MARKET DOMINANCE AND ANTITRUST POLICY doc

MARKET DOMINANCE AND ANTITRUST POLICY,
SECOND EDITION
Utton2 00 prelims 4/12/02 16:20 Page i
For Mark and Harry
Utton2 00 prelims 4/12/02 16:20 Page ii
Market Dominance and Antitrust
Policy, Second Edition
M.A. Utton
Professor of Economics, University of Reading, UK
Edward Elgar
Cheltenham, UK • Northampton, MA, USA
Utton2 00 prelims 4/12/02 16:20 Page iii
© M.A. Utton 2003
All rights reserved. No part of this publication may be reproduced, stored in a retrieval
system, or transmitted in any form or by any means, electronic, mechanical, photocopying,
recording, or otherwise without the prior permission of the publisher.
Published by
Edward Elgar Publishing Limited
Glensanda House
Montpellier Parade
Cheltenham
Glos GL50 1UA
UK
Edward Elgar Publishing, Inc.
136 West Street
Suite 202
Northampton
Massachusetts 01060
USA
A catalogue record for this book

is available from the British Library
Library of Congress Cataloging in Publication Data
Utton, M. A. (Michael A.), 1939–
Market dominance and antitrust policy / Michael A. Utton.–2nd ed.
p. cm.
includes index.
1. Industrial concentration. 2. Market share. 3. Antitrust law. I. Title.
HD2757 .U88 2003
338.8–dc21
2002029832
ISBN 1 84064 728 0 (cased)
Printed and bound in Great Britain by Biddles Ltd, www.biddles.co.uk
Utton2 00 prelims 4/12/02 16:20 Page iv
Contents
List of figures and tables vi
Preface to the first edition vii
Preface to the second edition ix
PART I ANALYTICAL AND INSTITUTIONAL BACKGROUND
1 The economic analysis of market dominance 3
2 Market dominance in practice: current perceptions and trends 27
3 The antitrust response: an outline of antitrust policy in Europe
and the United States 44
PART II MARKET DOMINANCE: HORIZONTAL ISSUES
4 The measurement and interpretation of market dominance 59
5 Market conduct of dominant firms: I 85
6 Market conduct of dominant firms: II 126
7 Market dominance and collusion 149
8 Horizontal mergers and market dominance 171
PART III MARKET DOMINANCE: VERTICAL ISSUES
9 Vertical integration and vertical mergers 207

10 Market dominance and vertical restraints 233
PART IV PRIORITIES AND PROSPECTS FOR ANTITRUST
11 Priorities in antitrust policy 273
12 Antitrust policy in an international perspective 291
References 309
Index 323
v
Utton2 00 prelims 4/12/02 16:20 Page v
Figures and tables
Figures
1.1 Price and output under simple monopoly 4
1.2 Price and output under competition and monopoly with
organizational economies 14
1.3 Price discrimination and allocative efficiency 17
1.4 Welfare losses from market dominance 20
4.1 Monopoly price and welfare loss 62
4.2 The dominant firm and the competitive fringe 64
4.3 Excess profits and market power 70
5.1 First degree price discrimination 87
5.2 Third degree price discrimination 91
5.3 Price and cost of the dominant firm 104
5.4 Price strategy of the dominant firm 107
6.1 Contestability and entry 130
7.1 Collusion to raise price 151
7.2 Duopoly with differing costs 153
8.1 The welfare trade-offs in horizontal mergers 174
9.1 Incentives to integrate under different production conditions 211
9.2 Double monopoly mark-ups and vertical integration 213
10.1 Welfare effects of resale price maintenance 237
12.1 Dumping and price 294

Tables
2.1 Average five-firm concentration for a sample of 121
manufactured products in the UK, 1958–77 (unadjusted and
adjusted for imports) 37
2.2 Long-run shares of market leaders in six UK industries 39
2.3 Entry barriers and market structure 42
8.1 Calculation of the Herfindahl (H) index in a market with six
firms 179
vi
Utton2 00 prelims 4/12/02 16:20 Page vi
Preface to the first edition
Even to the casual observer it might appear that problems of market dominance
and antitrust policy are almost daily in the news. Amongst the more prominent
examples are the following: officials of the European Commission pay unan-
nounced visits to the headquarters of some of the most famous companies in
the world and seize documents that appear to show that they have been colluding
on prices; a takeover bid for a highly respected and long-established UK con-
fectionery firm is made by a foreign company and is allowed to proceed
unhampered despite widespread protest; in the aviation industry a company whose
name is a household word is accused of using predatory tactics to ruin a much
smaller competitor; in the UK the most prestigious and successful brewing
companies are horrified to learn that the Monopolies and Mergers Commission
have recommended that they should be forced to divest themselves of a large
proportion of their retail outlets or pubs; in the US the most successful computer
software company is accused of anticompetitive behaviour. Many other examples
could be cited and in subsequent chapters we will look in detail at cases from
the EU, the US and the UK involving collusion, mergers, the market conduct
of dominant firms and the market power that may or may not derive both from
vertical integration and from vertical restraints.
The issues and institutions involved, like industry itself, are complex. The

