it was assumed that the two planes would
cover different market segments.
Using the two stylised representations on the supersonic and highcapacity market below, outline the conditions on development costs
under which it was sensible (i.e. an equilibrium strategy) for both
firms to locate in different market segments.
Figure 2: Normal form representation of ‘Supersonic’ and ‘Superjumbo’ market
157
MN3119 Strategy
4. If a firm believes that it is the only player in a technological field, it
may engage in basic R&D in this particular field. On the other hand, if
there are many players, R&D tends to be much more applied. Why?
5. It is often said that Microsoft’s operating systems are not truly
innovative because they don’t push the boundaries of the state-ofthe art in operating systems. ‘Real’ innovation is likely to come in
the shape of other operating systems (e.g. Linux). Outline potential
reasons for this.
Guidance on answering these Sample examination questions can be found
on the VLE.
Extended activity: discovering DNA
Read the following and answer the questions at the end.
The race for the prize – a case study on the discovery of the structure of
DNA.
The moment that James Watson and Francis Crick walked into the Eagle,
their local pub in 1953 and declared that they had ‘discovered the secret of
life’, they indelibly etched their names in the history books of science and
medicine. Their discovery of the structure of DNA has been widely hailed
as the most important scientific discovery of the last 50 years. Indeed as
time passes and more discoveries grow out of their original findings, it
may simply become one of the greatest discoveries of all time.
Background
DNA (deoxyribonucleic acid) is a large molecule with the complete set of
instructions for making all the proteins a cell needs. DNA consists of two
spiralling strands of millions of chemical building blocks or base pairs.
DNA is the genetic blueprint that determines such features as body height
and skin colour and allows the transmission of genetic information from
one generation to the next.
What is so well known now, that DNA carries the genetic matter of our
forefathers and in fact the secret of how human beings develop from
conception, was shrouded in mystery for centuries. Many theories were
expounded – one theory was that sperm from the male of the species
contained tiny humans that grew into babies in the womb. Another theory
suggested that the blood of the mother contained the features of the child
and this gave the child the nature that it eventually would have. Many
other such theories abounded and as bizarre as these may sound, they
were common currency in days past. And it should not be assumed that
these were archaic ideas – the realisation that DNA was indeed the bearer
of life, only came about in the 1940s and 50s.
The people involved
Linus Pauling was the most eminent chemist of his generation. He was
rightly nominated for the Nobel Prize in chemistry.1 This gave him
enormous clout around the world and in his own institution, CalTech
where he had significant lab resources at his disposal. Pauling was also
known for his intensity and drive when he had a problem he wanted to
solve, and would work rapidly and efficiently towards the solution. All
this coupled with his encyclopaedic knowledge of chemistry and solid
grounding in physics gave him the aura of the man to beat. It was widely
perceived that if Pauling was interested in DNA, then he would be the first
to unravel its structure.
158
1
Pauling did go on to
win the Nobel Prize
in Chemistry in 1954.
Indeed he also won
the prize for Peace in
1962 making him the
only person to win two
unshared Nobel Prizes.
Chapter 15: Research and development competition
Perhaps the greatest weakness that Linus Pauling possessed stemmed
also from his great strength, his intellect. It was generally acknowledged
that Pauling possessed a restlessness which meant that he was only ever
content when working on a dozen or so projects simultaneously. He would
always have trouble focusing all his attention on just one project.
Rosalind Franklin and Maurice Wilkins were researchers at King’s
College, London. Rosalind Franklin was an elite researcher and her arrival
was considered to be of great value to the research being done at King’s.
Maurice Wilkins was highly respected in his field of X-ray diffusion.
X-ray diffusion was particularly vital in the search for the DNA structure
as it would give the basic shapes and connections that would reveal the
underlying structure. In addition, Wilkins was known to be a methodical
and rigorous researcher, vital skills that would be required to slowly tease
out the mysteries of DNA. In many ways, the team of Franklin and Wilkins
were best equipped for the race; their talents and skills were perfect for
the task they had set themselves. Unfortunately, Wilkins and Franklin
never truly gelled as a team. Franklin especially was seen as a lone worker
and although appreciated the skills of her compatriot, she did not trust
him enough. This inevitably meant that she would never truly harness his
significant talents and abilities and indeed support.
Among the many ways in which James Watson and Francis Crick
were different from the other researchers in their field, was the stark
fact that neither of them were particularly qualified or well versed in
the chemistry and physics that would be required for the task they had
ahead of them. For much of his time, Watson was bored with his academic
work and did not distinguish himself in any particular way. It is with
much irony that one of the two people who uncovered the structure of
DNA has written, ‘I was principally interested in birds…and managed
to avoid taking any chemistry or physics courses which looked of even
medium difficulty. Briefly the Indiana biochemists encouraged me to learn
organic chemistry, but after I used a Bunsen burner to warm up some
benzene,2 I was relieved from further true chemistry. It was safer to turn
out an uneducated PhD than to risk another explosion’. Crick had been
born in Northampton halfway through the First World War. He had won
a scholarship to a small public school and went on to University College
London. He also found the courses at university irrelevant and narrow,
graduating with a second class degree. However, his maverick mind and
his determination to do serious academic work in any subject led on to a
PhD working on the design of an apparatus for the study of water viscosity.
For the non-chemists,
benzene being highly
flammable is not
generally ‘warmed up’
and especially not with
the naked flame of a
Bunsen burner.
2
Both Watson and Crick have been called mavericks and perhaps this
more than any other factor was the root cause of their success. There was
irreverence in their manner, of a refusal to realise that the odds were well
and firmly stacked against them. They were both hungry for a problem
to solve that they felt worthy of their attention. The decoding of DNA
provided a worthy opponent.
There was an additional factor that was crucial to their success. Although
from very different backgrounds and of very different manner, Watson and
Crick hit it off immediately. In a telling excerpt from Crick’s biography, he
recounts his first meeting with Watson in the Cavendish Laboratories: ‘Jim
and I hit it off immediately, partly because our interests were astonishingly
similar and partly I suspect, because a certain youthful arrogance, a
ruthlessness, and an impatience with sloppy thinking came naturally to
both of us.’
159
MN3119 Strategy
Setting the scene
In the early 1950s, James Watson was something of a wanderer. He
had tried his hand at various disciplines ranging from bacteriology to
zoology to ornithology, the study of birds. He had spent time in academic
institutions in the US, Copenhagen and Naples, but had found little
satisfaction or enjoyment in his work. In Naples however, he found himself
in a lecture by the British biophysicist Maurice Wilkins.3 Maurice Wilkins
was describing a procedure to use X-ray diffraction on a substance called
DNA.
Watson had been pondering the question of how genetic information
might pass from generation to generation and how these mechanisms
worked. He had not progressed with this thinking, unsure whether it was
indeed DNA that carried this information and whether the structure of
DNA was simply too complex to decipher. Wilkins’ talk dispelled those
doubts because of his confidence in DNA being the conduit of genetic
information and the regularity of the patterns that his X-ray diffractions
were showing. Within weeks of the lecture, Watson had secured a position
with which to follow his new interests. He had managed to find himself
a position at the Cavendish Laboratory in Cambridge under the inventor
of the X-ray diffraction technique, Sir Lawrence Bragg, then the leading
authority on the subject.
3
Who would become
on half of the WilkinsFranklin team at King’s
College.
Bragg instinctively knew to pair Watson up with Crick. Crick’s brilliance
was well known, as was his brashness and unwillingness to suffer lesser
minds gladly and he would often rub his more old-fashioned English
colleagues the wrong way. Not for him were the formality and fustiness
that characterised the Cambridge of the 1950s. Bragg’s judgment in this
matter was perfect, putting together the impatient Englishman with the
confident and extrovert American. Had it not been for Bragg’s insight into
the personalities of both men, it is likely that Watson would have passed
through Cambridge in the same way as he had his previous institutions,
without vigour or any substantial result to speak of.
Crick’s path to the question of how information was passed through the
generations, was a similarly unusual one. When the Second World War
ended, he was preparing to return to his original PhD4 when he became
captivated by a book then recently published. What is life? was written
by the physicist Edwin Schrödinger, and it approached biology from the
perspective of quantum mechanics. Schrödinger was suggesting that the
mysteries of genetics might be resolved by applying the laws of quantum
theory and building a picture based upon atomic physics and the design
of molecular structures. Crick was hooked on the idea and the potential it
offered.
It is often the case that great discoverers have a ‘Road to Damascus’
experience that illuminates their thoughts and in so doing energises and
propels them to the solution of the problem. For Watson, it seems to have
been the lecture by Maurice Wilkins and for Crick, Schrödinger’s book.
Max Perutz, for whom Crick originally worked when joining the Cavendish
Laboratories, recalled that Crick and Watson were clearly exceptionally
brilliant and when they first met they were bursting with ideas and
energy, although between them they knew almost nothing of the subject
they wanted to master. More clearly remembered, however, was Watson’s
overwhelming desire for fame and fortune – to win the Nobel Prize. Crick
was always hungry for the root answer to the problem and this was his
motivation. Crick would often say after they had solved the problem,
that Watson never had any doubts about his ability and was preoccupied
160
4
Crick’s original PhD
topic, on the design of
an apparatus for the
study of water viscosity,
highlights with startling
clarity the giant leap
between disciplines that
he eventually made.
Chapter 15: Research and development competition
neither with the means nor the method he used to find the answer, but
was principally concerned with finding the answer ‘as quickly as possible’
in order to gain recognition and fame.
That James Watson and Francis Crick were unique is without doubt, more
so their methods. They conducted very few if any real experiments, instead
using ideas from a variety of seemingly disparate disciplines to build an
intellectual edifice from the work of others. They would eagerly borrow
ideas and carefully make links from various experiments that had been
conducted and slowly but surely piece together a picture of what DNA
would look like. Being unencumbered by any one discipline or school of
science, they took freely from all comers.
