Content uploaded by Sean Sullivan
Author content
All content in this area was uploaded by Sean Sullivan on Aug 01, 2021
Content may be subject to copyright.
2. Market Definition
Sean P. Sullivan
1
July 31, 2021
I. INTRODUCTION
Monopolization, in the United States, and abuse of dominance, in the Eu-
ropean Union, embody different philosophies about how best to police single
firm conduct in competition law. Surprisingly, their disagreements end at mar-
ket definition. Both doctrines define relevant markets by similar processes and
use relevant markets for similar purposes. In some contexts, this type of agree-
ment would be a welcome sight. Here, it reflects a pocket of confusion in each
area of law.
This chapter describes the confusion of current market definition practices
and takes some initial steps toward a more coherent approach. It begins, in Part
II, with a brief review of the common justifications for defining markets in sin-
gle firm conduct cases, and a brief restatement of the common standards for
defining markets. Part III then turns a critical eye to the assumed connections
between current market definition practices and the substantive purposes the
resulting markets are meant to serve. Parts IV and V sketch the outline of a
more coherent approach to market definition in single firm conduct cases. The
idea behind this approach is that markets should not be defined by rote appli-
cation of omnibus standards, but should be defined by tests tailored to the spe-
cific purposes that relevant markets are meant to serve in a given application.
1
I am extremely grateful to Alexander Asawa for research assistance on this project. Amy
Koopmann and the University of Iowa Law Library also provided invaluable support and as-
sistance.
Final working paper for Sean P. Sullivan, Market Definition, in Research Handbook on
Abuse of Dominance and Monopolization (Pınar Akman, Konstantinos Stylianou, &
Or Brook eds., forthcoming 2022).
2
II. COMMON STANDARDS
Everyone knows the standards for defining markets in single firm conduct
cases. Advocates in both the US and the EU have spent decades intoning the
same familiar language about interchangeability and the substitutes to which
consumers may turn. The stated reasons for defining markets, and the uses to
which those markets are put, are also similar across jurisdictions.
A. Market definition in monopolization cases
In the United States, Section 2 of the Sherman Act makes it illegal for any
person to ‘monopolize, or attempt to monopolize, or combine or conspire …
to monopolize’ any trade within the scope of interstate commerce.
2
Interpret-
ing this sparse language, the US Supreme Court has held that the offense of
monopolization consists of two elements: ‘(1) the possession of monopoly
power in the relevant market and (2) the willful acquisition or maintenance of
that power as distinguished from growth or development as a consequence of
a superior product, business acumen, or historic accident’.
3
The related offense
of attempted monopolization edges away from the threshold requirement of
monopoly power, replacing it instead with the firm’s probable acquisition of
monopoly power as a result of its anticompetitive conduct.
4
Monopoly power, or its imminent acquisition, is thus a necessary element
in proving any violation of Section 2. So, what is monopoly power? The Su-
preme Court has described it as ‘the power to control prices or exclude com-
petition’.
5
But this is a very broad net. Nearly every competitor has some power
to control prices, as well as some capacity to exclude at least certain types of
competition.
6
Alternatively, monopoly power is sometimes defined as
2
15 USC § 2.
3
United States v Grinnell Corp 384 US 563, 570–71 (1966).
4
See Spectrum Sports Inc v McQuillan 506 US 447, 459 (1993) (‘[P]etitioners may not be
liable for attempted monopolization under § 2 of the Sherman Act absent proof of a dangerous
probability that they would monopolize a particular market and specific intent to monopo-
lize’.); Image Technical Services Inc v Eastman Kodak Co 125 F 3d 1195, 1202 (9th Cir 1997)
(observing that monopolization and attempted monopolization offenses ‘are similar, differing
primarily in the requisite intent and the necessary level of monopoly power’).
5
United States v E. I. du Pont de Nemours & Co 351 US 377, 391 (1956) (Cellophane case).
6
See Northern Securities Co v United States 193 US 197, 406 (1904) (Holmes J, dissenting)
3
something like substantial market power.
7
This, is a little more descriptive, but
only a little. If market power is defined as the ability to price above a competi-
tive level, then identifying monopoly power by this definition requires that two
judgements be made, neither guided by clear standards. First, a determination
must be made about what a ‘competitive’ price would be.
8
Second, another de-
termination must be made about whether the current or predicted price is ‘sub-
stantially’ greater than this competitive price.
The practical difficulty of applying either of the above definitions of mo-
nopoly power drives most litigants to follow a different path when seeking to
establish a violation of Section 2. Ever since the decision of United States v Alu-
minum Company of America in 1945, courts have endorsed an assumption that
a large share of a relevant market—something on the order of an 80 or 90%—
is evidence of a firm’s monopoly power in that market.
9
This is not to say that
a large market share is always indicative of monopoly power. But, according to
conventional thinking, a large share is at least a necessary condition for pos-
sessing monopoly power.
10
This, of course, requires the definition of relevant markets, and the Su-
preme Court has supplied several standards for the exercise. By one standard,
the scope of a relevant market is to be ‘drawn narrowly to exclude any other
product to which, within reasonable variations in price, only a limited number
of buyers will turn; in technical terms, products whose “cross-elasticities of de-
mand” are small’.
11
By another standard, relevant markets are ‘composed of
products that have reasonable interchangeability for the purposes for which
(‘According to popular speech, every concern monopolizes whatever business it does, and if
that business is trade between two states it monopolizes a part of the trade among the states.
Of course, the statute does not forbid that. It does not mean that all business must cease’.).
7
See Eastman Kodak Co v Image Technical Services Inc 504 US 451, 481 (1992) (‘Monopoly
power under § 2 requires, of course, something greater than market power under § 1’).
8
See Daniel A Crane, ‘Market Power Without Market Definition’ (2014) 90 Notre Dame
Law Review 31, 38–39.
9
See United States v Aluminum Co of America 148 F 2d 416, 424 (2d Cir 1945); Grinnell
(n 3) 571 (Alcoa) (‘The existence of [monopoly] power ordinarily may be inferred from the
predominant share of the market’.).
10
See, eg, Duncan Cameron, Mark Glick, and David Mangum, ‘Good Riddance to Market
Definition?’ (2012) 57 Antitrust Bulletin 719, 720–21.
11
Times-Picayune Publishing Co v United States 345 US 594, 612 n 31 (1953).
4
they are produced—price, use and qualities considered’.
12
By another standard,
relevant markets can be identified by the recognition of a product’s ‘peculiar
characteristics and uses’.
13
By yet another standard, markets—or, at least, ‘sub-
markets’—may be recognized by ‘practical indicia’ such as industry or public
recognition of a market, the presence of unique production facilities, or the
observation of distinct customers, distinct prices, or other distinct features.
14
Other tests supplement these standards. Initially developed for horizontal
merger analysis,
15
the Hypothetical Monopolist Test (HMT) defines markets
around groups of competitors who, if they were to band together as monopo-
lists, would elect to raise prices by at least a small but substantial amount for
some period of time. In the merger context, the HMT is an iterative algorithm
that starts from a candidate market as small as the merging parties and pro-
gressively expands the candidate market until the group of competitors it en-
compasses would have the joint market power and profit motive to exercise the
hypothesized price increase.