book is therefore structured in a way which we hope will allow the reader to
make sense of the complexity. There are four sections. Part I contains an
analysis of market dominance and its possible extent, with a preliminary review
of the institutions used to deal with it. The core of the book is then contained
in Parts II and III, distinguishing horizontal from vertical issues. In Part II,
particular attention is paid to the market conduct of dominant firms, which has
received so much recent theoretical attention. It also contains a discussion of
collusion, where the antitrust response has perhaps been the most uniform, and
horizontal mergers, where despite very intensive study many issues still remain
unresolved. In Part III, the emphasis switches to vertical issues: that is, those
involving the relationships between firms and their input suppliers or their dis-
tributors. Both theory and policy in this area have undergone significant changes
in recent years. Finally, in Part IV we raise a number of controversial questions
about the effectiveness of antitrust policy, including a discussion of the appro-
priate sanctions both against those who infringe the law and against those who
attempt to mould its application to their own purposes. We also take up sensitive
questions involving conflicts between antitrust and trade policies, the interna-
tional ‘reach’ of antitrust and foreign takeovers of domestic firms.
vii
Utton2 00 prelims 4/12/02 16:20 Page vii
viii Market dominance and antitrust policy
In courses in industrial and business economics in British universities,
antitrust policy tends to be relegated to a brief final chapter or passing reference
to a few well-known cases. Even though much of the preceding analysis may
have led up to some apparently important policy conclusions, the next step –
how these are or are not translated into actual policy – is often left unanswered
or merely given a fleeting reference. On the other hand, students of competi-
tion law or competition policy may acquire a detailed knowledge of many
cases without appreciating the economic analysis that may or, in some notorious
instances, may not underpin them. By bringing together in each chapter of Parts

II and III a discussion of the economic analysis and then the treatment of the
issues in European, British and US antitrust policy, we have attempted to
overcome this limitation.
The intention, therefore, is that the book should provide a useful accompa-
niment to courses in industrial and business economics, competition law and
institutions, and in some instances microeconomics where there is an emphasis
on market power issues. The level of economics assumed is no more than that
usually taught to first and second year undergraduates and what little algebra
is used has been largely relegated to the appendices.
Many of the topics have been discussed over several years with business
economics students at Reading, and I have benefited greatly from their scepticism
and readiness to challenge the conventional wisdom. I would especially like to
thank Lauraine Newcombe who coped superbly with the daunting task of deci-
phering my handwriting and preparing the final draft for the publisher.
M.A.U.
Utton2 00 prelims 4/12/02 16:20 Page viii
Preface to the second edition
I have used the opportunity of a second edition to make a number of substantial
changes. Some of these were necessary because the antitrust laws themselves
have changed, and some were required to discuss important recent cases.
In the first category was the British Competition Act passed in 1998 and which
came into effect in March 2000. The changes embodied in the Act were
probably the most substantial made in the fifty years of British competition policy.
In the process of aligning British policy closely with that of the European
Union, much of the previous machinery was swept away and existing institutions
were fundamentally altered. The detailed discussion of the previous policy,
especially that involving the Restrictive Practices Act, has therefore been
discarded to make way for coverage of the new policy. Similarly, fundamental
changes have been made to certain aspects of European policy, especially that
dealing with vertical restraints. The second edition, therefore, focuses on the

changes that came into effect in 2001.
In the second category are a number of cases which not only are highly
significant for antitrust policy but have been widely reported and extensively
discussed. Probably the most prominent was the case brought by the US
Department of Justice against Microsoft, allegedly for trying to monopolize the
market for operating systems. The proposed acquisition of McDonnell Douglas
by Boeing was widely reported for different reasons. The merger of two
quintessentially US companies was challenged by the European Union
competition authority. Until a compromise was agreed, the case threatened to
rupture US–EU commercial relations. Discussion of these and other cases are
included in the new edition. I have also renumbered references to Articles 85
and 86 as Articles 81 and 82 throughout, in accordance with the current Treaty
of Amsterdam.
Antitrust policy continues to evolve and inevitably, at the time of writing
(February 2002), the precise details of further changes in Britain (promised in
the Enterprise Bill) and in the EU (concerning the future structure of car
distribution) are not finalized. By the time the book is published they will be.
The outcome of these and subsequent changes will have to await a further edition!
M.A.U.
ix
Utton2 00 prelims 4/12/02 16:20 Page ix
Utton2 00 prelims 4/12/02 16:20 Page x
PART I
ANALYTICAL AND
INSTITUTIONAL
BACKGROUND
Utton2 01 chaps 4/12/02 16:21 Page 1
Utton2 01 chaps 4/12/02 16:21 Page 2
1 The economic analysis of market dominance
I Monopoly and market dominance