Watson and Crick had a sort of manifesto of how they would solve the
problem first. As Watson would say, their strategy was to imitate the
acknowledged and respected leader of the race and then beat him at his
own game. Watson and Crick put their efforts into keeping up with the
feverish pace of the great chemist himself – Linus Pauling.
Events unfolding
There were some key requirements in the race towards the structure of
DNA. One of them was most obviously the skills and knowledge of various
fields of science required to even know what to look for. The second
became more important as the work went on – an X-ray picture of DNA
would enable any researcher to start to have an insight into the DNA
molecule. Finally a model, either physical or mathematical, would be the
proof that the structure proposed was indeed coherent and plausible. Any
of the teams would realistically need all three to claim victory.
Initial situation
Linus Pauling started working on DNA at the same time as Watson and
Crick. The news of this would have generally defeated most young
scientists – Watson and Crick, however, were aware of the fact that if
they were to be beaten, it would be unlikely that they would lose any
face, indeed, most people would not even know that they had taken on
the great Linus Pauling. Additionally they were aware that the field was
relatively closed due to the very high investment required to get into it –
who was prepared to learn genetics, chemistry, physics, biochemistry and
X-ray diffusion?
As Wilkins and Franklin were clearly qualified for the task, Watson and
Crick were definitely in the worst position for the project.
Pauling also had a very poor quality X-ray diffraction picture of DNA which
was produced five years earlier. Crick, however, had obtained through his
old friend Maurice Wilkins, the latest superior images from Wilkins’ lab
at King’s College. This gave Watson and Crick a strong advantage against
the more skilled Pauling, however, their position relative to Franklin and
Wilkins was unclear.
Unexpected help
In October 1951, Crick invited Wilkins to visit Cambridge. In the small
confines of his flat, Crick, Watson, Wilkins and his wife discussed the
progress that Wilkins had made, especially with regard to the improved
images of DNA that Wilkins had managed to take. Eventually, Wilkins was
to become morose and share that the state of his working relationship with
Franklin had deteriorated to such an extent that she was not even sharing
her results with him. Instead she would spend long periods locked away in
161
MN3119 Strategy
her lab using the newer images that Wilkins had made, in isolation from
everyone else.
This conversation led to some important insights for Watson and Crick. One
was obviously that the King’s pair were not cooperating and this would
naturally slow them down. Second, that Franklin was still a long way from
a solution. Also they realised that Wilkins was willing to assist them in
their quest largely due to his poor relationship with Franklin. Perhaps most
importantly they learnt that Franklin would be delivering a seminar which
would sum up her year’s worth of research in the coming month.
Watson remembered the seminar as being a very dry affair. Franklin talked
through various numbers of the different types of chemical groups likely to
be in DNA, clues about the potential location of various molecules within
DNA and even the shape of the structure. There was a serious problem
though, much of the information was wasted as Watson, who never
took notes, found he had trouble remembering the finer details of what
Franklin had shared, indeed making some crucial mistakes in his recall.
For Crick it was very frustrating but they nonetheless began to construct a
model of what they thought DNA might look like.
Setbacks
After a week of slaving away they produced a structure constructed
around a central structure of three helices. This was essentially a guess –
Franklin in her seminar had suggested somewhere between two and four
helices. As soon as they had completed the model, they invited Wilkins to
come up and view it – they were surprised when he agreed and indeed
that Franklin was also planning to join him.
‘Where are the water molecules?’ the sarcasm in Franklin’s voice was
undisguised as she observed the model. Crick pointed them out only
to be told that there were nowhere near enough water molecules in
their model for it to be viable. With typical steeliness Franklin described
her visit to Cambridge as ‘a waste of time’ and in doing so effectively
humiliated and embarrassed Watson and Crick. Watson’s failure to make
notes in the earlier seminar had cost them dearly. Far worse, however,
was the umbrage that Franklin took to the fact that Watson and Crick
had attempted to build a model on the back of her research. She duly
complained to Lawrence Bragg, the supervisor of the Cavendish Institute.
The embarrassment of such an accusation caused Bragg to give both
Watson and Crick a severe dressing down. He then forbade them from
carrying out any more research on the topic.5 Bragg additionally insisted
they hand over their models to the King’s pair and that they stop any
further work they may otherwise have done.
This was a severe blow to Watson and Crick, perhaps the most significant
in their quest. Frustrated and angry, they had no option but to accept
Bragg’s demands. The closeknit community at Cavendish meant that it
would be impossible to carry on their work modelling DNA without it
getting back to Bragg.
The setback meant that it became impossible for them to go about
obviously working on DNA but they soon realised that they could keep
building mental models and talking to each other in private about their
progress. Indeed, they even continued borrowing from colleagues and
asking questions, albeit in a much more subtle and understated way. In
their usual watering hole, the Eagle, their questions were couched in a
language that implied that they were working on non-DNA related projects
but were always in fact directed towards that goal.
162
This episode highlights
what was then a
peculiarly British
research attitude.
Bragg and his peers in
powerful positions still
cherished the notion of
gentlemanly behaviour
and did not believe that
it was ethical for more
than one team to work
on a given problem,
even when the parties
were rival organisations.
5
Chapter 15: Research and development competition
That year, 1951, for all their mental efforts, was generally a bleak one and
progress was agonisingly slow.
Fortunate turn of events
In this period, known only to her research assistant, Rosalind Franklin was
making great strides forward. Early in 1952, Franklin had obtained the
clearest pictures of DNA that had ever been seen. Above all else it began to
confirm for the first time that DNA was helical in shape.
Picture 51
To the trained eye, ‘Picture 51’, with its broad dark X crossing the centre
of the image indicated that the structure of DNA was indeed helical. This
was a crucial step forward for Franklin and put her in the best position to
progress.
Franklin told no one of her results, not even Wilkins. In fact she even went
so far as to ridicule his beliefs of helical-shaped DNA by making a public
announcement that she had found proof that it was not helical – even
though she held the only proof that it was indeed helical.
Wilkins was livid and especially after he realised that the announcement
had been founded on a lie. Wilkins had been surreptitiously duplicating
Franklin’s work and had the same stunning images to prove that DNA was
indeed a helix.
Spring 1952 came round but brought nothing to lighten the mood at
Cambridge. Watson and Crick had not made progress and had not heard
anymore of Linus Pauling’s progress. They feared the great man’s powers
and were expecting a surprise paper that would shatter their hopes.
Throughout this time that had assumed that the work at King’s was also
not progressing.
Their worst fears were confirmed when Pauling was due to visit the UK
to attend a meeting of the Royal Society. They thought that it would be
inevitable that he would then meet up with the researchers at King’s, see the
X-ray pictures and thereby remove the slender lead they felt they held over
him. Convinced the battle would soon be over their spirits sunk to new lows.
On the eve of Pauling’s visit they heard shocking news. Pauling’s passport
had been withdrawn by US officials fearing that he was too threatening to
the capitalist cause and even a potential defector to the Russian side. He
had the indignity of being forced to return to California, potentially as a
Russian defector.
This news was greeted with outrage by the British scientific community.
For Watson and Crick, however, it was taken as an omen and they found
themselves even more determined to crack the DNA problem.
Their efforts over the summer of 1952 quickly bore fruit, in stark contrast
with the previous winter. One evening in the Eagle, Crick was to make
one of the great breakthroughs in the race. In discussion with John
Griffith, a mathematics postgraduate who was applying mathematics to
biochemical problems, he told Crick that there was only one consistent
arrangement for bases that was plausible for DNA to be bonded together.
The proposed structure with the bases in that arrangement meant that
two strands of DNA could carry enormous amounts of information. In a
moment of dazzling insight, Crick realised that because of the way the
bases bonded, the two helices of DNA could unzip, carry the code and pass
on instructions during reproduction. The secret of life. It was a giant leap
forward.
163
MN3119 Strategy
Pauling’s solution
In December of 1952, Watson and Pauling were informed through a
letter to Pauling’s son who was at Cambridge that he had outlined what
he thought to be a promising structure for DNA. No details were given
beyond to say that he hoped to publish early in the New Year.
In February 1953, the draft paper arrived on their desks. It was a moment
of intense despair and apprehension at learning that all was soon to be
lost. It was with trepidation that they began to read.
They were staggered. The most accomplished mind in chemistry had
proposed a structure that Watson and Crick knew to be impossible. He
had proposed a triple helix along with many other elementary errors. The
structure was too tightly packed, the chemical groups would not fit into
the space they had been allotted, the bonding structure was completely
wrong. Above all, the structure had nothing to say about how genetic
matter passed between generations. It was a stunning revelation that the
Pauling that they had so feared had got it so wrong.
Watson later declared, ‘The appropriate emotion was pleasure that a giant
had forgotten elementary chemistry.’
This emotion quickly gave way to fear. They realised that once Pauling
realised his mistake (and perhaps he already had), the humiliation would
spur him into a burst of activity that would inevitably culminate in the
correct structure. They believed that Pauling could have the structure
within weeks.
Final burst
Flying in the face of Bragg’s authority, Watson made another visit to
King’s. He was going to attempt to persuade Wilkins to part with more
information, possibly even new X-ray diffraction images. Perhaps as a
reflection of the deteriorating relationship between them Wilkins handed
over the images that Franklin had made some six months earlier, crucially
including a replica of Picture 51.
‘The instant I saw the picture my mouth fell open and my pulse began to
race’, Watson recalled.
This picture allowed them to leapfrog ahead of the other two teams, giving
them a slender but significant lead.
For Watson and Crick that data was everything they could have hoped for.
It told them that DNA was definitely helical but it also gave them crucial
information about the form of the helix and could allow them to calculate
the way the chains of the helix were arranged and how far apart the turns
of the helix would be. They both realised that this was the vital clue to the
structure. Combined with their ideas, it would allow them to elucidate the
structure.