16
The HMT can be used to define relevant markets
in monopolization cases,
17
but sometimes requires modification to function
usefully in this context—a point to which we will return shortly.
B. Market definition in abuse of dominance cases
In the European Union, Article 102 of the Treaty on the Functioning of the
European Union prohibits the ‘abuse by one or more undertakings of a domi-
nant position’ within at least a substantial part of the internal market.
18
In con-
trast to the Sherman Act, Article 102 spells out several of the specific acts it
aims to prohibit. It explains that abuse of a dominant position includes
12
Cellophane (n 5) 404.
13
United States v E. I. du Pont de Nemours & Co 353 US 586, 593–94 (1957) (du Pont-
General Motors case).
14
Brown Shoe Co v United States 370 US 294, 325 (1962); see also Grinnell (n 3) 572 (en-
dorsing, in dicta, the use this test in monopolization cases).
15
See Gregory J. Werden, ‘The 1982 Merger Guidelines and the Ascent of the Hypothetical
Monopolist Paradigm’ (2003) 71 Antitrust Law Journal 253.
16
United States Department of Justice and Federal Trade Commission, Horizontal Merger
Guidelines (19 August 2010) § 4.1.1.
17
See, eg, United States v Microsoft Corp 253 F 3d 34, 81 (DC Cir 2001).
18
Consolidated Version of the Treaty on the Functioning of the European Union [2012]
OJ C326/47, art 102.
5
‘imposing unfair purchase or selling prices’, ‘limiting production … to the prej-
udice of consumers’, and excluding rivals or otherwise distorting the competi-
tive process, among other things.
19
By its terms, Article 102 only applies to the conduct of an undertaking with
a ‘dominant position’.
20
So, what is a dominant position? In United Brands, the
Court of Justice described dominance as ‘a position of economic strength en-
joyed by an undertaking which enables it to prevent effective competition be-
ing maintained on the relevant market by giving it the power to behave to an
appreciable extent independently of its competitors, customers, and ultimately
of its consumers’.
21
In guidance, the Commission has also contributed its view
that dominance occurs when ‘competitive constraints are not sufficiently ef-
fective’, such that an undertaking ‘enjoys substantial market power over a pe-
riod of time’.
22
As in the case of US monopolization, these efforts to define dominance do
little more than gesture at the concept. Again, the definitions fail to clearly sep-
arate dominant undertakings from the large mass of undertakings with at least
some market power. Again, the effort to equate dominance with ‘substantial
market power’ stumbles at the difficulty of identifying substantial market
power. And, again, the cases have addressed these difficulties by interpreting
large shares of relevant markets as the primary indicator of dominance.
23
Mar-
ket shares of 70 to 80% have been described as ‘a clear indication of the
19
Ibid.
20
Ibid.
21
Case 27/76 United Brands Co v Commission [1978] ECR 207, para 65.
22
Guidance on the Commission’s enforcement priorities in applying Article 82 of the EC
Treaty to abusive exclusionary conduct by dominant undertakings [2009] OJ C45/7, para 10.
23
See Case C-62/86 AKZO Chemie BV v Commission [1991] ECR 3359, para 60 (stating
that ‘the Court has held that very large shares are in themselves, and save in exceptional cir-
cumstances, evidence of the existence of a dominant position’, citing Case 85/76 Hoffman-La
Roche & Co AG v Commission [1979] ECR 461, and holding that the inference of dominance
applies ‘where there is a market share of 50% such as that found to exist in this case’); Com-
mission Notice on the Definition of Relevant Market for Purposes of Community Competition
Law [1997] OJ C372/5, para 10 (‘[A dominant position] would usually arise when a firm or
group of firms accounted for a large share of the supply in any given market, provided that
other factors analysed in the assessment (such as entry barriers, customers' capacity to react,
etc.) point in the same direction’. (internal footnotes omitted)).
6
existence of a dominant position’,
24
with shares greater than this extending be-
yond dominance into something closer to monopoly.
25
This places market definition in roughly the same position in abuse of
dominance cases that it occupies in monopolization cases. The processes of
market definition are also similar. For example, implementing regulation de-
fines a relevant market by language similar to the US standard of reasonable
interchangeability: ‘A relevant product market comprises all those products
and/or services which are regarded as interchangeable or substitutable by the
consumer, by reason of the products’ characteristics, their prices and their in-
tended use’.
26
The cross-elasticity standard is also used: ‘Factors relevant to the
assessment of the relevant product market include the analysis of … cross-
price elasticity of demand’.
27
As explained by the Commission in the Notice on
Market Definition, the goal of market definition is to identify the competitive
constraints that act upon an undertaking: ‘Basically, the exercise of market def-
inition consists in identifying the effective alternative sources of supply for the
customers of the undertakings involved’.
28
This recalls the HMT, which is also
expressly adopted as a potential standard when defining relevant markets in
abuse of dominance cases.
29
III. COMMON PROBLEMS
Much could be said about the various standards for defining relevant mar-
kets,
30
but the more intriguing puzzle is the assumption—common to both the
US and the EU—that current market definition practices connect relevant
24
Case T-30/89 Hilti AG v Commission [1991] ECR II-1439, para 92.
25
See Hoffman-La Roche (n 23) para 39 (distinguishing a dominant position from ‘a mo-
nopoly or a quasi-monopoly’).
26
Commission Implementing Regulation (EU) No 1269/2013 of 5 December 2013 amend-
ing Regulation (EC) No 802/2004 implementing Council Regulation (EC) No 139/2004 on the
control of concentrations between undertakings [2013] OJ L336/1, Annex 1, sec 6.1; see also
Commission Notice on the Definition of Relevant Market (n 23) para 9 (adopting this defini-
tion in other contexts).
27
Commission Implementing Regulation (n 26).
28
Notice on Definition of Relevant Market (n 22), para 13.
29
Ibid paras 15–17.
30
See Sean P Sullivan, ‘Modular Market Definition’ (2021) 55 UC Davis Law Review
(forthcoming).
7
markets to the underlying concerns of the substantive law. What makes a large
share of a relevant market evidence of monopoly power or occupancy of a
dominant position? What makes any of the market definition standards helpful
in assessing allegedly exclusionary or abusive conduct? The conventional an-
swers to these questions are not inspiring. They do, however, offer glimpses
into the confusion that currently envelops market definition practice.
A. Markets as threshold tests
One conventional argument for defining markets in single firm conduct
cases is the notion that only the conduct of a firm with a large enough share of
a relevant market warrants special scrutiny. Consistent with this idea, market
shares below about 50 percent are treated as inadequate to establish monopoly
power in US courts,
31
while something like 40 percent or a little less gates en-
forcement of Article 102 in the EU.
32
This positions market definition, and
market share analysis, at the threshold of the competitive effects inquiry,
33
with
market shares below minimum thresholds ending inquiries before they get to
31
See, eg, Bailey v Allgas Inc 284 F 3d 1237, 1250 (11th Cir 2002) (‘A market share at or
less than 50% is inadequate as a matter of law to constitute monopoly power.’); Rebel Oil Co v
Atlantic Richfield Co 51 F 3d 1421, 1438 (9th Cir 1995) (collecting ‘numerous cases’ that hold
‘a market share of less than 50 percent is presumptively insufficient’); see also Phillip E Areeda
and Herbert Hovenkamp, Antitrust Law, vol 2B (4th edn, Wolters Kluwer 2014) para 532c
(collecting other holdings).