The immediate task of this opening chapter is to discuss some of the basic
concepts that will be used throughout the book. We begin by setting out in
simple terms the economic case against market dominance using the well-
known tools of static theory to highlight the inefficiencies that can arise from
monopolized compared with competitive markets. The discussion also allows
us to compare the ‘monopoly’ of economic theory with the looser notion of
‘market dominance’ which often finds its way into antitrust cases.
Although the static analysis highlights the core of the problem, real markets
do not remain frozen but are constantly undergoing change, even if the major
participants try to resist. In Section II, therefore, we extend the preliminary
analysis into the difficult area of dynamics and introduce a question which will
recur at many subsequent points in our discussion: namely, if in many cir-
cumstances positions of dominance generate a faster rate of growth or can be
eroded by pressures in the market, is ‘benign neglect’ a more efficient solution
to the problem than direct intervention by antitrust action which may be costly,
cumbersome or even wrong? In other words, we need to consider the question
of how selective antitrust policy should be. In Section III we try to allay the
suspicions of those who feel that antitrust action is unimportant because the
losses it tries to correct and repair are trivial. There is now quite a lot of evidence
that the costs, broadly interpreted, of market dominance can be considerable
and, although not as important, say, as hyperinflation or the depletion of the
ozone layer, are nevertheless great enough to merit detailed enquiry.
A monopolist in economic theory is the sole producer of a good or service for
which there are no close substitutes. Some impediment also exists which
prevents other firms from entering the market and competing with the incumbent.
Under these circumstances the firm can choose that output and thence price
which maximizes profit. The barrier to entry will also ensure that, whatever the
size of the profit, there will be no competition to affect the firm’s performance.
It will be useful for our subsequent discussion to represent these familiar
results in a simple diagram. Thus, in Figure 1.1, the monopolist’s demand and

marginal revenue are shown as AR and MR, respectively, and long-run average
cost, denoted by LRAC, is assumed to be roughly L-shaped and constant beyond
output Q
E
. This representation of the average cost curve has substantial empirical
support (see Johnston, 1960; Wiles, 1961). If average costs are constant beyond
Q
E
, marginal costs, denoted LRMC, will equal average costs over this range. For
profit maximization the firm will equate marginal cost with marginal revenue,
3
Utton2 01 chaps 4/12/02 16:21 Page 3
resulting in an output OQ
M
which can be sold for a price of P
M
. The monopoly
profit is represented by the rectangular area ABCP
M
.
We can now use this simple analysis to get a preliminary answer to some
basic questions of concern in antitrust. First, in what sense, if any, is the
monopoly represented in Figure 1.1 inefficient? We can also introduce the
related question of whether efficiency is the legitimate concern of antitrust
policy. Secondly, what is the character and role of the impediment to entry in
the market shown? Thirdly, what is the significance of the profits shown in the
figure? And fourthly, what relation is there between the monopoly illustrated
and firms in an antitrust case in a position of market dominance?
In order to address the first question we need to focus on the relationship
between marginal cost and price at the equilibrium output. It is clear from the

figure that at output OQ
M
there is a difference CB between demand price and
marginal cost. Since this difference is fundamental to the economic case against
monopoly we shall spend some time examining its full implications. Any point
on the demand curve represents the valuation at the margin by consumers of the
unit of output specified. Thus, assuming the output is infinitely divisible,
4 Market dominance and antitrust policy
Figure 1.1 Price and output under simple monopoly
Utton2 01 chaps 4/12/02 16:21 Page 4
consumers’ marginal valuation of the Q
M
unit of output is CQ
M
. The marginal
resource cost of this unit, however, is only BQ
M
(in the absence of production
externalities). The difference, shown as CB in Figure 1.1, amounts to the mark-
up of price over marginal cost. Over the entire output range Q
M
Q
C
consumers
would be prepared to pay a price (as indicated by the demand curve) greater than
the resource cost of production, indicating that such output would generate a
consumer benefit. In principle consumers would be prepared to pay the
monopolist a sum equal to the area ABCP
M
in exchange for an increase in output