They had a problem – time was not in their favour and working in a
clandestine manner as they had been would slow them up considerably.
They decided to come clean and tell Bragg what they were up to.
Feeling extremely pessimistic, they came to the interview expecting a stern
dressing down. To their astonishment, their boss’s attitude was starkly
different from before. Having learnt that Pauling was so clearly in the race,
ethical codes and the need for gentlemanly conduct vaporised. Bragg and
Pauling were old rivals. Pauling had previously succeeded in overtaking
Bragg in the determination of the alpha-helix for proteins, showing up
Bragg’s efforts with an elegant and convincing structure and clearly
164
Chapter 15: Research and development competition
exerting his superiority. Now facing the spectre of Pauling gaining the
greatest biochemical prize in history, his sense of ethics was ditched.
Bragg’s approval was absolute. Not only did he give Watson and Crick
moral support, he provided them with the full resources of the Cavendish
Lab, including the workshop to fashion model parts for them and X-ray
diffraction equipment along with technical support.
During this period Franklin had decided to leave King’s and was serving
out her final weeks. This did not matter to her as she was planning
to continue to work on elucidating the structure wherever she went.
However, it did mean that her work with Wilkins came to a total stop
during this period as her leaving was largely due to the frictions between
them. She hence had no inkling that Wilkins had passed on to Crick and
Watson the crucial information from her experiments.
It is now known from looking at the notebook she kept over that period
that she had decided that the double helix was indeed the correct structure
and was working on the mathematical analysis to establish a clear proof of
the double helix backbone.
Watson and Crick were at this stage racing ahead. They had constructed
a model of the basic double spiral but were still struggling to position
the bases of the molecule in a chemically consistent way. In a casual
conversation with Jerry Donahue, a colleague in the chemistry department
who specialised in organic bases, Watson described the problems he and
Crick were having. Donahue immediately spotted their problem – they
had been using the structure of bases that were accepted as standard in
chemistry textbooks but had recently been shown to be unlikely to occur in
nature. Donahue proposed the structure of the bases that recent research
had shown to be naturally occurring. This made all the difference to
Watson and Crick as this now allowed the structure to fit together and the
bonds to be of the right length within the backbone structure of the helix.
Franklin had mathematically and independently come to the same
conclusions. Certain crucial aspects still eluded her – however, it is almost
certain that in due course she would have cracked these also. Her notes of
that period certainly pointed in that direction.
Time, however, was something neither she nor Pauling were to have.
A few days after discovering the correct bases, the final pieces of the
puzzle were falling into place for Watson and Crick. They had a logically
consistent theoretical model in which the four bases were fitted into the
interior of the helices and bonded correctly together. They then ordered
the parts from the workshop and began building their model.
A week later on 7 March, their model was completed.
Conclusion
On an overcast spring morning Linus Pauling is chauffeured to the
Cavendish Laboratories. He is met by Sir Lawrence Bragg. Pauling is on his
way to the Solvay Conference in Brussels and had been asked to make a
detour to inspect the model that Watson and Crick have constructed.
For Watson and Crick this is the definitive test along the road to
acceptance of their work. All the researchers at the Cavendish Labs,
Rosalind Franklin and Maurice Wilkins have inspected the structure and
conceded that it was indeed the representation of DNA.
Pauling paced the floor in front of the model, peering closely at all its
details. It was not pretty to look at by any means. But it consistently and
165
MN3119 Strategy
logically showed the bonds between chemicals and displayed how the
complex code of the living beings was held between its helices.
Perhaps more than any other, Pauling appreciated this. A few days later at
the Solvay conference, Linus Pauling beaten to the most important finding
of twentieth-century biology, but magnanimous in defeat, announced
to the gathering: ‘I think that the formation of the structure by Crick
and Watson may turn out to be the greatest development in the field of
molecular genetics in recent years’. It was a prescient comment, and an
understatement.
Wilkins and Franklin were also magnanimous in defeat and asked only
that they were allowed also to publish their contribution to the elucidation
of the structure when Watson and Crick published theirs. The 25th April
1953 issue of Nature, contained not one but three pieces that announced
the elucidation of DNA – one by Wilkins, one by Franklin and the seminal
piece by Watson and Crick.
Epilogue
An extract of the speech by Professor A. Engström, member of the Staff of
Professors of the Royal Caroline Institute when awarding the Noble Prize
for Physiology or Medicine in 1962 read as follows.
‘Your discovery of the molecular structure of the deoxyribonucleic acid,
the substance carrying the heredity, is of utmost importance for our
understanding of one of the most vital biological processes. Practically
all the scientific disciplines in the life sciences have felt the great impact
of your discovery. The formulation of double helical structure of the
deoxyribonucleic acid with the specific pairing of the organic bases, opens
the most spectacular possibilities for the unravelling of the details of the
control and transfer of genetic information.’
An achievement by any standards, but especially for two wanderers.
The Nobel award for Physiology or Medicine that year was awarded to
Francis Crick, James Watson and Maurice Wilkins, acknowledging the
work all three had done. Rosalind Franklin was omitted necessarily as the
Nobel Committee does not make posthumous awards.
All three scientists continued to work within the field.
Rosalind Franklin died in 1958, at the age of 37, through ovarian cancer.
It is thought likely that the cancer was brought on in part due to the
excessive exposure she had to X-rays in the name of her research –
research that was ultimately crucial in finding the structure of DNA.
Source
Cheryan, D. under the supervision of Dr T. Kretschmer ‘The race for the prize – a
case study on the discovery of the structure of DNA’, LSE.
Questions
1. The discovery of DNA had the character of an R&D race. Using
concepts of R&D efficiency, R&D costs (including opportunity cost) and
the riskiness of R&D, introduce the three players (or teams) and their
chances of success.
2. Of all the three players, Linus Pauling seemed best equipped to win
the race if he set his mind to it. How do you think the other two teams
changed their strategies in response to Pauling? Why? Explain using a
game-theoretical framework.
166
Chapter 15: Research and development competition
3. Pick one episode from the case where a new discovery was made/
the situation changed significantly/etc. Describe the players’ initial
situation and justify how the players reacted to the new situation and
the expected responses to the other players.
167
MN3119 Strategy
Notes
168
Chapter 16: Technology adoption
Chapter 16: Technology adoption
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
describe technology adoption by building game-theoretic models
•
discriminate between and explain different models of technology
diffusion.
Essential reading
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) Chapter 14.
Further reading
Bryson, A., R. Gomez and T. Kretschmer (2005) ‘Catching a wave: the adoption
of voice and high-commitment workplace practices in Britain’, 1984–98. CEP
Discussion Paper DP 0676. http://cep.lse.ac.uk/pubs/abstract.asp?index=2154
Fudenberg, D. and J. Tirole ‘Preemption and rent equalization in the adoption of
new technology’, Review of Economic Studies 52(3) 1985, pp.383–402.
Geroski, P. ‘Models of technology diffusion’, Research Policy 29(4–5) 2000,
pp.603–25.
Griliches, Z ‘Hybrid corn: an exploration in the economics of technological change’,
Econometrica 1957. Reprinted in Griliches, Z. (ed.) Technology, education, and
productivity. (New York: Basil Blackwell, 1988), pp.27–52.
16.1 Introduction
Firms change their technologies all the time – an old one becomes obsolete,
a new one is introduced into the market, an existing production process
is inadequate for an improved product and so on. In this chapter we will
highlight some specific features of new technologies and the effects they
have on the adoption process.
16.2 Adoption dependence
Benefits from the adoption of new technologies will often depend on
previous and future adoption decisions of other users. For example, as more
users adopt, information about the technology increases – there are users’
reports, information gets out by word of mouth, and the simple fact that
plenty of others are using it may say something about the quality of the
technology. Also, a technology may get objectively better or cheaper as more
people have adopted it – the learning curve discussed in Chapter 7, or user
feedback may make adopting a technology that has been tried and tested
by others more attractive. Network effects are also an important feature of
many new technologies – adopting a technology with a large network of
users is more attractive because there are more people to share, say, software
with. We will discuss this in the next chapter. All these phenomena suggest
that it is more beneficial to adopt a new technology late, but counteracting
this are first-mover advantages that might originate from brand building,
occupying crucial market space and so on. First-mover advantages will make
it more attractive for users to get in early rather than wait.
169
MN3119 Strategy
For the analysis of technology it is important to distinguish between
individual and aggregate adoption dependence.1 As mentioned
above, the benefits from adoption often depend on other users’ decisions.
However, for practical purposes it is important to estimate whether one
user’s decision is going to make a large or a small impact on one’s own
utility. If the impact is large, game-theoretic analysis is likely to be useful –
after all, we want to accurately predict what other adopters will do if their
decision may dramatically change our pay-offs. On the other hand, if small
changes in the number of other users do not change pay-offs significantly,
game-theoretic analysis would be ‘analytical overkill’. After all, trying to
predict one player’s adoption decision would not make a big difference
so understanding the overall outcome and hence, does not make sense.
Situations like these are best analysed using diffusion studies, which we
will introduce in the second part of this chapter.
16.3 Strategic technology adoption – option value
Consider the following situation: a new technology has just been
introduced, and there are two firms considering adoption. The technology
can either be ‘Good’ or ‘Bad’ in the sense that their NPV net of adoption
cost is positive or negative. Both firms do not know if the technology is
‘Good’ or ‘Bad’, but if one of the firms adopts, the quality of the technology
becomes known to both players. The game has two periods, and both firms
can adopt in period 1, adopt in period 2 or not adopt at all. Waiting, that
is, adoption in period 2, is discounted by d < 1, the discount factor.
Let us split our analysis in two parts. First, we assume that the technology
is expected to have a negative NPV. That is, the weighted (by their
probability of occurring) average of ‘Good’ and ‘Bad’ is negative.