32
Guidance on the Commission’s enforcement priorities (n 22) para 14 (‘The Commission
considers that low market shares are generally a good proxy for the absence of substantial mar-
ket power. The Commission’s experience suggests that dominance is not likely if the under-
taking’s market share is below 40% in the relevant market.’); Commission, ‘DG Competition
discussion paper on the application of Article 82 of the Treaty to exclusionary abuses’ (2005)
para 34 <https://ec.europa.eu/competition/antitrust/art82/discpaper2005.pdf> accessed 10
June 2021 (DG competition discussion paper)31 (‘[A]lthough also undertakings with market
shares below 40% could be considered to be in a dominant position…. [U]ndertakings with
market shares of no more than 25% are not likely to enjoy a (single) dominant position on the
market concerned.’).
33
See William M Landes and Richard A Posner, ‘Market Power in Antitrust Cases’ (1981)
94 Harvard Law Review 937, 972 (describing the usual treatment of market power analysis in
rule of reason cases as ‘“a threshold condition’”); Guidance on Enforcement Priorities (n 22)
para 9 (‘“The assessment of whether an undertaking is in a dominant position and of the degree
of market power it holds is a first step in the application of Article 82.’”).
8
competitive effects analysis.
34
That raises a question: why should low shares of
a relevant market have this abortive effect?
One possible justification for the approach is the frequent suspicion that a
firm without a large share of a relevant market cannot have the market power
necessary to raise monopolization or abuse of dominance concerns. This might
make sense if market definition controlled the price elasticity of market de-
mand. But no current market definition standard except the HMT has any se-
rious connection to market demand elasticity. The problem with this justifica-
tion is thus that the relationship it posits is not a reliable product of current
practice. As William Landes and Richard Posner observed—over forty years
ago—without controlling for demand elasticity, ‘a given market share is neither
necessary nor sufficient for a firm to be able to raise prices above the competi-
tive level’.
35
While tests like the HMT offer one context where market share
may have a consistent relationship with market power, the traditional practice
of lumping together all different tests of market definition leaves Landes and
Posner’s critique as true today as it was back then.
36
A large share of a relevant
market does not always indicate market power, and a small share does not pre-
clude it.
37
Another possible justification for the threshold-test interpretation is that a
firm without a large share of a relevant market is not economically significant
enough to warrant scrutiny under monopolization or abuse of dominance law.
A focus on economic significance motivated the US Supreme Court in its ar-
ticulation of the practical indicia standard,
38
and similar justifications have
been offered for the size of a hypothesized price increase under the HMT.
39
34
See Frank H Easterbrook, ‘The Limits of Antitrust’ (1984) 63 Texas Law Review 1, 17–
18 (advocating for the use of a market power ‘filter’ in antitrust analysis).
35
Landes and Posner (n 33) 952–53.
36
See Commission Notice on the Definition of Relevant Market (n 23) para 52 (contem-
plating the definition of relevant markets by a combination of different factors); Sullivan (n
30) § I (observing the apparently simultaneous application of different market definition
standards in antitrust opinion).
37
Franklin M Fisher, ‘Diagnosing Monopoly’ (1979) 19(2) Quarterly Review of Economics
& Business 7, 18 (commenting on false belief that small market share shows the absence of
monopoly power while large market share shows its presence and noting that what matters is
what happens to market share when monopoly profits are sought).
38
See, eg, Brown Shoe Co v United States 370 US 294, 325 (1962).
39
See Areeda and Hovenkamp (n 31) para 537a.
9
This focus on economic significance might make sense if market definition did
indeed reliably identify markets of significant economic importance. But cur-
rent practice does no such thing. Markets defined by the HMT are often nar-
row
40
and bear no guarantee of economic significance.
41
Markets defined by
substitutability inquiries similarly focus on the closeness of products, not the
economic significance of those products as a class. The practical indicia test
might capture markets of reliable economic significance
42
but, again, the typi-
cal practice of defining markets by lumping together different standards dilutes
this relationship. In short, under current practice, a firm’s large share of a rel-
evant market says little to nothing about its size or importance in a local, na-
tional, or global economy.
B. Markets as sets of substitutes
Another common argument for defining markets in single firm conduct
cases is the idea that doing so helps to identify sets of reasonable substitutes for
the products of the firm in question. The intuitive justification for this project
appears to be a belief that markets are identifiable scopes of trade in the world,
bounded by the similarity of products in serving particular purposes or satis-
fying particular needs.
43
The current author and others have argued that it is a
category error to think of relevant markets as constructs with any presence in
the world or permanence outside of a given inquiry.
44
But, even setting aside
this error, what relevance would identification of a given set of substitutes have
for assessing monopolization and abuse of dominance concerns?
40
See Horizontal Merger Guidelines (n 16) § 4 para 8 (‘Relevant antitrust markets defined
according to the hypothetical monopolist test are not always intuitive and may not align with
how industry members use the term “market.”’); Sullivan (n 30) § II.A (discussing the tendency
of most modern market definition standards to identify narrow markets focused on particular
concerns).
41
See David Glasner and Sean P Sullivan, ‘The Logic of Market Definition’ (2020) 83 An-
titrust Law Journal 293, 339–40.
42
See Sullivan (n 30) § I.B.
43
See ibid 303–07 (discussing attempts to identify ‘natural’ market boundaries by refer-
ence to substitutability).
44
Ibid 307–12; Magali Eben, ‘The Antitrust Market Does Not Exist: Pursuit of Objectivity
in a Purposive Process’ (2021) Journal of Competition Law and Economics (forthcoming); Ste-
ven C. Salop, ‘The First Principles Approach to Antitrust, Kodak, and Antitrust at the Millen-
nium’ (2000) 68 Antitrust Law Journal 187, 188–89.
10
That question can be sharpened by the well-known Cellophane fallacy. As
generations of students have learned, the US Supreme Court made a serious
mistake in its market definition analysis in United States v E.I. du Pont de
Nemours & Co.
45
Translating the error into HMT terms, the Court defined the
relevant market by asking whether DuPont would have found it profitable to
increase the price of its cellophane product. Finding this would not be profita-
ble—since DuPont was presumably already pricing as high as its market power
would allow—the Court expanded the relevant market to include products that
were only realistic substitutes at DuPont’s already elevated prices. This expan-
sion of the relevant market reduced DuPont’s share of the market—perversely
allowing DuPont’s exercise of market power to masquerade as evidence of
DuPont’s lack of market power.
The error of the Cellophane fallacy has been well aired in the scholarly lit-
erature,
46
and it now motivates cautionary qualifications in prescriptions of the
HMT for use in monopolization and abuse of context cases.
47
For example, it
is frequently said that a competitive price must be used as the baseline price
when attempting to apply the HMT to single firm conduct analysis.
48
This is
not exactly right. We will shortly return to this point with examples where the
Cellophane fallacy does and does not apply in the single firm conduct context.