to Q
C,
sold at a price OA. The consumers’ net gain would then be the equivalent
of the triangle BCD. The fact that the monopolist would restrict output below
Q
C
implies that resources are being misallocated: too few resources are devoted
to monopoly production in order that price can be maintained above marginal
cost. The monopolist’s pricing behaviour therefore leads to allocative ineffi-
ciency. At the heart of economists’ case against monopoly is this price–cost
divergence and the resource misallocation that results.
The concept of efficiency, however, is one that is often heard in discussions
of the relative merits of competition and monopoly and, unless the several
different meanings are kept clear, confusion is likely to occur. Thus although
the monopoly may be allocatively inefficient it produces at a technically
efficient scale of operations. By this we mean that the firm has built a plant of
a size large enough to take advantage of all available economies of scale, where
we include not only physical plant but also optimal organizational and marketing
practice. By combining all factors in a way which embodies best available
practice the firm is able to achieve minimum unit costs for producing and dis-
tributing its output Q
M
. As shown in Figure 1.1, costs can be minimized for a
scale of operations Q
E
or greater. For scales smaller than Q
E
unit costs would
be higher, as shown by the LRAC curve. The monopolist is thus technically
efficient but allocatively inefficient.

Those familiar with the perfectly competitive model will know that in long-
run equilibrium the industry will also be efficient in this sense. All firms will
be producing on the minimum point of their long-run average cost curves which
in this model are assumed to be U-shaped, owing to the influence of disec-
onomies of scale beyond the optimum output. On the assumptions currently
made, therefore, the main difference between monopoly and competition is
that, in the former, prices are greater than marginal cost and this leads to
allocative inefficiency. Both forms of market organization, however, are
technically efficient in the long run.
1
It is useful at this stage to introduce a further type of efficiency to which we
will refer in subsequent chapters. A number of writers, stretching back as far
as Adam Smith, have noted that, whereas the representation of costs in Figure
1.1 refers to the firm using the best available technology and organizational
methods so as to obtain minimum costs for any level of output specified, when
The economic analysis of market dominance 5
Utton2 01 chaps 4/12/02 16:21 Page 5
firms are in a protected market, incentives to achieve minimum cost may be
blunted. In particular the amount of effort put into achieving the firms’
objectives may be reduced at all levels of the organization so that it may operate
with a considerable amount of slack. The term used for this kind of internal dis-
organization is ‘X-inefficiency’. In terms of Figure 1.1, this would mean that,
instead of operating at a point B on the LRAC and LRMC curve, the firm would
be operating with a cost level above B for an output Q
M
. However, if that was
the case its chosen output would be to the left of Q
M
, since in effect it would
be attempting to obtain its objectives from a higher perceived unit and marginal

cost curve than that shown. We have hesitated to use the term ‘profit maxi-
mization’ in this context because if the firm allows its costs to rise in the way
indicated its objectives have clearly shifted and it may be pursuing ‘employee
and management satisfaction’, rather than profit maximization.
2
For our
purposes, however, it is sufficient to note that the stronger the protection for a
monopoly the weaker are the incentives for X-efficiency. If competitive forces
can be strengthened in a market then X-inefficiency will tend to disappear.
It is appropriate at this point to introduce the question of whether efficiency
should be the major goal of antitrust policy or whether broader or indeed
narrower objectives would lead to more satisfactory results. In Chapter 3 we will
look in detail at what are the stated objectives of antitrust policy in the EU, the
USA and the UK. For the moment we consider the issue in more general terms.
One approach is to argue that ultimately the consumer interest should be
paramount in antitrust questions and that this can be best achieved by the pursuit
of greater efficiency through antitrust policy. Bork certainly takes this view:
‘The whole task of antitrust can be summed up as the effort to improve
allocative efficiency without impairing productive efficiency so greatly as to
produce either no gain or a net loss in consumer welfare’ (Bork, 1978, p. 91).
Thus, on this view, the reduction of monopoly by removing impediments to
entry is likely to have the effect of improving allocative and X-efficiency. If the
incumbent firm continues to restrict output and pays insufficient attention to
the level of its costs it will lose market share and its profits will be eroded.
Consumers will gain through lower prices. The argument can be sustained even
though the increased competition may ultimately lead to the demise of one or
more of the competitors. In fact if the process of competition is effective we
expect over time some firms to disappear either through bankruptcy or merger,
and their place to be taken by new organizations. The speed with which the
process works is one of the main concerns of Chapter 2.