By backward induction, we can then conclude that if no player has
adopted in period 1, neither of the players will adopt in period 2. After
all, the technology is expected to yield negative pay-offs. Pay-offs then for
delaying adoption if both delay are (0,0). If firm 1 adopts in period 1, its
expected benefit will be P < 0. If firm 2 adopts in 1 and the technology
turns out to be successful, firm 1 will adopt in 2 at a pay-off of dpG. On
the other hand, if early adoption reveals a bad technology, firm 1 will not
adopt and obtain 0. This information can be combined to give the pay-off
matrix illustrated in Figure 16.1.
Figure 16.1: Technology adoption I.
170
1
You might wonder how
network effects and
adoption dependence
are related. We will
discuss network
effects in more detail
in the next chapter,
but basically adoption
dependence can
originate from network
effects as well as
learning, information
transmission and so on.
Chapter 16: Technology adoption
Activity 16.1
Show that the Nash equilibrium to the game above is (D, D).
Guidance on this activity can be found in the VLE.
The outcome of the game is that a new technology is not even tried out,
even though there is a likelihood of (p) that it is successful and gives a
positive NPV. Since on average, however, the technology is not profitable,
one might argue that ‘no adoption’ is not such a bad outcome. We now
analyse a game where the expected value of a technology is positive, i.e.
P > 0. This could be either because the likelihood of a technology being
‘Good’ is very high (p ≈ 1), or because the benefit from a ‘Good’ technology
is considerably higher than the risk of an unsuccessful one
(G > B).
The first result we can note is that if no adoption has taken place in period
1, firms 1 and 2 would still adopt the technology in period 2, although
they still do not know the technology’s quality and they have delayed
adoption. The pay-offs from late adoption under uncertainty are then dP,
and the rest of the game remains the same.
Figure 16.2: Technology adoption II.
The outcome of this game depends crucially on the value of the discount
factor. Suppose that the discount factor is close to 1, so that the cost of
waiting is very slow. It would seem intuitive that firms will wait and adopt
late in the hope that the other player might adopt early. On the other
hand, if d is very low, waiting is very costly, and a firm might be willing to
take the risk of adopting early to avoid the loss of waiting.
The following analysis shows that the dominant strategy for low d is to
adopt early and consequently the unique Nash equilibrium is (E,E), and
that there are two Nash equilibria (D,E) and (E,D) if d is high.
If
δpG < pG + (1 − p )B 
dominant strategy to adopt early

 (1 − p ) B 
i.e. δ < δ = 1 − 

 p G
If
δΠ < Π
δ > δ , pG > pG + (1 − p )B → NE = (D, E ), (E , D )
The analysis illustrates that there is a cut-off value of the discount factor
which determines the choice of strategy. We can see that for low values of
d, our intuition of early adoption is correct. On the other hand, for high
values of d, even though it is not very costly, there is still an incentive
171
MN3119 Strategy
to adopt early, which leaves two Nash equilibria (and possible mixed
strategies to solve the game).
The option value of waiting until more becomes known about a technology
may introduce a number of inefficiencies – either a technology is not even
tried out if the perceived risk of a failure is too high (as in the first game
we analysed), or adoption takes place late or firms play a waiting game in
the hope that the other firm will adopt and reveal information (as in the
second game when d is relatively high)
Activity 16.2
Consider the situation analysed above. How would the game and its equilibria change in
the following situations?
a. A new technology following the one currently on the market has been announced
and will represent an improvement of D over the current one.
b. Adopters are not homogeneous. A firm can be either a high-valuation adopter
with valuations G, B, or a low-valuation adopter with valutions g, b. The following
parameter restrictions apply: G > g > 0 > B > b.
Guidance on this activity can be found in the VLE.
16.4 Technology diffusion models
As mentioned above, if there is a larger number of adopters of a
technology, game-theoretic analysis becomes increasingly difficult and
increasingly less useful. There are two alternative approaches to dealing
with such situations. Either, the population of potential adopters is divided
into consumer groups that are assumed to act more or less the same,
which then brings back the possibility of game-theoretic analysis. On the
other hand, sometimes it makes more sense to look at aggregate adoption
figures, that is, the diffusion of a technology. The first approach borrows
heavily from the marketing and market research literature to identify the
different consumer groups. We will focus on the second approach and
introduce three different models of technology diffusion.2 First, however,
we document one of the most robust findings in the economics literature –
the s-curve of technological diffusion.
Starting with Griliches’ (1957) study on hybrid corn, numerous scholars
have identified s-shaped diffusion patterns in the economy (if they are
successful at all – the other regular ‘pattern’ is a technology not getting any
meaningful level of market penetration). New technologies such as mobile
phones and the compact disc have followed a similar pattern. Other
‘technologies’, such as high-performance workplace practices (Bryson,
Gomez and Kretschmer, 2005) follow a similar adoption path. The sheer
regularity of this pattern throws up two important questions. First, if
a new technology is good enough to eventually be adopted by a large
number of users, why is adoption not instant? In other words, why wait if
it is a good technology? The second puzzle is that these technologies are
seemingly very different, and the fact they all display the same diffusion
behaviour suggests that they have some common characteristic that drives
the diffusion process.
172
2
These models are
summarised in more
detail in Geroski (2000).
Hybrid corn
Chapter 16: Technology adoption
100
Per cent
80
Iowa
60
Wisconsin
Kentucky
Texas
Alabama
40
20
10
0
1932 ‘34
‘36
‘38
1940
‘42
‘44
‘46
‘48
1950
‘52
‘54
‘56
Years
Figure 16.3: Hybrid corn.
Mobile phones in France
as a %
70
60
60
62
55
47
50
40
30
24
20
10
11
5
0
Jan 1998
Jan 1997
Jan 1999
June 2000
June 2001
CD players in the UK
July 2002
June 2003
Figure 16.4: Mobile phones in France.
97
19
95
19
93
19
91
19
89
19
87
19
85
19
19
19
83
80
70
60
50
40
30
20
10
0
81
% of population
CD Diffusion UK
Figure 16.5: CD players in the UK.
As it turns out, there are four dominant models of technology diffusion
– the heterogeneity (or probit) approach, the epidemic approach, the
population approach model and the network effect approach. We will deal
with network effects in the next chapter and introduce the other three
now.
173
MN3119 Strategy
16.4.1 Heterogeneous adopters
The first approach is based on the notion that adopters are heterogeneous
in their propensity to adopt a new technology. If we are dealing with a
large number of adopters, it is likely that the distribution of tastes will
have one peak in the middle3 – most users will be neither complete
enthusiasts for a new technology, nor complete opponents of it, but will
be somewhere in between. If a technology now develops over time, either
by becoming better or cheaper with time, early adoptions will be made
by the enthusiasts, because everybody else will not find it worthwhile to
adopt yet. As we have said, enthusiasts are typically few and far between.
As it gets more attractive to adopt, more people will, and eventually the
mass market – a large ‘spike’ – will adopt. Mass market adoption implies
that diffusion speed is much faster than in the early stages. After the mass
market has adopted, all that is left are the small numbers of steadfast nonadopters. As the technology becomes even cheaper and more attractive, or
the existing technology
goes increasingly
outmodel)
of date, even those unwilling
Heterogeneous
Adopters
( Probit
users will eventually adopt, but at a slow rate, which explains the final
adoption slowdown before market saturation see (Figure 16.6).
Figure 16.6: Heterogeneous Adopters (Probit model).
So far, we simply assume that adopters are heterogeneous in their
propensity to adopt. What could be factors that determine the likelihood
of adoption?
174
•
The size of a firm can have an important effect on the propensity to
adopt. For example, if there is a fixed cost of adopting a technology
that reduces output costs, it will be more useful (and therefore
adopted earlier) by firms who are producing high quantities already
and therefore stand to save significantly from a cost-reducing
technology.
•
The indirect adoption cost may differ by firm. Indirect adoption costs
include the switching cost from an existing technology, cost of learning
to use the new technology or even search cost of identifying the best
new technology.
•
A firm’s absorptive capacity is the ability to cope with new
organisational realities and situations. It is plausible to assume that
‘smarter’ firms are more likely to adopt a new technology as the
adjustment period will be shorter and less costly.
•
Use of complementary technologies is another important determinant
of a firm’s incentives to adopt a new technology. For example, a firm
with an up-to-date computer system is likely to benefit more from a
new piece of software than a firm with an outdated system. Conversely,
This is called a
unimodal distribution.
3
Chapter 16: Technology adoption
if a firm is using other software applications that are complementary to
the existing generation of software, it will be less inclined to switch to
the new generation.
These are just a selection of factors that will determine firms’ incentives to
adopt. For every technology, it is important to identify the main drivers of
adoption and to be able to rank firms according to it.
Activity 16.3
What do you think are the main drivers of adopter heterogeneity for the following
technologies?
a. High-commitment workplace practices (for firms)4
b. Mobile telephony (for end consumers)
c. Computer aided design (CAD) (for architecture studios)
d. Airbag (for car manufacturers).
Guidance on this activity can be found in the VLE.
16.4.2 Epidemic model
High-commitment
workplace practices
cover HR practices like
job enrichment, job
rotation, incentive pay,
internal promotion,
etc. - effectively creating
a workplace in which
employees can learn and
improve continuously
and are rewarded for
good performance.
4
Knowledge about new technologies is often said to spread by ‘word of
mouth’. Can this generate an s-curve of diffusion? To simplify things,
suppose that everybody who hears about a new technology will adopt it.5
Everybody who has adopted the technology talks about it to a (randomly
selected) fraction a of the population. Of this fraction, some will have
heard about the new technology already and some will not.6
5
Geroski (2000) also
discussed a model in
which knowledge does
not immediately imply
adoption, but it does not
make a huge difference
in terms of the intuition.