45
Cellophane (n 5) 380–81.
46
The earliest clear articulation of the error seems to be George W Stocking, ‘Economic
Tests of Monopoly and the Concept of the Relevant Market’ (1957) 2 Antitrust Bulletin 479.
For modern recitations, see Salop (n 44) 197–98, Richard A Posner, Antitrust Law (2nd edn,
University of Chicago Press 2001) 150–51.
47
Eg, Commission Notice on the Definition of Relevant Market (n 23) para 19 (noting the
Cellophane problem); OECD, Abuse of dominance in digital markets (2020) 14
<www.oecd.org/daf/competition/abuse-of-dominance-in-digital-markets-2020.pdf> accessed
10 June 2021 (‘The core concepts of market definition in abuse of dominance cases are the
same as those applied in merger cases. However, care must be taken when applying analytical
techniques such as the hypothetical monopolist test in markets whose conditions may already
have been shaped by market power.’); DG Competition discussion paper (n 32) para 15 (‘It is
essential to take account of the fact that the SSNIP-test normally is based on the assumption
that prevailing prices constitute the appropriate benchmark for the analysis. This assumption
often does not hold in Article 82 cases.’).
48
Eg, Patrick Massey, ‘Market Definition and Market Power in Competition Analysis:
Some Practical Issues’ (2000) 31 Economic & Social Review 309, 323 (‘The cellophane trap
means that a different approach is required in abuse of dominance cases.’).
11
First, however, a few additional points about current practice need to be ad-
dressed.
The underlying problem at issue in the Cellophane fallacy extends beyond
the HMT. All efforts to define markets as sets of substitute products suffer from
the same fundamental problem that beguiled the Court in Cellophane. Reason-
able interchangeability, cross-price elasticity, the HMT—every substitution-
based test of market definition is exposed to an interpretive problem known to
economists since well before any of these tests were developed. As Fritz
Machlup explained in 1952:
[O]ne must have certain prices, price ranges, or price rela-
tions in mind when one speaks of particular demand elastic-
ities. The cross-elasticity of demand may be of very different
magnitudes at different price relations and, hence, when
making an estimate of the elasticity, one obviously thinks of
the currently existing price relations. If we now say that the
cross-elasticity of demand between two products is zero, it
may refer only to the given price relation, while at others the
cross-elasticities may be positive. Where this is the case, the
monopoly position exists only within certain price ranges
and is therefore an imperfect one.
49
Put another way, the very project of trying to identify a single set of substi-
tute products is economically unsound. There is no one set of substitute prod-
ucts, only different sets of substitutes at different prices.
50
To focus on a useful
set of substitutes, one must start from a clear understanding of what price range
is relevant to an inquiry. That is possible—as addressed in more detail below—
but it is not how relevant markets are defined or interpreted today.
49
Fritz Machlup, The Economics of Sellers’ Competition (John Hopkins Press 1952) 546–
47.
50
See Franklin M Fisher, ‘Economic Analysis and “Bright-Line” Tests’ (2008) 4 Journal of
Competition Law and Economics 129, 132 (‘The relevant facts of Cellophane are undisputed.
At a high enough price for cellophane, there was substitution of other flexible wrapping papers.
At lower, still profitable prices, there was not. Once one has stated that (and specified the
prices), one has said all there is to say.’).
12
C. Markets as collections of competitive constraints
One last argument for defining markets in single firm conduct cases is that
doing so helps to identify the competitive constraints acting upon the firm in
question. Franklin Fisher was an early advocate of this understanding of mar-
ket definition in antitrust cases. In 1979, he explained that, if market definition
is to facilitate analysis, it needs ‘to place in the relevant market those products
and services and firms whose presence and actions can serve as a constraint on
the policies of the alleged monopolist’.
51
Fisher contended that the breadth of
a relevant market should be whatever was needed to embrace the ‘significant
constraints’ on market power:
[A] ‘market’ is something that can be monopolized. If you
have left out significant constraints on power, the ‘market’ is
too small. If you have kept in firms and products or services
that are not significant constraints, the ‘market’ is too large.
52
Something analogous to Fisher’s constraints interpretation of market defi-
nition motivates enforcement practices in the EU. A 2005 DG Competition
discussion paper states: ‘The main purpose of market definition is to identify
in a systematic way the immediate competitive constraints faced by an under-
taking’.
53
This language parallels that of the Commission’s 1997 Notice on the
Market Definition,
54
which devotes an entire section to commenting on the
various competitive constraints that may be identified in the process of defin-
ing relevant markets.
55
For abuse of dominance analysis, this focus on constraints is tethered to the
substantive understanding of dominance itself. The Commission explains this
point in the same 1997 notice:
The concept of relevant market is closely related to the ob-
jectives pursued under Community competition policy. …
Under the Community’s competition rules, a dominant
51
Fisher (n 37) 13.
52
Fisher (n 50) 133.
53
DG Competition discussion paper (n 32) para 12.
54
Commission Notice on the Definition of Relevant Market (n 23) para 2.
55
Ibid paras 13–24.
13
position is such that a firm or group of firms would be in a
position to behave to an appreciable extent independently of
its competitors, customers and ultimately of its consumers.
56
Other Commission guidance explains dominance as a situation in which
‘the undertaking’s decisions are largely insensitive to the actions and reactions
of competitors, customers and, ultimately, consumers’.
57
Insensitivity to cus-
tomers and competitors is thus seen as a necessary condition for dominance
and market power.
58
There is undoubtedly something helpful in the constraints approach, but
the usual articulation fails to capture what it is. To start, the simple equation of
market power with the absence of competitive constraints is inconsistent with
basic economics. All firms face competitive constraints. Even those with sig-
nificant market power. Examples include workhorse models in industrial or-
ganization economics. A single price monopolist facing a competitive fringe
exercises durable market power despite the presence of competitive con-
straints.
59
In many oligopoly models, firms likewise exercise durable market
power despite the presence of direct competition from other firms.
60
These ex-
amples illustrate a general point. As courts have intuited since the dawn of
modern competition law, all market power is exercised within the bounds of
some set of competitive constraints.
61
56
Ibid para 10. This definition of dominance was articulated in Hoffman-La Roche (n 23)
[38].
57
Guidance on the Commission’s enforcement priorities (n 22) para 10.
58
Ibid para 11 (‘[A]n undertaking which is capable of profitably increasing prices above
the competitive level for a significant period of time does not face sufficiently effective com-
petitive constraints and can thus generally be regarded as dominant’); DG Competition dis-
cussion paper (n 32) para 23 (‘For dominance to exist the undertaking(s) concerned must not
be subject to effective competitive constraints. In other words, it thus must have substantial
market power.’)
59
See, eg, Louis Kaplow and Carl Shapiro, ‘Antitrust’ in A Mitchell Polinsky and Steven
Shavell (eds), Handbook of Law and Economics, vol 2 (Elsevier 2007) 1073, 1180–83.
60
See, eg, ibid 1083–86.