Although the ultimate objective to antitrust policy may be greater efficiency
and consumer welfare, by placing a greater emphasis on the competitive means
to that end some writers may unwittingly allow antitrust authorities to give
undue emphasis to the protection of competitors rather than to competition.
Thus if the goal of antitrust is seen primarily in terms of maintaining competition
6 Market dominance and antitrust policy
Utton2 01 chaps 4/12/02 16:21 Page 6
this may be interpreted by lawyers, judges or antitrust administrators as
requiring the continued presence of existing firms even though it might mean
damaging the consumer interest because they are inefficient and are attempting
to use the antitrust machinery to preserve their position. The pressures for this
outcome should not be underestimated, especially where an existing market
structure is dominated by one firm and where the complaint or action comes
from a smaller rival. Unless the antitrust body keeps efficiency and the consumer
interest firmly in mind it may serve the interests of weak competitors.
We turn now to the second question mentioned above, the character and role
of impediments to entry in the monopolized market. Some impediments or
barriers are quite straightforward and their effects clear-cut. The government
grant of a monopoly to a single firm or individual, for example, means that no
other firm can legally enter that market in the short or long run. The Stuart
kings found this a useful means of raising revenue and the governments of some
developing countries have granted exclusive import licences to individuals
owed a favour.
Other barriers are more complex and ambiguous. The government grant of
exclusive rights to a new product or process in the form of a patent has the
specific purpose of encouraging and rewarding invention and innovation even
though, for a time, a monopoly performance can be expected. The optimum
level of innovation and change can be thought of as dynamic efficiency. There
is clearly a trade-off involved between static and dynamic efficiency. Static
efficiency is impaired by the use of the legal monopoly (the patent) but the

grant of the patent may be the means by which new products are introduced or
the costs of existing ones reduced, so that over time there are positive benefits
to consumers. Furthermore, even with a patent system the impediment to new
entry is likely in many instances to be less complete than in the case of a
government-granted monopoly. Invention without patent infringement may
frequently be possible, especially by firms already working along similar
avenues of research.
Control of crucial raw materials is also often cited as an important
impediment to entry. The most extreme case would be where known high-grade
mineral deposits are very limited and owned by a single firm. Competitors must
then either make do with inferior (more costly) materials or depend on the
monopolist for their supplies. In either case their costs will be higher than those
of the monopoly. A similar effect may be felt through control of prime sites
for distributing a product. If the existing firm controls the best locations, other
firms will have to be content with slightly inferior ones, which would again
put them at a cost disadvantage. Although such extreme cases do undoubtedly
exist and have been maintained sometimes for long periods, others have been
much less secure simply because the very scarcity and high price of the resource
The economic analysis of market dominance 7
Utton2 01 chaps 4/12/02 16:21 Page 7
provides strong incentives for finding new deposits or other methods of
production and distribution.
If we move a little further down the hierarchy of entry barriers the ambiguity,
not to say controversy, increases. In the markets for consumer products where
incumbent firms may have very large market shares they may also have built
up over a long period strong consumer loyalty and brand preference through
their advertising and marketing policies. Such firms thus have a very valuable
intangible asset which entrants may be unable to create for themselves in the
short term, even though they may have access to the same advertising and
marketing media as the incumbent. Even in the long term the degree of

uncertainty surrounding the creation of a comparable intangible asset is likely
to be far greater than that associated, for example, with operating a production
line. In this sense, therefore, the entrant is at a cost disadvantage compared with
the incumbent. Whether this kind of cost disadvantage should rank with a
government-granted monopoly in the analysis of entry barriers has, however,
recently been the subject of considerable dispute. A number of writers have
argued that the new entrant faces similar expenditures in creating consumer
brand loyalty as the incumbent firm (or firms) have incurred in the past. As
long as the entrant has access to the same facilities as the incumbent, then no
long-term impediment exists. To sustain their argument they can rely on a well-
known definition of entry barriers by Stigler: ‘a cost of production (at some or
every rate of output) which must be borne by a firm which seeks to enter an
industry but is not borne by firms already in the industry’ (Stigler, 1968, p. 67).
If the entrant can create a similar intangible asset by replicating the past expen-
ditures of the incumbent, on this view, no entry barrier exists. The difficulty
arises, however, precisely because although similar costs may be incurred by
an entrant the market environment that it faces with an established dominant
firm is not the same as when the latter was building its market share.
Similar considerations surround the treatment of economies of scale. If an
entrant to a market where such economies are substantial can effectively
reproduce the production facilities used by an incumbent firm, there is no
impediment to entry, according to the Stigler view. Yet, if entry on a large and
therefore technically efficient scale were to occur in such a market, the increase
in output might be so great as to ensure that the market price would fall below
average costs. Entry would not be viable and therefore is unlikely to take place,
even though in principle a potential entrant could achieve the same cost level
as the incumbent.
Clearly there are many sources of impediment to entry to particular markets
and often different types will occur simultaneously. It is also evident that their
precise impact on the performance of existing and entrant firms will vary con-