This again generates an s-curve: In the early stages, there are very few
people who know about the new technology and ‘spread the word’. This
implies that only a small number of new users can be recruited in each
period. Once the number of existing adopters is higher, ‘word of mouth’
really starts to work – diffusion is very fast. This will only work, however,
as long as there are still enough ‘non-users’ in the population. If everybody
knows about the technology, it is increasingly likely that users will talk to
other users, thus slowing down diffusion.
6
The reason why this
model is called the
‘epidemic model’ is that
diseases spread in very
much the same way
throughout populations
- infected people come
into contact with
infected and uninfected
people and spread
the disease to the
uninfected.
Activity 16.4
There are N = 1001 potential users in a population and currently one of them is using the
internet. He talks to 10 people per day and tells them about the internet. Everybody who
hears about it adopts it immediately and talks to 10 people about it. How long will it take
for the internet to reach a market penetration of 25 per cent?
Guidance on this activity can be found in the VLE.
16.4.3 Population ecology model
The final model we discuss is that of population ecology. The difference
to the previous two is that the population ecology model is based on two
competing effects, the legitimisation and the competition effects.
Consider a technology that can potentially confer a competitive advantage
over non-users. In the early stages of the technology, there is considerable
uncertainty about its quality and potential adopters are reluctant to
take the risk, so that adoption is slow. As more adopters start using the
new technology, non-users grow more confident about the technology
and become more likely to use it as well – the technology becomes
legitimised. However, as more users adopt it, the competitive advantage
that can be derived from it shrinks since everybody is now using it.7 This
7
It may help to think
of the new technology
as a new market of
uncertain size - as more
adopters go into the
new market, it becomes
more competitive and
the potential profits
decrease.
175
MN3119 Strategy
is called competition effect, since competition in the new technology
decreases the profitability of the technology and consequently slows
down adoption. Again, and not surprisingly, we find that this generates an
s-curve, as at first the legitimisation effect causes diffusion to speed after
a slow start, while the competition effect slows down diffusion of a fully
legitimised technology.
Technology diffusion does not come in all shapes and sizes – in fact, the
only shape it seems to come in is an S. This does not necessarily mean,
however, that all new technologies follow the same diffusion process. We
have introduced three approaches to technology diffusion that all generate
such a pattern, but are based on very different premises. In the assessment
of real-life diffusion processes, it is important to judge which of these
processes is most important in the particular case, and what implications
can and should be drawn from this in terms of the levers to accelerate
diffusion.
Activity 16.5
The following technologies have elements of different diffusion models. Discuss the
relative importance of the models for each case and outline the diffusion process for each
of them.
a. E-commerce
b. Hybrid corn
c. The internet
d. Renewable energy sources.
Guidance on this activity can be found in the VLE.
16.5 Key concepts
•
Adoption dependence
•
Individual and aggregate adoption dependence
•
S-curve
•
Diffusion models
•
Heterogeneous adaptors
•
Epidemic model
•
Population ecology model
16.6 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
describe technology adoption by building game-theoretic models
•
discriminate between and explain different models of technology
diffusion.
16.7 Sample examination questions
1. The Zip disk drive was never particularly successful, even though it
was undoubtedly superior to the then industry standard, the 3.5 inch
disk. Using an option value model, how do you think the likelihood
that rewriteable CDs would come out relatively soon after the Zip disk
affected the Zip disk’s success?
176
Chapter 16: Technology adoption
2. Can the three models of technology also be used to explain
unsuccessful technologies? How would you explain a failed technology
using the heterogeneity approach, the epidemic approach and the
population ecology approach?
3. How would you try and distinguish between the three models of
diffusion if you were confronted with a real-life diffusion process?
4. Below are graphs of the diffusion of CD players and CDs (note that the
timeframe for CD sales is larger than CD player sales). It seems that
while CD player sales have been slowing down in the late 1980s, CD
sales growth shows no signs of subsiding. What could be reasons for
that?
Millions of CD players
2
1.5
1
0.5
0
84
85
86
87
88
89
90
Figure 16.7.1: CD player sales, 1984–1990.
140
Millions of CDs
120
100
80
60
40
20
0
83
84
85 86
87 88
89
90 91
92 93
94
Figure 16.7.2: CD sales, 1993–1994.
Guidance on answering these sample examination questions can be
found in the VLE.
Extended activity: the adoption of Botox
Read the following and answer the questions at the end.
Technology Adoption: Botox Population Ecology
The United Nation’s demographic projections show that, in the more
developed countries, populations across the board are proportionally
ageing, with an average of 32 per cent of the population aged over 60 in
2050 versus 20 per cent today.
Allergan, Inc. is a $2.2B global healthcare company that plans to
profit from this inescapable trend. Among its lines of business in the
pharmaceutical and biotechnology industries, the company specialises
in aesthetic-related pharmaceuticals and over-the-counter products
competing in the anti-ageing sector of the cosmetics market. In the US
alone, the demand for R&D based ‘cosmeceutical’ products is expected to
increase 11 per cent per year, the market reaching $7 billion in 2008.
177
MN3119 Strategy
Based in California, Allergan was arguably optimally positioned for the
launch of cosmetic applications of its star product Botox, a treatment
based on injections of a purified form of the toxin botulinum, which blocks
nerve impulses to muscles. The notion of routinely absorbing a neurotoxin
might raise some common sense misgivings, and clinical studies are based
only on short-term data and so cannot rule out the possibility of long-term
detrimental health effects. At upwards of $300 a session every three to
four months, the treatment isn’t cheap. Still, from a market perspective,
the technology has been a success story to date. Everybody’s doing it:
the treatment is now practically de rigueur among the Hollywood set,
where physical beauty rivals talent and personal contacts as the ultimate
currency; and from Kansas to Korea, women as young as 30 appear to be
following suit.
Botox was approved for cosmetic use in the US in 2002, although
considerable off-label use for the treatment of facial wrinkles occurred
before that time, driving sales to $300 million by 2001. By 2003,
total Botox sales reached over $500 million, of which 40 per cent was
attributable to its cosmetic application. In 2004, it was cited as becoming
America’s fastest growing cosmetics procedure. In 2005 international sales
for all applications (medical and cosmetic) of Botox reached more than
$800 million. By 2008 analysts predict sales from the cosmetic applications
alone to reach $650 million. The patented technology is now Allergan’s
main growth driver, delivering a near monopoly in the world market for
the unique toxin. The company has recently announced a deal with Glaxo
to develop and promote Botox in China and Japan. While Botox sales will
continue to grow, the chart below shows how the rate of growth in the
numbers of people adopting this new technology has peaked.
A 2004 lawsuit brought unsuccessfully against Allergan by Irena Medavoy,
the wife of a Hollywood executive, shined a spotlight onto the safety issues
raised by the use of Botox. It also brought to light the fact that Allergan
has paid Arnold W. Klein, a top celebrity dermatologist, nearly half a
million US$ to promote the drug. One year after Botox’s market release,
Klein proudly claimed to have personally injected the product into 90,000
patients.
Does Botox have the final word in the quest for eternal youth? Vogue
magazine has already likened the ‘frozen’ Botox face to frosted hair and
too-tight facelifts as examples of commonplace beauty fixes gone awry.
Some market analysts predict the rapid acceleration of the market for
dermal fillers such as the Swedish Restalyne, which lasts longer and
appears to have fewer health risks. Distributors in this segment are
already nudging along market acceptance through clever marketing and
loyalty schemes that reward customers handsomely for adopting this
new technology as part of their beauty regimen. Physician-advocates and
marketers of both Botox and Restalyne seem to have consensus in the view
that Botox is best for treatment of the upper face, while Restalyne is best
for the lower – a convenient supposition that creates two markets from
one face.
Sources
‘Allergan, Inc. thinking ahead in Botox Glaxo Development Deal’, JP Morgan North
America Equity Research, 3 October 2005.
‘Botox getting bigger’, Global Cosmetic Industry, Allured Publishing Corp. 173(6)
2005.
‘Cosmetic dermatology, beyond Botox – the expanding market for dermal fillers’,
Deutsche Bank, 27 September 2004.
178
Chapter 16: Technology adoption
Gellene, D. ‘“Mr Botox” case raises some brows’, Los Angeles Times 22 August 2004.
Gentry, C. ‘Injection for perfection’, Tampa Tribune 3 October 2005.
White, J. ‘Fill in the bl’, The Kansas City Star 16 May 2004.
Question
Interpret the diffusion of Botox as a population ecology case. Calculate a
diffusion curve given the data below, assuming a starting point of 787,000
patients receiving Botox in the year 2000.
Year
YoY growth rate in number of patients receiving Botox
2000E
2001E
9%
2002E
31%
2003E
157%
2004E
27%
2005E
22%
2006E
18%
2007E
13%
2008E
10%
Figure 1: Historic and projected growth in numbers of patients receiving Botox
from plastic surgeons.
179
MN3119 Strategy
Notes
180
Chapter 17: Network effects
Chapter 17: Network effects
Learning outcomes
By the end of this chapter, and having completed the Essential reading and
Activities, you should be able to:
•
define network effects and network market structures
•
explain technology diffusion patterns using network theory
•
discuss the four generic strategies firms can pursue in network markets
as described by Shapiro and Varian (1999).
Essential reading
Cabral, L.M.B. Introduction to Industrial Organization. (Cambridge, MA: MIT Press,
2000) Chapter 17.
Further reading
David, P. ‘Clio and the economics of QWERTY’ in The American Economic Review
75(2) 1985, pp.332–37.
Farrell, J. and G. Saloner ‘Standardization, compatibility, and innovation’, Rand
Journal of Economics 16(1) 1985, pp.70–83.
Katz, M and C. Shapiro ‘Systems competition and network effects’, Journal of
Economic Perspectives 8(2) 1994, pp.93–115.