61
See, eg, United States v Addyston Pipe & Steel Co 85 F 271, 292 (6th Cir 1898) (observing
that firms possessed a window of market power despite the presence of competitive con-
straints: ‘Within the margin of the freight per ton which Eastern manufacturers would have to
pay to deliver pipe in pay territory, the defendants, by controlling two-thirds of the output in
14
A similar error underlies the notion that dominance over a market traces
to the independence or insensitivity of a firm to the actions of others in the
market. Consider the textbook single-price monopolist. No firm better cap-
tures the idea of dominance, yet the single-price monopolist is emphatically
sensitive to, and dependent upon, the actions of its consumers. The monopolist
raises its price until it is in the elastic portion of the demand curve, and the
limit of its price is the threatened exit of the marginal consumer still willing to
buy at the monopolistic price. The same point applies for competitors. Even a
monopolist is typically sensitive to the prices of marginal substitute products
and to changes like the expansion of firms in its competitive fringe.
62
Again, all
market power is subject to constraints.
There is, of course, an important role for competitive constraints in single
firm conduct cases. In any given equilibrium, the addition of a new constraint
might depress a firm’s market power while the elimination of an existing con-
straint might increase it. The identification of these competitive constraints can
be helpful in evaluating potential changes in a firm’s market power as a result
of its conduct.
63
What the identification of these constraints does not do is shed
much light on the thing it is commonly said to illuminate: whether a given firm
is a monopolist, occupies a dominant position, or is presently able to exercise
substantial market power.
IV. A MORE COHERENT APPROACH
The single most reliable way to get market definition wrong is to try to de-
fine markets before deciding what questions need to be answered. Contrary to
what seems to be standard practice today, market definition cannot usefully
come before a determination of what exactly the competitive concern entails.
64
pay territory, were practically able to fix prices.’); Alcoa (n 9) 426 (similarly finding that ‘within
the limits afforded by the tariff and the cost of transportation, ‘Alcoa’ was free to raise its prices
as it chose’).
62
Lawrence J White, ‘Market Power and Market Definition in Monopolization Cases: A
Paradigm Is Missing’ in W Dale Collins (ed), Issues in Competition Law and Policy, vol 2
(American Bar Association Section of Antitrust Law 2008) 913, 920 (‘[D]emand for a monop-
olist’s product should be expected to be sensitive (at the margin) to the prices of sellers of some
substitutes … and thus the monopolist’s price should be expected to vary as well.’).
63
See Eben (n 44) 30 (suggesting something similar through a focus on the ‘competitive
constraints … most appropriate to the alleged conduct and theory of harm’).
64
See Glasner & Sullivan (n 41) 312–15 (describing the theory-dependence of market
15
A single market also cannot generally serve every analytical purpose that might
be needed. Helpful and reliable market definition starts from the questions that
need to be answered and reasons backwards to the identification of markets
responsive to those needs.
65
This may compel the definition of different rele-
vant markets for different parts of the evaluative process.
66
All depends on what
questions need to be answered.
So, what questions do need to be answered in single firm conduct cases?
Let us focus, for now, on traditional market power concerns. Three categories
of concerns could arise from the challenged conduct: (1) it could allow the firm
gain new market power, (2) it could allow the firm to maintain existing market
power, or (3) it could reflect the firm’s exercise of its existing market power.
Within each of these broad categories of concern, subcategories could arise.
Single firm conduct is not conventionally broken apart to this level of granu-
larity—and we must save for another day a full exploration of all the possible
permutations of concern—but this deconstruction helps to uncover some use-
ful rules and guides to follow when defining relevant markets.
A. Conduct that increases market power (results in higher prices)
To start, consider a firm not presently in possession of monopoly power or
a dominant position, but which threatens to acquire that status through its
challenged conduct. The concern is a substantial increase in the firm’s market
power: something that would allow it to raise prices relative to where they are
now. Example fact patterns could include a merger from duopoly to monopoly
or the acquisition of a essential patent by one of several current competitors.
In the US, this type of conduct could be challenged as attempted monopoliza-
tion.
67
In the EU, Article 102 arguably does not fit the facts—the conduct in
question is creating a dominant position, not exploiting it—but a flexible
definition).
65
See Sullivan (n 30) 33 (proposing a purpose-based approach to selecting between differ-
ent methods of market definition).
66
For specific discussion of defining multiple relevant markets, see Glasner and Sullivan
(n 41) 330–33; Sullivan (n 30) 26–27.
67
See n 2 (prohibiting ‘attempt to monopolize’).
16
enough interpretation of dominance could capture many of the interesting
cases.
68
Helpful market definition, in this context, can borrow from the standard
practices of market definition in horizontal merger analysis. Where exclusion-
ary conduct is used to obtain new market power, a helpful relevant market col-
lects those competitors that currently constrain the firm, and thus whose ex-
clusion could result in an increase in the firm’s market power. Something anal-
ogous to the HMT, starting from the current price, would be an appropriate
test for identifying this type of market. The problem of the Cellophane fallacy
does not apply, here, since these markets are being defined to evaluate potential
increases in market power relative to whatever market power the firm may al-
ready have. The HMT may, however, require a larger-than-average hypothe-
sized price increase to approximate the concept of substantial market power,
at least where the firm in question appears to have little market power at pre-
sent.
69
Next, consider a firm already in possession of monopoly power or a domi-
nant position, but which threatens to obtain even more market power via the
challenged conduct. In the US, this type of conduct could be challenged as sim-
ple monopolization.
70
In the EU, the dominant position of the firm would bring
its conduct within the reach of Article 102. In either case, the inquiry involves
two steps: (1) assessing whether the firm has monopoly power or a dominant
position, and (2) assessing whether the firm’s conduct is anticompetitive or
constitutes an abuse of a dominant position. Each of these inquiries may be
aided by a distinct relevant market.
As discussed above, large shares in a relevant market are often interpreted
as proof of substantial market power. Without necessarily endorsing this prac-
tice, we can observe three properties of market definition necessary for a large
share of a relevant market to support this interpretation. First, the market
68
See n 18 (prohibiting ‘abuse … of a dominant position’ (emphasis added)); Pinar Akman,
The Concept of Abuse in EU Competition Law: Law and Economic Approaches (Hart Publishing
2012) 94 (tracing the original intent of Article 102 to the prevention of ‘the dominant under-
taking receiving advantages that would not be possible but for its dominance’). But cf n 32
(providing some flexibility for dominance to be found even at moderate share thresholds).
69
See Glasner & Sullivan (n 41) 316–17 (explaining why the size of the hypothesized price
increase in the HMT must match the magnitude of the anticipated exercise of market power).
70
See Grinnell (n 3) 570–71.
17
would need to be defined by a process that controls in some way for demand
elasticity.
71
A process like the HMT does this indirectly through its hypothe-
sized price increase: by construction, a monopolist in an HMT market has at
least enough market power to implement a price increase of the hypothesized
magnitude. Markets defined by abstract reference to substitutability or inter-
changeability offer no such guarantee. Second, the hypothesized price increase
in the HMT would need to be large enough to equate a large market share with
the possession of substantial market power.
72
Exactly what constitutes substan-
tial market power may be a fact-bound question and has not been adequately
explored to date. Third, the problem of the Cellophane fallacy emphatically
does apply in this situation, so care must be taken not to assess patterns of sub-
stitution at already elevated prices. If something like the HMT is used, then it
must be implemented as a hypothesized price increase above an estimate of a
competitive base price.