siderably. In Parts II and III of the book we will be very much concerned with
these effects. For the moment we simply underline the significance of entry
8 Market dominance and antitrust policy
Utton2 01 chaps 4/12/02 16:21 Page 8
barriers to the performance of the monopoly shown in Figure 1.1. The presence
of some impediment to entry ensures that the firm retains its monopoly not
only in the short term but also in the long term. If the impediment were removed
the firm would still be able to charge a monopoly price in the short term but
the profits that resulted would attract new resources to the market and over
time the increased supplies that result would cause the price to fall to the
competitive level.
We thus come to the third question posed above: what is the significance
of profits shown in the simple monopoly case illustrated in Figure 1.1? As long
as the firm is protected by barriers to entry its policy of output restriction and
simple monopoly pricing will generate the excess profit shown as ABCP
M
.By
‘excess’ we mean an amount greater than that required to retain the resources
in their present use. On the usual convention that amount, the ‘normal’ return
or opportunity costs of capital, is included in the long-run average cost curve.
In a market economy the key role of returns greater than ‘normal’ is to attract
additional resources. The presence of entry barriers frustrates this mechanism.
To the extent that additional resources are kept out of a market because of
some barrier (of the kind mentioned above) which allows the incumbent firm
or firms to earn excess returns, the resource misallocation identified in
Figure 1.1 will persist.
This conclusion would appear to have important implications for antitrust
policy. Excessive profits identified amongst monopoly firms might give a good
indication of a poor allocative performance, and thus be of direct policy concern.
The British Monopolies and Merger Commission (MMC) often used such

indicators to assist its investigations, but they have to be treated very cautiously
and may be misunderstood or completely misleading.
3
In practice it may be
very difficult to distinguish short- from long-term influences and those that
have a benign effect (due to innovation) from those that are malign (due to
entry deterrence). It may also be very difficult to obtain data that reflect even
approximately the concepts used in the economic analysis of the problem.
Furthermore, if monopoly firms are prone to X-inefficiency as we suggest
above, recorded profits may appear modest simply because internal slack
accounts for the rest. On the other hand, a monopoly which has earned only
modest returns may be constrained by the threat of entry. Simply because a
firm has a very large market share does not mean that it can sustain a monopoly
price and earn a monopoly return. Only if its share is buttressed by barriers to
entry will this be possible.
So far we have deliberately used the notion of a monopolist as the sole seller
of a product for which there are no close substitutes, taken from the theory of
monopoly. We have thus been able to introduce a number of concepts that
will be particularly useful in the succeeding chapters. Since, however, a major
task of the book is to show how economic analysis can be used to analyse
The economic analysis of market dominance 9
Utton2 01 chaps 4/12/02 16:21 Page 9
antitrust problems we need now to address the question of the relationship
between the monopoly of economic theory and firms in antitrust cases in
positions of market dominance.
By ‘market dominance’ we mean the ability of a firm or group of firms per-
sistently to hold price above long-run average costs without thereby losing so
many sales that the price level is unsustainable. We have deliberately used the
term ‘market dominance’ rather than the more frequent ‘market power’ in order
to emphasize the central role played in many markets by one or a few large

firms. The terms can, however, be used interchangeably. In the light of the
foregoing discussion we can note the following points about this definition.
First, a market may be dominated by more than one firm and they may either
act together (through secret collusion, for example) or tacitly arrive at a price
solution close to the monopoly level by acting on what they believe is their
best strategy, given their anticipations of what the policy of the others will be.
Thus, whereas our previous discussion proceeded on the basis of a single firm
monopoly, in practice we have to recognize that a frequent case will be
oligopoly and the mutual interdependence that that implies.
Secondly, the price–cost difference has to be persistent. Temporarily high
prices in relation to costs may be caused by a variety of factors, including short-
run fluctuations in demand or input prices. High short-term profits may result
but they will not be due to market dominance. An unforeseen increase in
demand will generate windfall gains even in a highly competitive market. In
antitrust cases, however, ‘persistence’ has to be properly interpreted. Clearly
whether increases in demand cause prices to rise for one, two or five years
depends very much on the type of industry involved. In some cases additional
supplies from new entrants may be available within weeks or months of a
demand increase. They may come either from imports which were previously
uneconomic because of transport costs or from new firms entering the market.
In other cases technological complexities may mean that additional output from
entrants will only be available after several years. Nevertheless the price increase
caused by the shift in demand will not be sustainable against new entry, and the
existing firms cannot be said to dominate the market in the sense defined.
Thirdly, the definition refers to prices successfully held at a level greater
than average costs. Although not involving directly the concepts of perfect
competition and barriers to entry, it is clear from our previous discussion that
they have an important indirect bearing on the definition. In a perfectly
competitive market, prices will eventually fall to the level of marginal cost.
Even though there is no suggestion that the real world is populated by perfectly