Koski, H and T. Kretschmer ‘Survey on competing in network industries: firm
strategies, market outcomes and policy implications’, Journal of Industry,
Competition and Trade 4(1) 2004, pp.5–31.
Ohashi, H. ‘The role of network effects in the USVCR market, 1978–86’, Journal of
Economics and Management Strategy 12(4) 2003, pp.447–94.
Postrel, S. ‘Competing networks and proprietary standards: the case of
quadraphonic sound’, Journal of Industrial Economics 39(2) 1990, pp.169–85.
Saloner, G., A. Shepard and J. Podolny Strategic Management. (New Jersey: Wiley,
2006) Chapter 12.
Shapiro, C. and H. Varian Information rules. (Cambridge, MA: HBS Press, 1999)
Chapter 7.
17.1 Introduction
In earlier chapters, we referred to network effect quite a few times. This
chapter will go into more detail and look at some of the aspects specific
to markets with strong network effects. To start with, however, we need
to define network effect in general, and distinguish between two different
types, direct and indirect network effects.1
Network effects – definition
A product is said to have network effects if the net benefits from buying and using the
product increase with the number of other users using the same.
Direct network effects originate from the ability of consumers to directly communicate
with other users of the same network, indirect network effects originate from an
increased incentive for providers of complementary goods to produce more or a wider
variety of them.
The literature on
network effects is vast
and we can only hope
to cover a relatively
restricted range of
topics. For an extensive
review of the recent
literature, see Koski and
Kretschmer (2004).
1
181
MN3119 Strategy
The following figures illustrate the mechanics of direct and indirect
network effects. In Figure 17.1, we can see that as another user joins
the network, the previous two users directly benefit because they can
communicate with one more user, which they have not been able to
previously.
Figure 17.1: Direct network effects.
In Figure 17.2, network effects materialise because a complementary good
producer has more incentive to produce for a larger network, which in
turn makes it more attractive to join a larger network.
Figure 17.2: Indirect network effects.
Activity 17.1
Which of these products have direct network effects? And indirect?
a. Languages
b. Telephones
c. Diesel cars
d. Video games.
Guidance on this activity can be found in the VLE.
17.2 Network market structures
The general notion of network effects is that a larger network of users
is more attractive for new users to join. What does this imply for market
structure in industries with significant network effects? We have a look at
a number of industries with network effects.
17.2.1 Video cassette recorders (VCRs)
Figure 17.3 illustrates market shares of VHS and Betamax in the US.2
In the late 1970s both standards were relatively close in terms of their
market shares, but VHS kept on pulling away, so that by the late 1980s
Betamax virtually disappeared from the market. Some of the main reasons
for VHS’s success was said to be its liberal licensing policy and the support
of the largest video rental store, Blockbusters.
182
2
Source: Ohashi (2003).
Chapter 17: Network effects
Market shares VHS, Beta
100
80
60
VHS
40
BETA
20
0
78
79
80
81
82
83
84
85
86
Figure 17.3: VCR by format, 78–86, US.
Other successful network technologies include the compact disc, FM
stereo radio, eBay auctions, and so on. For almost every success story,
however, there is also a story of dismal failure. Quadraphonic sound was
a technology that set out to replace stereo in the 1970s, but never caught
on,3 the DAT tape never managed to replace the compact cassette, the
Dvorak keyboard layout never displaced the ubiquitous QWERTY layout.4
These are just a small number of spectacular failures of technologies that,
while some claim they were not significantly better than the successful
ones, they were certainly not much worse than them. One of the main
points of interest about network markets therefore is how and why
technologies fail, and what firms can do to avoid this.
17.2.2 Excess inertia and excess momentum5
One of the reasons for the failure of perfectly acceptable technologies with
network effects is that the incumbent technology is superior due to their
installed base, which provides users of the incumbent technology with
a higher utility than switching to the new technology. It is quite possible
that they would prefer the new technology if everybody else switched, but
the uncertainty about the switching behaviour of others holds back all but
the most enthusiastic adopters of the new technology. Therefore, if a new
technology has network effects, is up against an incumbent technology with
a significant installed base, and there is some uncertainty about whether
other adopters will switch as well, a new technology may suffer from
excess inertia, the socially inefficient failure to adopt a new technology.
3
See Postrel (1990).
4
See David (1985).
This section is based on
Cabral (2000) Chapter
17.
5
Another scenario that might emerge with network effects is the opposite,
where the user of an existing technology feels he has to conform to other
users who have switched to a new technology, even though the incumbent
technology is still perfectly up to date. So if a technology has network
effects, and early users have a strong impact on late users
through the need for compatibility, excess momentum could emerge,
the socially inefficient adoption of a new technology.
17.3 Technology diffusion with network effects
In the previous chapter, we discussed three approaches to technology
diffusion – the epidemic, heterogeneity and population ecology approach.
We now introduce network effects to a story of technology diffusion.
Suppose a population of potential users is considering adopting the
compact disc. Compact discs are a network product with direct and
indirect network effects, so we would expect an early user to obtain
relatively low benefit from buying a CD player, while later users, who have
183
MN3119 Strategy
access to a large library of CDs, will derive more benefit. Assuming then
that adopters could do something else with their money if they did not buy
a CD player, every potential user has an incentive to wait and see if the
CD catches on. We can see immediately that there are two Nash equilibria
among the many potential CD users: either nobody adopts because nobody
is expected to adopt, or everybody adopts because everybody is expected
to adopt. Which one of these equilibria will emerge is obviously of prime
importance to sponsors of the technology, but also to users who are
convinced of the value of a new audio system, but would not want to take
the risk on their own.
If we take the intuition of this story into a dynamic setting, we will find
that for low expectations sales will fall short of expected sales, starting a
vicious cycle that will end in no sales of the new technology, and that for
high expectations sales will exceed expected sales, which in turn generates
a virtuous cycle that ends in market success. By continuity, there will be a
level of expectations below which the long-term equilibrium will involve
zero sales, and above which it involves full market penetration. This
threshold level of expectations is called critical mass.
Let us look at a numerical example. Suppose that there are N potential
adopters in each period and that demand in each period depends on the
price charged (p) and the expected demand (Ne) in the following way:
N = 1 – p + 2Ne
Apart from the last term, this is a well-behaved demand function as
demand decreases the price of the product. On the other hand, the higher
the expectations are for the product, the more consumers are willing to
pay for the product, which captures the assumption of network effects.
We look purely at the demand side to begin with. If expectations are
updated in some way so that the new expectations (i.e. expected sales in
the next period) are some kind of combination of last period’s expectations
and last period’s sales, for every price p there will be a point for which
actual sales exceed expected sales, which causes expectations to be
adjusted upwards and a self-reinforcing process to emerge.
Activity 17.2
Find the critical mass point for p = 3/2, p = 7/4 and p = 1. How does critical mass
depend on prices?
Guidance on this activity can be found in the VLE.
We now introduce the supply side of a network technology. Suppose for
example that a technology has marginal cost of MC = 1/2 and that a
review article states that expected sales of the technology are 1/3.6 At
what price should the product be sold? You should be able to derive the
profit function for the firm and find the profit-maximising price:
6
We are sidestepping
the question of the
starting price here –
suppose that the review
article simply states
the starting level of
expectations.
This is just making
things easier for us.
We could just as well
assume that secondperiod expectations
are some combination
of last period’s
expectations and real
sales, i.e. Ne(t+1) = a
Ne(t) + (1 – a ) N(t).
What effect would this
have on the dynamics of
the market?
7
N = 1 – p + 2*(1/3)
P = (p – 1/2) * (1 – p + 2/3)
p* = 13/12, N* = 7/12, P* = 49/144 ≈ 0.34
What is interesting about the optimum prices and sales is that at the
optimal price, actual sales exceed expected sales. Suppose now that in the
second period, consumers expect sales to be 7/12, that is, the first-period
sales level.7 If we derive the optimal price and demand, we find that
p* = 4/3 and N* = 5/6. This seems rather counterintuitive since prices
have gone up in the second period, yet demand has gone up as well. What
184
Chapter 17: Network effects
has happened? As we have established before, the demand function is well
behaved, so demand should decrease with price. Indeed it does, but only
for constant levels of expected sales. If we let expected sales vary,
we are in fact talking about two different demand curves.
Activity 17.3
Draw demand curves for the first and second period. Suppose now that expected sales
are a linear combination of past expected and real sales, i.e. Ne(t+1) = a Ne(t) + (1 – a)
N(t), where a = 50 per cent. What would be the demand curve in the second period?
Guidance on this activity can be found in the VLE.
The general profit function and optimal prices and quantities (for any level
of expectations) are (you should be able to derive this):
p* = 3/4 + Ne, N* = 1/4 + Ne, P* = (1/4 + Ne)2
We can see that profits increase in expectations, which makes sense
given our assumption of network effects. How can firms use this to
their advantage? One way to do it in our simple model is to influence
expectations in the first period in order to enjoy higher demand in the
second period. Leaving aside marketing activities and other strategies
outside the model, the firm could try and price strategically.
One strategy would be, for example, to price such that sales in the first
period are N(t) = 1, which would raise expectations in the second period
to Ne(t+1) = 1. The downside is that profits in the first period would
be lower than under optimal (static) pricing, the upside of it is higher
second-period profits. To see if this strategy is worth it, we need to solve
our model for both periods, assuming our initial demand curve and
expectations.
First-period prices p1 must be such that:
1 = 1 – p1 + 2/3, i.e. p1 = 2/3.
Calculating profits, we find that P1 = 1/6, which is much lower than
the first-period profits we found previously (0.34). In the second period,
however, we enjoy higher expectations and a demand function of:
N=1–p+2
This gives an optimal price of p2 = 7/4, N2 = 5/4, and P2 = 25/16, which
is much higher than 0.39, the second-period profits under regular dynamic
pricing. Ignoring discounting for a moment, combined first- and secondperiod profits are significantly higher under strategic pricing than under
non-strategic dynamic pricing.