73
The above considerations apply only to the first inquiry in the challenge.
Now consider the second inquiry. Even if the firm possesses significant market
power, is its conduct reasonably capable of increasing that market power?
74
If
market-based inferences are to be used in this second inquiry,
75
it will typically
be helpful to define a different relevant market when doing so. The reason for
this is that the purpose of the relevant market is now different. The point is no
longer to establish the firm’s market power—the purpose of the first relevant
market—but to help evaluate the firm’s ability to obtain additional market
power through its challenged conduct. Intuitively, this second market defini-
tion exercise is analogous to that involved in the attempted monopolization
71
See generally Landes and Posner (n 33).
72
See n 69.
73
Ibid 320–21.
74
See Salop (n 44) 195 (observing that a firm’s ability to price above marginal cost ‘does
not mean that the firm can maintain or enhance its power by engaging in specific conduct
alleged to be anticompetitive’).
75
See Herbert Hovenkamp, Federal Antitrust Policy: The Law of Competition and its Prac-
tice 110–11 (5th edn. 2016) (citing predatory pricing, foreclosure, and tying as examples of
conduct offenses that only makes sense if the firm in question has a large market share).
18
challenge. Helpful market definition can again follow the pattern of horizontal
merger analysis and the Cellophane fallacy again ceases to be a problem.
76
As discussed above, the equation of monopoly power or a dominant posi-
tion with a firm’s possession of ‘substantial market power’ requires the hypoth-
esis of an appropriately substantial price increase when using shares of an
HMT market to assess the firm’s possession of monopoly power or a dominant
position. Does the same apply to HMT markets used to assess the firm’s chal-
lenged conduct? How large must the hypothesized price increase be in this sec-
ond market? These are difficult questions. While insignificant changes in mar-
ket power may not warrant intervention, one would expect both monopoliza-
tion and abuse of dominance to prohibit conduct which, though insignificant
in isolation, may in aggregate have significant effects.
77
A reasonable minimum
requirement is that the conduct ‘relates’ to the firm’s current market power
78
or that it is reasonably capable of contributing to that market power.
79
Beyond
this, some form of sliding scale may be appropriate, such that larger exclusion-
ary effects are demanded of firms with lower initial market power. This would
smooth the distinction between monopolization and attempted monopoliza-
tion in US law.
80
B. Conduct that maintains market power (preserves current prices)
Now, consider single firm conduct that does not risk the creation of new
market power, but instead risks entrenching, preserving, or otherwise operat-
ing to maintain the firm’s existing market power. Example fact patterns may
include a dominant incumbent’s acquisition of a nascent but potentially dis-
ruptive rival or a monopolist’s adoption of contracting practices that limit the
76
Here, the focus of concern is an increase in market power, and that may be appropriately
modeled by a price increase over the current-price baseline.
77
See Phillip E Areeda and Herbert Hovenkamp, Antitrust Law, vol. 3 (4th edn, Wolters
Kluwer 2015) para 651g (‘Any single exclusionary act may seem trivial. … [Yet] it may be fit-
ting to presume the exclusionary act ‘significant’ or ‘causally related’ to the monopoly power
being challenged.’).
78
Ibid (‘[I]t must at least appear plausible to an informed observer that the exclusionary
act could have had, or would probably have, a significant relationship to the defendant’s mo-
nopoly.’).
79
Ibid (rephrasing as restraints that ‘reasonably appear capable of making a significant
contribution to creating or maintaining monopoly power’).
80
See n 69 and accompanying text.
19
ability or incentive of customers to trade with its rivals. In the US, conduct
which has the effect of maintaining existing market power can be challenged
as monopolization under Section 2.
81
Recent cases have, however, narrowed
the situations in which a monopolist may be obligated to allow its competitors
to grow.
82
In the EU, conduct that would limit production or development in
order to maintain a dominant position is proscribed by Article 102.
83
Indeed,
undertakings which hold a dominant position are frequently said to have ‘a
special responsibility not to allow [their] conduct to impair genuine un-
distorted competition in the internal market’.
84
As in the previous discussion
of monopolization challenges and their EU analogues, multiple relevant mar-
kets may be helpful in evaluating maintenance-of-market-power concerns.
The first relevant market to consider is a market defined and used to estab-
lish the firm’s current possession of substantial market power. This is the same
exercise discussed in relation to the first inquiry in an acquisition-of-market-
power challenge and does not require repetition here.
85
It should, however, be
noted that the effectiveness of some tools of exclusion may depend upon a
firm’s size and importance relative to its rivals. Examples include exclusion
through volume discounts or exclusivity contracts, where the exclusionary po-
tential of the conduct arises primarily because of the large size or importance
of the firm. In some cases, the same relevant market used to establish the firm’s
81
See Grinnell (n 3) 570–71 (‘willful acquisition or maintenance of [monopoly] power’
(emphasis added)); Aspen Skiing v Aspen Highlands Skiing 472 US 585, 602 (1985) (‘purpose
to create or maintain a monopoly’ (emphasis added)). See also Herbert Hovenkamp, ‘The Mo-
nopolization Offense’ (2000) 61 Ohio State Law Journal 1035, 1041 (‘[A] great deal of strategic
behavior is concerned with the ‘maintenance’ rather than the acquisition of monopoly
power.’).
82
See Verizon Communications Inc v Law Offices of Curtis V. Trinko LLP 540 US 398, 415–
16 (2004) (‘The Sherman Act ... does not give judges carte blanche to insist that a monopolist
alter its way of doing business whenever some other approach might yield greater competi-
tion.’).
83
See Hoffman-La Roche (n 23) para 91 (including in the concept of abuse, behavior which
‘has the effect of hindering the maintenance of the degree of competition still existing in the
market or the growth of that competition’ (emphasis added)).
84
Case C-52/09 Konkurrensverket v TeliaSonera Sverige AB [2011] ECR 527, para 24; see
also Eleanor Fox, ‘Monopolization and Abuse of Dominance: Why Europe is Different’ (2014)
59 Antitrust Bulletin 129, 135 (‘[D]ominant firms have a special responsibility not to erect or
maintain barriers that frustrate the access of nondominant firms to markets.’).
85
See nn 71–73 and accompanying text.
20
present market power may double as a way of establishing the firm’s relative
size and importance when evaluating this type of exclusionary conduct.
86
Since a firm’s current possession of substantial market power is no guaran-
tee that its challenged conduct operates to maintain that power,
87
a separate
inquiry is needed to identify which competitors may be limited or excluded by
the challenged conduct and to assess whether these competitors would have
more tightly constrained the dominant firm—eroding the firm’s market
power—but for the effect of the challenged conduct. As Lawrence White has
explained, this aspect of maintenance-of-market-power challenges can itself be
usefully deconstructed into two questions: (1) what would the relevant rivals’
performance be (or have been) but for the exclusionary conduct of the domi-
nant firm, and (2) what would the effect of this performance be (or have been)
on the behavior of the dominant firm and the options available to consumers?