competitive markets, nevertheless, in those where a reasonable approximation
is found, prices will in the long run be aligned with costs. The corollary is that,
where serious and prolonged divergences occur between price and cost, the
cause can be traced back to some impediment to entry. We may note in passing
10 Market dominance and antitrust policy
Utton2 01 chaps 4/12/02 16:21 Page 10
that the definition can clearly encompass different degrees of dominance
implying large or small divergences between price and cost. Clearly, given the
limited resources that will always be available for antitrust enforcement, it is
desirable that serious cases should take precedence over minor cases, not least
because these may be expected to create the largest welfare losses for
consumers.
We have left until last perhaps the most difficult question of all arising from
the definition of market dominance. This concerns the apparently innocuous
reference to ‘the market’. If a firm or firms are dominating ‘the market’ and if
this can have important policy implications it is evident that we need to have
a clear idea of what constitutes the relevant market in a particular case. In Figure
1.1 we were able to avoid the practical problems by simply drawing a
downward-sloping demand curve facing the monopolist, on the assumption
that the product in question had no close substitutes. The greater the gap in the
chain of substitutes, the smaller will be the price elasticity of demand, ceteris
paribus. Hence a firm or group of firms controlling all of the supply of a product
for which no close substitute exists will have greater scope for raising price
above cost.
When we move, however, from theory to antitrust policy some assessment
of what actually constitutes the affected market has to be made. As we shall
see in later chapters, large quantities of ink have been spilt and many hairs have
been split in trying to catch the elusive concept. Clearly the firms themselves
will argue in defence of their position that the correct market is very wide and
that consequently their dominance is relatively small. Complainants and antitrust

authorities are likely to think otherwise and draw the boundaries more narrowly.
One approach which has received much attention but which in practice may
be very difficult to employ involves the use of cross-elasticity of demand and
elasticity of supply. The cross-elasticity of demand between products X and Y
may be defined as the percentage change in the quantity of X demanded
resulting from a small percentage change in the price of product Y. In effect this
is merely a more formal way of identifying products which are close substitutes
for each other. Where the cross-elasticity is high, that is where a small change
in the price of Y results in a relatively large change in the quantity of X
demanded, the products can be regarded as part of the same market. Consumers
will evidently switch from one product to another very readily.
Although the concept may help to clarify thinking about the scope of markets,
there are practical reasons why it is of limited policy use. Estimates of cross-
elasticities are not readily available or easily made. An antitrust case does not
provide the ideal environment for making such estimates. Furthermore it is not
clear where the cut-off points should be, even if estimates were available. In the
context of an antitrust case, what constitutes a ‘relatively large’ cross-elasticity:
2, 5, 25 or higher? Additional adjustments would also have to be made to take
The economic analysis of market dominance 11
Utton2 01 chaps 4/12/02 16:21 Page 11
account of the fact that ‘close substitutability’ depends not only on physical
characteristics but on geographic location. Identical commodities produced
several hundred miles apart and costly to transport do not constitute part of the
same market.
Similar considerations apply to the elasticity of supply. The concept is defined
as the percentage change in the quantity of X supplied in response to a small
percentage change in the price of X. Thus, if a small increase in the price of X
causes a disproportionate increase in the supply of X, supply elasticity will be
high. In other words existing producers face few production difficulties in
increasing their output but also other firms, using flexible production methods,

can readily switch to producing X following a price increase. A further source
of additional supplies may be imports. If a small price change makes the
difference between profitable and unprofitable imports, clearly existing
domestic firms may have little room to manoeuvre on price; that is they will
have limited market dominance. In principle therefore high supply elasticity
will have an important influence on the way the true ‘market’ is perceived.
4
The principle is of increasing importance in European Union (EU) cases where
the continued removal of trade and other barriers between member states has
had the effect of ‘widening’ the market in the way indicated. This is not to say,
of course, that statistical estimation of supply elasticities is actually carried out,
because the same problems arise as in the case of cross-elasticities of demand.
The underlying logic, however, does have a bearing on the way a particular
market is viewed and consequently how much discretion the leading firm or
firms have.
The joint emphasis in the elasticity approach to both demand and supply sub-
stitutabilities is therefore correct and avoids the mistake of taking too narrow
a view of the market by simply accepting at face value the alleged market share
of existing firms. A seemingly different approach has been suggested by Areeda
and Turner, who define a market as ‘a firm or group of firms which, if unified
by agreement or merger, would have market power in dealing with any group
of buyers’ (Areeda and Turner, 1978, p. 347). The definition appears to go
directly to the heart of the problem by drawing the boundary of the market
around those firms which, if acting together successfully, could raise prices. It
is close to the Guidelines now used by the US antitrust authorities in merger
cases which we discuss fully in Chapter 8. Some assessment has to be made both
about the substitutability in demand of the outputs from the firms to be included
and about the possible switches in production that may take place in firms
ostensibly supplying a different market. In fact commercial history is strewn
with the remains of many restrictive agreements that collapsed precisely because

