Economics Points

Market Failures, Transaction Costs and Article
101(1)
TFEU Case Law
By
Pablo Ibanez Colomo
Reprinted from European Law Review
Issue 5, 2012
Sweet & Maxwell
100 Avenue Road
Swiss Cottage
London
NW3 3PF
(Law Publishers)
Market Failures, T ransaction Costs and Article
101(1) TFEU Case Law
Pablo Ibanez Colomo
*
Department of Law , London School of Economics and Political Science
Anti-competitive practices; Competition law; EU law; Inter-state effect
Abstract
EU competition law is incr easingly informed by economic analysis. However , the tools of this discipline
are rarely ever used systematically for positive purposes. This article gives a unifying picture of art.101(1)
TFEU judgments based on familiar economic concepts (market failur es and transaction costs). It is
submitted that a formalisation of case law based on these concepts has greater explanatory power than
the prevailing approaches currently found in textbooks and policy instruments. The article shows, first,
that the availability of an efficiency explanation for an agreement is the default starting point followed
by EU courts when drawing the line between restrictions of competition by object and by effect. Secondly ,
it explores how market failur es and transaction costs influence the analysis of restrictive effects on
competition under art.101(1) TFEU. These insights are equally useful to define the scope of art.101(3)
TFEU and the relevance of non-economic considerations.
Introduction
Economic analysis is increasingly relied upon to rationalise the enforcement of EU competition law . It is
unfortunate that this trend has not contributed to clarifying the scope of art.101(1) TFEU. The interest
with which recent judgments
1
have been received by commentators shows that the instances in which an
agreement is likely to be deemed to restrict competition by object are notoriously unclear .
2
This is not, to
be sure, the only remaining area of uncertainty relating to the interpretation of this provision. In particular,
*
Email: P .Ibanez-Colomo@lse.ac.uk. I am grateful to Eduardo Baistrocchi, Jarleth Burke, Martin Cave, Alfonso
Lamadrid de Pablo and the anonymous referee for their comments on previous drafts. Most of the ideas found in this
article originated while I was preparing my PhD under the guidance of Heike Schweitzer, to whom I am very much
grateful for our discussions.
1
T-Mobile Netherlands BV , KPN Mobile NV , Orange Nederland NV and V odafone Libertel NV v Raad van bestuur
van de Nederlandse Mededingingsautoriteit (C-8/08) [2009] E.C.R. I-4529; [2009] 5 C.M.L.R. 11; Competition
Authority v Beef Industry Development Society Ltd and Barry Brothers (Carrigmore) Meats Ltd (C-209/07) [2008]
E.C.R. I-8637; [2009] 4 C.M.L.R. 6; Football Association Premier League Ltd v QC Leisure; and Karen Murphy v
Media Protection Services Ltd (C-403/08 and C-429/08) [2012] Bus. L.R. 1321; and Pierre Fabre Dermo-Cosmétique
SAS v Président de l’Autorité de la concurrence and Ministre de l’Économie, de l’Industrie et de l’Emploi (C-439/09)
[2011] 5 C.M.L.R. 31.
2
Some of the recent articles addressing this question include A. Andreangeli, “From Mobile Phones to Cattle: How
the Court of Justice is Reframing the Approach to Article 101 (Formerly Article 81 EC) of the EU Treaty” (2011) 34
World Competition 215; D. Bailey, “Restrictions of Competition by Object under Article 101 TFEU” (2012) 49
C.M.L. Rev. 559; J. Goyder, “Cet Obscur Objet: Object Restrictions in V ertical Agreements” (2011) 2 Journal of
European Competition Law and Practice 327; A. Jones, “Left Behind by Modernisation? Restrictions by Object
Under Article 101(1)” (2010) 6 European Competition Journal 649. S. King, “The Object Box: Law, Policy or Myth?”
(2011) 7 European Competition Journal 269.
541 (2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
the question of whether, and to what extent, the pro-competitive effects of an agreement should be accounted
for under art.101(1) TFEU is perpetually contentious.
Arguably , the limited contribution of economic analysis to these debates has to do with the fact that
the tools of this discipline have only been used, by and large, for normative purposes, and in particular to
assist the Commission in the (re)definition of its enforcement priorities. The scope of art.101(1) TFEU
as defined by EU Courts is still explored, from a positive perspective, with the aid of an amalgam of
concepts that do not necessarily belong together . While the Commission now accepts that restrictive effects
on competition are unlikely to be manifested if the agreement does not create or strengthen market power,
3
it seems clear that this insight alone provides only a very partial understanding of art.101(1) TFEU case
law . As a result, authorities and commentators alike typically have recourse to an array of legal notions
when they seek to capture the logic underpinning the relevant judgments in its full complexity .