88
If market-based inferences are to be used in addressing either part of this
inquiry, one market concept responsive to the maintenance-of-market-power
concern would be something like a reverse HMT: a relevant market defined as
a group of competitors who, if unrestrained by the firm’s conduct, would likely
expand or take other actions that would lead to a small but substantial depres-
sion of prices for some product sold by the monopolist or dominant firm. Note
that this definition of a relevant market is easier to state than it is to operation-
alize. The hypothesis involves predicting both the future actions of firms and
the effects of those actions on prices. Also, markets defined by this process may
not be helpful in every case. Where only a single competitor is included in the
relevant market, for example, the market adds nothing to a direct evaluation of
White’s two inquiries. But where multiple competitors are simultaneously con-
strained, or where future competitive constraints could arise from different
86
See Hovenkamp (n 81) 1041 (‘Much monopolistic conduct is rational behavior only on
the premise that the firm is already a monopolist, and frequently the conduct is designed not
so much to create monopoly in a secondary market as to maintain the dominant firm’s position
in the primary market.’).
87
See n 74; Thomas G Krattenmaker, Robert H Lande and Steven C Salop, ‘Monopoly
Power and Market Power in Antitrust Law’ (1987) 76 Georgetown Law Journal 241, 255 (of-
fering similar comments).
88
White (n 62) 923.
21
sources, the above process of defining a relevant market may provide helpful
context for the requisite analysis.
89
Two additional observations about market definition are noteworthy. First,
relevant markets defined for purposes of evaluating the firm’s conduct may be
quite different from relevant markets defined for purposes of evaluating the
firm’s market power. In United States v Microsoft, for example, Microsoft was
found to have market power in the worldwide market for ‘Intel-compatible PC
operating systems’.
90
This would be the relevant market for the first inquiry
above. But Microsoft’s exclusionary conduct occurred in a broader scope of
trade including ‘middleware’ products like Java and Netscape.
91
That broader
scope of trade could be seen as the relevant market for the second inquiry. In
Microsoft, the DC Circuit rightly rejected Microsoft’s argument that the targets
of alleged exclusion needed to be included in the relevant market—that is, the
first relevant market.
92
The court rested this conclusion on a temporal distinc-
tion between nascent and present substitutes,
93
but it would have been clearer
to observe that different relevant markets may be defined for different pur-
poses. Here, the relevant market used to evaluate Microsoft’s current market
power simply involved a different set of competitors than the relevant market
used to evaluate Microsoft’s ability to maintain its market power against future
erosion by middleware products.
94
Second, a remaining consideration, applicable to most single firm conduct
cases but especially important for maintenance concerns, is the role that the
dominant firm’s size can play in explaining its incentive to exclude rivals or
89
See Thomas G Krattenmaker and Steven C Salop, ‘Anticompetitive Exclusion: Raising
Rivals’ Costs to Achieve Power over Price’ (1986) 96 Yale Law Journal 209, 255 (‘[E]nough of
the purchaser’s actual and potential rivals must suffer the price increase so that remaining un-
excluded rivals cannot or will not prevent the purchaser from exercising power over price.’).
90
Microsoft (n 17) 52.
91
See ibid 53–54 (discussing the future threat that middleware products posed for Mi-
crosoft).
92
Ibid. See also Case T-83/91 Tetra Pak International SA v Commission [1994] ECR II-
00755, para 116 (rejecting the argument that abuse of dominance necessarily requires abuse to
occur in the same market that is dominated).
93
Ibid 54 (‘Because middleware's threat is only nascent, however, no contradiction exists.
Nothing in § 2 of the Sherman Act limits its prohibition to actions taken against threats that
are already well-developed enough to serve as present substitutes.’).
94
See n 75.
22
raise their costs.
95
In many cases, the larger the dominant firm’s base of existing
sales, the more benefit it will extract from obstructing the competitive inroads
of rivals. Intuitively, a 5 percent market-wide price reduction may equate to a
far greater profit loss for the dominant firm with an 85 percent market share
than it would for smaller competitors in the same market. The dominant firm
may thus find it profitable to invest in forestalling future price decreases where
smaller competitors would not.
96
When shares of a relevant market are used to
assess the comparative incentives of firms to engage in exclusionary or cost-
raising conduct, it seems intuitive that the relevant market for the second in-
quiry will often be the appropriate basis for share computations. This is only
intuition, however, and exceptions may arise.
C. Conduct that exercises market power (sets current prices)
Finally, consider single firm conduct risking neither the creation nor the
maintenance of market power but instead merely reflecting the exercise of
market power. Examples of this type of conduct include a firm charging mo-
nopolistic prices or insisting upon terms of trade which could not be demanded
if it did not already possess substantial market power. Under US law, this con-
duct falls squarely outside the reach of Section 2. Cases have long held that a
monopolist commits no wrong if it obtains monopoly power without engaging
in prohibited types of exclusionary conduct,
97
and recent cases have settled
what little doubt may have remained about the legality of exercising monopoly
power, lawfully obtained.
98
In Pacific Bell v linkLine Communications, the
95
Franklin M Fisher, ‘The IBM and Microsoft Cases: What's the Difference?’ (2000) 90
American Economic Review 180, 180 (‘An anticompetitive act by a single firm is one that is
not profit-maximizing without the monopoly rents that it creates or maintains but is profit-
maximizing with those rents included.’)
96
See Krattenmaker, Lande and Salop (n 87) 259 (‘The greater the disparity in market
shares between the firm seeking to raise its rivals’ costs and the rivals, the greater the firm's
anticipated reward for achieving a higher price for its output. Hence, such a firm would be
willing to spend more in attempting to exclude rivals to gain power over price.’).
97
See Grinnell (n 3) 571 (distinguishing illegal monopolization from ‘growth or develop-
ment as a consequence of a superior product, business acumen, or historic accident’); Alcoa (n
9) 429 (inquiring whether a firm had ‘monopolized’ a market or whether ‘monopoly [had]
been thrust upon it’).
98
Cf Areeda and Hovenkamp (n 77) para 650a (‘[A]s the cases have clearly held, §2 is
concerned not only with remedying monopoly power to which undesirable conduct has made
a substantial contribution, but also with enjoining the conduct itself.’); Trinko (n 82) 407 (‘The
23
Supreme Court declared that ‘antitrust law does not prohibit lawfully obtained
monopolies from charging monopoly prices’.
99
For the same reasons, Section 5
of the FTC Act is unlikely to prohibit the mere exercise of market power to-
day.
100
Under EU law, dominant undertakings are less free to exercise their market
power. Article 102 includes in its definition of abuse the act of ‘imposing unfair
purchase or selling prices’. Case law supports the obvious interpretation of this
language as reaching something like monopolistic pricing. As the Court of Jus-
tice held in United Brands, ‘charging a price which is excessive because it has
no reasonable relation to the economic value of the product supplied would be
[within the scope of the abuse of imposing an unfair selling price]’.
101
The Com-
mission has also issued guidance identifying as illegal conduct which is ‘directly
exploitative of consumers, for example charging excessively high prices’.
102
At
least in principle, this means that excessive pricing, or equivalent contracting
practices, could be challenged as abuse of dominance under EU law.