the participants failed to take account of the increases in supply that would
occur if the price was raised. Participants clearly have a very sharp incentive
to try to ensure that all suppliers are included in the agreement to make it
12 Market dominance and antitrust policy
Utton2 01 chaps 4/12/02 16:21 Page 12
effective. Yet they often fail. As has often been pointed out, the managements
of firms have a far greater knowledge of their own industry than either lawyers
or antitrust authorities. It is difficult therefore to envisage either doing any better
in their attempts to define a market by drawing a ring round all actual and
potential suppliers.
There is thus a considerable gulf between the concepts defined in economic
analysis and their close approximation in an actual case. This does not detract
from the importance of the analysis which, as we shall see, is a powerful tool
for distinguishing the relevant from the misleading or false.
II Market dominance: extensions to the preliminary analysis
The caution is reinforced when we move away from the simple, static analysis
to consider more complex cases. At this stage we merely introduce three issues
which will be of more detailed concern later on: significant scale economies,
price discrimination and incentives for innovation.
The first and most obvious point arising out of the case illustrated in Figure
1.1 is that we implicitly assumed that costs were unchanged whether the market
was supplied by one firm or by a larger group of firms. If there are significant
economies of large-scale organization, as seems evident from the experience of
some industries, then monopoly may be the most efficient form of organization.
In particular, if the cost-reducing economies are important enough, price may
be lower and output higher with monopoly than with a multi-firm competitive
structure. These results can also be usefully illustrated in Figure 1.2. In the
figure, market demand is shown by AR and, if production is organized by a
monopoly, MR is marginal revenue. The monopolist’s long-run average and
marginal revenue curves are shown by LRAC and LRMC, respectively. We

have thus retained the assumption that under monopoly unit costs decline to a
minimum (at output Q
N
) and then remain constant over the relevant range. The
profit-maximizing monopoly price is then P
M
and output Q
M
. All of this is the
same as in Figure 1.1. However we now show the competitive supply curve as
S
C
. In other words, if production is organized by a large number of small-scale
firms, their aggregated costs are embodied in the horizontal line S
C
.
5
In this
case the market-clearing price is P
C
and corresponding output Q
C
. Price would
therefore be higher and output lower than under monopoly.
What can we say about the comparative efficiency of the two contrasting
cases? As far as technical efficiency is concerned there is no question that the
monopoly is much more technically efficient than the competitive industry. It
uses far fewer resources to produce output Q
C
or its preferred output Q

M
. The
resulting lower price means that consumers who, as it were, enter the market
at prices between P
C
and P
M
have their demands satisfied by the monopolist,
whereas they would have remained uncatered for by the competitive industry.
6
As far as allocative efficiency is concerned the answer to the question is more
The economic analysis of market dominance 13
Utton2 01 chaps 4/12/02 16:21 Page 13
complex. The fact that a greater output is sold at a lower price under monopoly
implies that consumers are better off than under competition, and yet it is clear
from Figure 1.2 that the monopoly output is sold at a price greatly in excess of
marginal cost. At output Q
M
marginal cost is FQ
M
, whereas price is EQ
M
, giving
a mark-up over cost of EF. Thus despite the lower price the monopolist is
generating considerable allocative inefficiency. Consumers are better off but
they could be made even better off if a means were found of making the
monopolist charge a price P
R
, equal to marginal cost. Consumers would then
have the full benefit of the scale economies, and the allocative inefficiency

would be eliminated. We state the possibility without suggesting that there is
a means readily to hand which would produce these desirable results.
Indeed, in a very real sense, the possible conflict between technical and
allocative efficiency lies at the heart of many antitrust questions. In addition to
the case shown in Figure 1.2, it frequently arises in slightly less extreme form
in large horizontal mergers where there may be a strong probability that the
merger would increase market dominance but at the same time reduce costs
through economies of reorganization.
7
Similarly, a fragmented market structure
14 Market dominance and antitrust policy
£
Output
LRMC
LRMC'
LRAC
A
E
S
C
P
M
F
MR
P
C
Q
N
Q
C

Q
M
AR
0
Figure 1.2 Price and output under competition and monopoly with organi-
zational economies
Utton2 01 chaps 4/12/02 16:21 Page 14