Activity 17.4
a. Suppose that second-period profits are discounted by a factor d < 1. How low would
d have to be in order to render strategic pricing unprofitable?
b. Consider the demand function N = 1 – p + (1/10) * Ne. Using our two-period model
and initial expectations of Ne = 1/3, would a firm want to set first-period prices
strategically so that Ne(t+1) = 1? (That is, is it profitable to do so compared to regular
(static) pricing)?)
Guidance on this activity can be found in the VLE.
185
MN3119 Strategy
17.4 Generic strategies in network markets8
As mentioned before, entering into a market with an incumbent
technology with network effects is not always straightforward and
carries the risk of excess inertia. In particular, it makes it difficult to
establish a marginally better, but incompatible technology in the market
because consumers will be unwilling to switch and abandon their current
technology for one with uncertain benefits and small network size.
Firms or sponsors of new technologies in general will have to consider two
dimensions – the degree of technological improvement and the degree
of control over the new technology. Shapiro and Varian (1999) have
developed a list of generic strategies in network markets in light of these
two dimensions.
17.4.1 Backward compatibility versus performance
New technologies that are designed to be backward compatible with the
incumbent technology are constrained in their improvements – clearly, if
some of the product characteristics need to be shared by the old and the
new technologies, improvements in these characteristics is impracticable.
A radical new technology on the other hand is not constrained by the
existing. However, a technology that is backward compatible will be able
to link into the existing network of consumers, thus somewhat lowering
the disadvantage of having a small installed base for a new technology.
The main trade-off in choosing (assuming of course it can be chosen)
the degree of technological improvement is that backward compatibility
lowers the installed base problem, but renders the technology less superior
to the existing one.
15.4.2 Open standard versus control
Choosing the level of control over the new technology has important
implications for the ability to appropriate profits from the technology.
A closed standard (or ‘control’ strategy) will imply that if it succeeds in
replacing the incumbent technology as industry standard, the technology
sponsor has monopoly power over the technology. On the other hand, an
open standard is attractive to potential users because different suppliers
offer slightly differentiated versions of the new technology and prices are
likely to be much more competitive, which lowers the risk of being ‘lockedin’ to an all-powerful producer of an industry standard. Here again, both
strategies have advantages and disadvantages – control over a technology
ensures a larger share of potential future profits, while a relatively liberal
licensing policy or even an open standard makes adoption by the market
more likely.
The combination of these two dimensions yields four generic strategies –
controlled migration (control/compatibility), performance play
(control/performance), open migration (openness/compatibility), and
discontinuity (openness/performance).
186
Control
Openness
Compatibility
Controlled migration
Open migration
Performance
Performance play
Discontinuity
This section is based
on Shapiro and Varian
(1999), Chapter 7. The
strategies described here
are laid out in much
more detail there.
8
Chapter 17: Network effects
Activity 17.5
The following are all examples of new technology introductions – which of the four
generic network strategies do they represent?
a. Atari è Nintendo Entertainment System
b. Windows 95 è Windows 98
c. 1G (analog) mobile phones è 2G (digital) mobile phones
d. 2G mobile phones è 2.5G mobile phones
e. 3.5’’ diskettes è Zip disks (by IOmega)
f. Vinyl records è Compact discs
Guidance on this activity can be found in the VLE.
17.5 Fighting a standards battle
In the previous section, we only considered a single new technology trying
to replace an existing one. Quite frequently, however, there is more than
one candidate for the next generation of a technology – the VCR battle
between VHS and Betamax is just one of the many examples. In a market
with network effects, one of the technologies is likely to emerge as the
industry standard, as we have seen earlier in this chapter. We now turn to
some of the factors that affect the outcome of a standard battle and discuss
how firms can influence these factors.
First, the price of a technology relative to the rival technology plays an
important role. This may not be very surprising, but the consequences
of pricing too high may be much more dramatic than in conventional
markets. Since future sales are affected by the level of past sales, a small
disadvantage early on can translate into a major shortfall in the future –
therefore the incentives for strategic pricing will be even higher if there is
competition than if there is not.
Second, the relative qualities of the technologies play a role. This
again is not unexpected, but it should be noted here that a firm faces the
trade-off of trying to introduce a technology early to gain an installed
base advantage over its rival, but waiting as long as possible to ensure
the best possible technology. Pre-emption motives may be important in
network industries, but it is always important to maintain a sufficient level
of technological quality in order to make switching attractive in the first
place.
Third, as mentioned previously, backward compatibility may have an
impact on the outcome of a standards battle. If one technology sponsor
maintains backward compatibility, it can gain a huge advantage over its
competitors, but possibly at the expense of a technological advantage over
the incumbent technology.
Fourth, consumer expectations may play a much more important role
in network markets than in regular industries as we have seen, since high
expectations are likely to translate into a higher willingness to pay in the
future, and consequently even bigger advantages later on. Firms fighting
a standards battle will therefore make sure they portray their technology
as eventual winners even in the early market stages, for example by
highlighting past successes, comparing their technology favourably to their
rivals, gaining endorsements from credible ‘lead consumers’, or investing
in large production capacities that will only ever be used if the technology
becomes the industry standard.
187
MN3119 Strategy
Finally, a technology with a sufficient supply of complementary goods
can gain a lead over its competitors – the chicken-and-egg problem
of producers of complementary goods (why would I produce for no
consumers, and why would consumers choose a technology with no
complementary goods?) is often solved by the technology sponsor itself.
For example, by tapered integration where some of the complementary
good is produced in-house and some by outside suppliers. This helps
build up expectations and reaching critical mass, while ensuring sufficient
variety of complementary goods.
17.6 Key concepts
•
Direct and indirect network effects
•
Excess inertia
•
Excess momentum
•
Generic strategies in network markets
•
Compatibility
•
Standard battles.
17.7 A reminder of your learning outcomes
Having completed this chapter, and the Essential reading and Activities,
you should be able to:
•
define network effects and network market structures
•
explain technology diffusion patterns using network theory
•
discuss the four generic strategies firms can pursue in network markets
as described by Shapiro and Varian (1999).
17.8 Questions for discussion
1. Postrel (1990) describes the process of the failure of quadraphonic
sound to replace stereo sound. Which of the four generic strategies
was chosen by quad suppliers and why do you think it failed?
2. Sony and Philips joined forces to develop and market the compact disc
standard – what are the advantages and disadvantages of developing
jointly with one’s biggest competitor?
3. Excess inertia may occur for technologies with direct or indirect
network effects. How can a supplier of a technology with direct
network effects try and overcome excess inertia? Can the same
techniques be applied for an ‘indirect network effects technology’?
Guidance on this activity can be found in the VLE.
Extended activity: Skype and digital cinema
Read the following and answer the questions at the end.
Direct Network Effect ‘caselet’: Skype Technologies SA
Founded in 2002, the web-based communications company Skype
Technologies has over 60 million users in 225 countries, and has recently
been acquired by the online auction site eBay for $2.5 billion. Skype’s
aim is to revolutionise the $1 trillion global telecom industry through
its service of offering free calls using Voice over Internet protocol (VoIP)
technology.
188
Chapter 17: Network effects
Skype’s revenue model is distinct from other VoIP players: the company
does not charge its users monthly or per-minute fees. Instead, it lets users
make free calls to other registered Skype users, and charges them only
when calling landlines or users registered with a competing VoIP provider.
Network Effects: Digital Cinema in the USA
Digital cinema refers to the creation, transmission and projection of film
using digital technology – bits and bytes and satellites – instead of physical
reels of 35mm film.
Universal transition to digital cinema is ultimately expected to benefit all
parties on the value chain of this industry – the studios and distributors,
the digital technology and hardware developers, exhibitors (movie
theatres) and audiences – as more and better films can be made and
distributed more cheaply, flexibly and readily than ever before possible.
In 2002 the major studios – Disney, Fox, MGM, Paramount, Sony and
Warner – formed the Digital Cinema Initiative (DCI) to ‘establish and
document voluntary specifications for an open architecture for digital
cinema that ensures a uniform and high level of technical performance,
reliability and quality control’. The DCI was also committed to facilitating
the development of business plans and strategies to help spur deployment
of digital cinema systems in movie theatres.
The DCI was hailed as being the first bold step towards making digital
cinema a reality. It was expected to provide a catalyst for the market
transition and hasten its development. Most importantly creating a
universal format meant that the technology developers would not be able
to offer competing standards at different studios, creating a Betamax/VHS
scenario in which purchasers risked investing in the ‘wrong’ technology.
In July 2005, the DCI completed its technical specifications for digital
cinema, and large-scale rollout of the technology is now widely deemed
to be imminent. Sceptics, however, say the transition is still several years
off due to lack of infrastructure, ready supply of necessary technology and
support, and, finally, the lack of agreement about the economics of the
entire endeavour.
The chief impediment to progress lies in the fact that the costs of the
technology transition fall to the exhibitors, while the cost benefits accrue
to the studios/distributors. With installation costing up to $100,000 per
screen, say JP Morgan analysts, transferring to a digital cinema system is
a major capital expenditure. Conversely, a move to digital saves studios
$2,000–3,000 per print of each reel of 35mm film made for distribution,
while filming costs go down dramatically. For example, the 220 hours
it took to film Star Wars, Episode II: Attack of the Clones digitally cost
$16,000, as opposed to the estimated $1.8 million it would have cost using
traditional film. Unlike some European countries, where governments are
subsidising the digital cinema roll-out, particularly for the independent
sector, the US is leaving it to the commercial players to work out between
themselves just how the costs are split among them.