103
In evaluating exercise-of-market-power concerns, it is hard to find any an-
alytical need for relevant markets or market definition.
104
The traditional infer-
ence of substantial market power from a large share of a relevant market could
still be drawn. (And, if this inference is to be drawn, the relevant market should
be defined subject to the practices discussed above.) But, unless something like
excessive pricing is to be inferred from the possession of market power alone,
mere possession of monopoly power, and the concomitant charging of monopoly prices, is not
only not unlawful; it is an important element of the free-market system.’); Ibid. (‘To safeguard
the incentive to innovate, the possession of monopoly power will not be found unlawful unless
it is accompanied by an element of anticompetitive conduct.’).
99
555 US 438, 454 (2009).
100
See, eg, Official Airline Guides, Inc. v. FTC, 630 F.2d 920, 927–28 (2d Cir. 1980) (ac-
knowledging, in a Section 5 case, ‘the long recognized right of trader or manufacturer … freely
to exercise his own independent discretion as to parties with whom he will deal’ and conclud-
ing that ‘even a monopolist, as long as he has no purpose to restrain competition or to enhance
or expand his monopoly, and does not act coercively, retains this right’).
101
United Brands (n 21) paras 248–254.
102
Guidance on the Commission’s enforcement priorities (n 22) para 7.
103
For a discussion of apparent legal standards for the offense of excessive pricing, see
Akman (n 68) 194–95.
104
See Glasner & Sullivan (n 41) 336–39 (critiquing the possibility that market definition
should be legally required even where not analytically helpful).
24
evidence of the exercise of market power must still be produced, and whatever
evidence is produced in that regard would seem better proof of the firm’s pos-
session of market power than the market definition exercise would support.
105
Perhaps this limits the role of market definition to the—presumably unusual—
situation in which the challenged exercise of market power is not itself a max-
imal exercise of the firm’s available market power.
V. QUESTIONS OF SIZE
While the previous discussion focused on traditional market power con-
cerns, monopolization and abuse of dominance are not necessarily so limited.
Other concerns, and other aspects of what it might mean to dominate or mo-
nopolize trade, can be usefully collected under the heading of a firm’s size.
These, too, may influence how relevant markets should be interpreted and de-
fined.
A. Does size matter?
As a thought experiment, suppose that fictional firm ZedCo has a 90%
share of a relevant market containing only it and one other competitor. This
market was defined by the HMT with a current-price baseline and a hypothe-
sized price increase of 5%, so if ZedCo’s rival were eliminated, ZedCo would
be able to raise its prices by at least 5%. Now suppose ZedCo pays a critical
input supplier to refuse future deliveries to its rival, eventually driving the rival
out of business. Does it matter, for monopolization or abuse of dominance pur-
poses, that ZedCo is a single-employee business with market power limited to
the isolated rural village in which it operates? Many would say yes: neither mo-
nopolization nor abuse of dominance feel like appropriate labels for the con-
duct of a firm this small,
106
even if the firm has substantial market power in its
limited scope of operation.
Now change the facts. Suppose the same conduct presents only a 3% chance
of driving the rival from the market and suppose further that successful exclu-
sion of the rival would only buy ZedCo the power to raise its prices by 0.001%.
105
See generally Louis Kaplow, ‘Why (Ever) Define Markets?’ (2010) 124 Harvard Law
Review 437, 466 (providing a related but broader critique of market definition).
106
White (n 62) 923 (‘[T]here must be some de jure or de facto lower limit on a defendant's
size and importance …. Antitrust enforcement should not care if Joe (the owner of Joe's
Unique Coffee Shop) insists that his manager Nora signs a “non-compete” clause when she
decides to leave and start her own eatery ….’).
25
Does it matter, for monopolization or abuse of dominance purposes, that
ZedCo is a tech giant with yearly revenue greater than the GDP of many coun-
tries? Again, many would say yes. Indeed, there is an economic basis for this
distinction. As Landes and Posner observed long ago, ‘actual economic injury’
is a function not only of the change in price ‘but also the amount of economic
activity over which the [change in price] occurs’.
107
Put another way, a modest
market power gain could have disastrous welfare consequences when multi-
plied across a large enough base of transactions.
Our intuition in this thought experiment contrasts with the limited role
that size plays in modern market definition practice. Relevant markets are not
regularly defined by processes designed to capture economically significant
scopes of trade.
108
Nor are they interpreted in these terms. True, consistent at-
tention to market shares keeps the relative size of firms in focus. But an over-
emphasis on relative size may actually obscure the role that absolute size—
amount of economic activity affected—also plays in explaining competitive
harm.
Particularly with the growing attention being paid to large tech companies,
matters of size may soon become a more important part of single firm conduct
cases. This is no justification for changing market definition standards—mak-
ing economic significance a necessary requirement of market definition, for
example. But it is excellent justification for making economic significance a
regular part of the interpretation of relevant markets and how the challenged
conduct would affect them.
B. Does size alone matter?
The previous discussion might undersell the role that size plays in monop-
olization and abuse of dominance concepts. One could take the position that
size alone is a proper concern in single firm conduct cases. Under this view,
firms too great in size are a danger themselves, separate from any market power
concerns.
109
Markets dominated by large firms may similarly be seen as
107
Landes & Posner (n 33) 953; see also Louis Kaplow, ‘On the Relevance of Market Power’
(2017) 130 Harvard Law Review 1303, 1361 n 130 (2017) (noting the potentially countervailing
concern that the costs of enforcement errors may also scale with firm size).
108
Sullivan (n 30) § II.A.
109
For a variety of arguments in this vein, see Tim Wu, The Curse of Bigness: Antitrust in
the New Gilded Age (2018).
26
deficient themselves, separate from any specific connection between concen-
tration and power in pricing or negotiation. Competition policy could simply
prefer more and smaller competitors in important areas of trade. Here, again,
growing agitation over large tech companies is surfacing many of these policy
positions in debates about enforcement practices and priorities.
It is not obvious that market definition is needed to apply ‘size alone’ con-
siderations. But, if relevant markets are to be used in the evaluating single firm
conduct under these policy goals, then new processes of market definition may
be required. Markets defined by a process like the HMT, though potentially
helpful in assessing the market power implications of different types of con-
duct, have no necessary relationship to these concerns. A firm’s control of 99%
of a relevant market defined by the HMT is simply not reliable evidence that
the firm is economically or politically significant, or that any other non-mar-
ket-power objective would be furthered by reduced concentration in this mar-
ket. Helpful relevant markets must be defined by processes matched to the con-
cern to be addressed. For policy objectives not recently at the forefront of com-
petition policy, there is little reason to suppose that previous market definition
processes match the concern.
VI. CONCLUSION
The aspiration of this chapter is to take some initial steps toward a more
coherent approach to market definition in monopolization and abuse of dom-
inance cases. In brief, the proposed approach is to match market definition
with the concerns to be addressed. If we want markets to be useful in evaluating
potential wrongs, then we must define markets by processes that are sensitive
to the wrongs we seek to avoid. For single firm conduct, this places the onus
on the substantive law to clearly identify what specific wrongs it seeks to avoid.
Greater clarity in substantive law is the most promising path to coherence in
this area of market definition.