In between the studios and exhibitors are the companies such as Dolby,
Barco, Christie, Technicolor and Access Integrated Technologies, who
make the digital cinema systems and projectors. These companies have
a clear objective, which is to sell their products, and thus have emerged
as ‘brokers’ of the potential cost-sharing agreements between studios/
distributors and exhibitors. As such, the technology providers have created
a variety of different business plans that offer exhibitors different financing
arrangements. For example, some business plans propose that exhibitors
189
MN3119 Strategy
do not pay for initial installation costs but only for ongoing maintenance
and support of their new digital cinema systems. Under this arrangement
the costs are effectively borne by the studios over time, through the
payment to technology providers of ‘virtual print fees’ per film distributed
over a given number of years.
To date, none of the technology developers appear to have been able to
secure the critical mass of participants from either side of the equation –
studios or exhibitors – that would push the transition fully into motion.
Even once a watershed agreement is secured with one of the big exhibitor
chains, there is the issue of a current lack of hardware – for example, there
are currently only 85 digital projectors in stock in all of North America.
Aligning the supply with demand and getting the timing right will be
challenging, so long as there is such uncertainty in this market. What
is certain is that the film industry will reach the digital age eventually,
however bumpy the road.
Sources
Coster, P. ‘Digital cinema industry update: near the end of the reel’, JP Morgan
North America Equity Research, 20 October 2005.
‘Double Standards’, Media, Haymarket Publishing Services, Ltd, 9 September 2005.
Sperling, N. ‘Industry panel discusses 3-D projection’, The Hollywood Reporter 11
October 2005.
Sperling, N. ‘Access Integrated Technologies and Christie Digital Systems are
learning that rolling out digital cinema can be difficult’, The Hollywood Reporter
2 September 2005.
Questions
1. How would you characterise Skype’s business strategy? What is its
potential for success versus other VoIP players?
2. Describe and analyse the direct and indirect network effects of digital
cinema. What factors are most likely to affect the roll-out of digital
cinema?
190
Appendix 1: Sample examination paper
Appendix 1: Sample examination paper
Important note: The Sample examination paper reflects the
examination and assessment arrangements for this course in
the academic year 2013−2014. The format and structure of the
examination may have changed since the publication of this subject
guide. You can find the most recent examination papers on the VLE
where all changes to the format of the examination are posted.
Time allowed: three hours
Candidates have to answer a total of eight of the following twelve
questions: all four from Section A (12.5 marks each) and four of eight
(12.5 marks each) from Section B. Candidates are strongly advised to
divide their time accordingly.
The maximum length for the answer to each question is 300 words.
Section A
Answer four questions from this section (50 marks in total).
1. Explain the term ‘time compression diseconomies’.
2. You have been asked to define the market for private financial services,
e.g. small credit financing. Which technique(s) would you use and
what kind of data would you need to collect to do it?
3. The transition from 2G to 2.5G mobile telephony has been
characterised as ‘open migration’. Explain what this means and what
advantages or disadvantages it has for owners and consumers of the
new technology.
4. What are the major barriers to entry into the brewing industry and
how did the microbrewers get around them?
Section B
Answer four questions from this section (50 marks in total).
5. Explain how some firms might use ‘tapered integration’ to lower the
quasi-rents they are facing in the context of a particular transaction.
6. Product introductions in this market come at relatively regular
intervals. If consumers expect a new technology to supersede the
existing one soon, how will their incentive to adopt the current
technology be affected? What, if anything, could a supplier of razor
systems do to counter this tendency?
7. The intensity and success of research and development is said to
be strongly affected by the present market structure. What are the
replacement and efficiency effects for a monopolist? Under what
conditions is one stronger than the other?
8. Tapered integration is a way to combine some of the advantages
of vertical integration and outsourcing. Which motives for vertical
integration still hold when integration is not full, but tapered? Are
there advantages that only occur in the case of tapered (but not full)
integration?
191
MN3119 Strategy
9. One of the problems Matsushita/JVC faced while working on their
own digital audio standard was that they did not have easy access
to an affiliated software firm like Sony (in CBS) and Philips (in
Polygram). Assuming that their technology was of a similar quality,
what strategic options did Matsushita/JVC have? Analyse these options
(including the likely reactions by Sony and Philips) and discuss their
expected profitability.
10. How can resale price maintenance (RPM) help solve the problem of
investment externalities in case of two retailers in the same shopping
mall?
11. An industry has a C4-ratio of 80%.
a. What is the minimum Herfindahl-Hirschman index (HHI) for this
industry? Show your calculation.
b. What is the maximum HHI for this industry? Show your
calculation.
12. You are a manager of the (at the moment) only ice-cream parlour in a
large shopping mall. Someone told you about how endogenous sunk
costs can help your business. How could you use endogenous sunk
costs to maximise profits in this case?
END OF PAPER
192
Appendix 2: Guidance on answering the Sample examination paper
Appendix 2: Guidance on answering the
Sample examination paper
This course examines your ability to structure situations where there
is strategic interdependence between firms (or within firms). Special
emphasis is given to the application of concepts introduced in the subject
guide to new (real-life) situations. The weighting is such that passing
the course is not possible by simply regurgitating facts from the subject
guide; you have to give the examiners some evidence that you can transfer
the knowledge to new situations. Giving additional evidence of ‘thinking
outside the box’, for example by challenging a concept or outlining the
limits to its applicability, or giving detailed examples with some of the
mechanisms present, will usually be the difference between an upper
second and a first-class answer.
While there are some questions that have a right and wrong answer, many
questions will have cumulative answers where pure repetition gives some
credit (but not very much), some transfer of knowledge puts your mark
between a pass and a high upper second mark, and a genuinely new
concept or connection warrants a first-class mark. In other words, it is
possible for students to give a correct, but very ‘thin’ answer and still not
get high marks because they did not go far enough in their application of
the concept in the examination.
It is important that answers are concise and selective. 300 words is not a
lot, so candidates will have to make sure to select the most important facts
and concepts in their answer.
The following notes are guidance on answering the questions on the
sample examination paper (Appendix 1).
Section A
Question 1
This is more or less bookwork – which means it will be something that
the subject guide and reading will have described clearly and if you
have studied these carefully you should be able to provide a clear and
correct definition which will ensure a pass grade. Some examples of
time compression diseconomies, limits to the concept’s applicability, a
discussion of the implications of time compression diseconomies will
cumulatively enable you to get higher marks. Mentioning at least two
(better three) of all these factors would warrant a distinction mark.
Question 2
The focus in this question is on a competent application of the techniques
of market definition for a particular case or scenario. Simply listing all
available techniques will at most result in a low passing grade if the
techniques are presented very well. Proposing a select set of techniques
to this particular situation will get higher marks, and a practical guide of
the kinds of data and the method of generating this data (if it needs to be
generated) warrants a distinction.
193
MN3119 Strategy
Question 3
This question asks you to apply the generic strategy ‘open migration’ to
the mobile phone industry. Explaining ‘open migration’ competently is
sufficient for a pass mark, but linking it to the specific characteristics
of this particular case will be necessary for a higher mark. An informed
discussion of the pros and cons will include some mention of the other
generic strategies and why open migration has specific advantages or
disadvantages over them. The distinction in the pros and cons between
owners and consumers is necessary for a first-class mark.
Question 4
Barriers to entry have to be defined and applied to the case –
indiscriminately listing barriers to entry is not sufficient to pass! Specific
application to the case is necessary, and the strategic responses by entrants
are necessary for a distinction mark.
Section B
Question 5
Here, tapered integration must be defined briefly, but this is not enough
to pass. An understanding of quasi-rents is essential, and will be needed
for a pass mark. Linking these two concepts is relatively challenging, so a
good demonstration of how these two are linked will warrant a high upper
second class mark. Use of a concrete example with some specific industry
(or firm) knowledge then takes an answer up to the distinction or first
class level.
Question 6
A discussion of the incentives of adopters to purchase a new technology
is required here. Factors like uncertainty, network effects, previous usage,
pre-announcements by rivals could all be mentioned here. Overall, this
question tests your ability to discuss adoption incentives in a novel setting.
Question 7
Explaining innovation incentives is necessary here. You should include
a discussion on the factors influencing these effects and a sensible
assessment of when one of these effects dominate.
Question 8
This question is asking for a precise comparison of two scenarios. A list of
factors should be produced, with an informed judgment on the direction of
the inequality - i.e. are the advantages higher or lower with tapered than
with full integration?
Question 9
You are asked to be creative here in outlining different strategic options
for JVC/Matsushita and their respective advantages and disadvantages.
The strategies proposed should be sensible and practical to achieve a pass
mark, and they should include a good discussion of the pros and cons for a
higher mark.
194
Appendix 2: Guidance on answering the Sample examination paper
Question 10
This question requires you to understand the principle of investment
externalities – the fact that some of the benefit from an investment into a
brand does not accrue to the investor, but other firms. In particular, firms
can charge slightly lower prices and benefit from the investment of others.
RPM lowers this effect because prices are homogenous. A good answer will
detect this link and explain it successfully.
Question 11
Candidates have to realise that a C4 of 80% can still hide some
heterogeneity among the top 4 firms or the smaller firms. The minimum
is an even spread of 20% each of the four largest ((20*20)*4 = 1600)
and a mass of atomistic firms (=0), so that the minimum HHI is 1600.
The maximum HHI can be derived by knowing that larger firms get
increasingly larger weights, so that there could be one firm with 80%
market share and the rest of the market being atomistic, giving an HHI of
6400. A good answer will show this and show the workings.
Question 12
This question asks for an application of endogenous sunk costs to deter
entry. Bookwork of explaining endogenous sunk costs is not sufficient
to pass, it is important to show that entry can be deterred through this.
Moreover, a very good answer would include a concrete example of what
an ice cream parlour in a mall could do by way of endogenous sunk costs.
195
MN3119 Strategy
Notes
196
Notes
Notes
197
MN3119 Strategy
Notes
198
Notes
Notes
199
MN3119 Strategy
Notes
200