4
The purpose of this article is to further explore the potential of economic analysis to improve the
interpretation, prediction and criticism of judgments defining the scope of art.101(1) TFEU. It is submitted
that the use of standard tools can make a substantial contribution to the study of the patterns common to
the different rulings and to the identification of the exact terms and rationale of actual and alleged
contradictions between them.
It must be clarified at the outset that this analytical approach does not imply that economic notions were
expressly considered by EU Courts in their analysis, nor is it based on the claim that there is (or there
should be) a single overarching objective guiding intervention in the field. The fundamental idea underlying
the following pages is instead that EU Courts have regularly displayed a remarkably solid, if often intuitive,
understanding of the transactions examined under art.101(1) TFEU, and that the concepts used in
mainstream economics can help systematise this understanding.
Commentators have long noted that EU Courts tend to interpret this provision more flexibly than the
Commission.
5
However, they have rarely ever had recourse to economic analysis to make sense of this
divergence. In particular, the question of how market failures and transaction costs influence the outcome
of cases has rarely been examined systematically .
6
This is somewhat surprising considering that it is
3
See in this sense the statement of principle found in the Guidelines on the application of art.81(3) of the Treaty
[2004] OJ C101/97(Guidelines on the application of art.101(3) TFEU), para.25.
4
Authors very often refer to the so-called “ancillary restraints” doctrine (the idea that some restraints are not deemed
to restrict competition insofar as they are ancillary to a legitimate commercial transaction) to explain judgments such
as Metro SB-Großmärkte GmbH & Co KG v Commission (Metro I) (26/76) [1977] E.C.R. 1875; [1978] 2 C.M.L.R.
1; Remia BV v Commission (42/84) [1985] E.C.R. 2545; [1987] 1 C.M.L.R. 1; and Pronuptia de Paris GmbH v
Pronuptia de Paris Irmgard Schillgallis (161/84) [1986] E.C.R. 353; [1986] 1 C.M.L.R. 414. The ancillary restraints
doctrine is often relied upon to make sense of other cases, including Gøttrup-Klim Grovvareforeninger v Dansk
Landbrugs Grovvareselskab AmbA (C-250/92) [1994] E.C.R. I-5641; [1996] 4 C.M.L.R. 191; or J.C.J. Wouters, J.W .
Savelbergh and Price Waterhouse Belastingadviseurs BV v Algemene Raad van de Nederlandse Orde van Advocaten
(C-309/99) [2002] E.C.R. I-1577; [2002] 4 C.M.L.R. 27. For an overview of the different interpretations of this
doctrine, and different perspectives on its relevance in EU competition law , see in particular J. Goyder and A. Albors
Llorens, Goyder’ s Competition Law, 5th edn (Oxford: Oxford University Press, 2009), pp.123–126; R. Whish and
D. Bailey, Competition Law, 7th edn (Oxford: Oxford University Press, 2012), pp.129–134 (these authors prefer the
notion of “commercial ancillarity”); or G. Monti, EC Competition Law (Cambridge: Cambridge University Press,
2007), pp.33–36.
5
V . Korah, “EEC Competition Policy — Legal Form or Economic Efficiency” (1986) 39 Current Legal Problems
85; and R. Whish and B. Sufrin, “Article 85 and the Rule of Reason” (1987) 7 Y earbook of European Law 1. See also
B. Hawk, “System failure: V ertical restraints and EC competition law” (1995) 32 C.M.L. Rev. 973.
6
Bailey, “Restrictions of Competition by Object under Article 101 TFEU” (2012) 49 C.M.L. Rev. 559, refers to
market failures as a relevant factor in the conclusion, but does not systematically explore its role in the body of the
paper (the primary purpose of which seems to be a different one). Lianos, in turn, relied on Oliver Williamson’ s work
on transaction costs to interpret the flexibility shown by the ECJ in some vertical restraints cases. See I. Lianos, La
transformation du droit de la concurrence par le recours à l’analyse economique (Brussels: Bruylant, 2007).
542 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
nowadays well established that firm co-operation (in particular in the context of vertical restraints) may
be an effective means to address the former or to reduce the latter .
7
The formalisation of case law principles on the basis of these insights is illuminating in several respects.
It is an approach that provides an interpretation of the notion of restriction of competition by object which,
it is submitted, has greater explanatory power than prevailing ones. Similarly , an analysis of how market
failures and transaction costs influence the evaluation of restrictions of competition sheds light on the
reasons why EU Courts have regularly balanced the pro- and anti-competitive effects of agreements under
art.101(1) TFEU (in spite of occasional statements to the contrary).
Because the definition of the scope of art.101(1) TFEU has direct implications for the interpretation
and implementation of art.101(3) TFEU, this line of analysis is equally useful to define the precise
boundaries of the latter . By the same token, this reading of the case law improves the understanding of
the exact impact of public policy considerations on the analysis of restrictive agreements. T o the extent
that such non-competition concerns very often constitute a reaction to a market failure, they are not,
contrary to what is often suggested or implied, necessarily in conflict with the core objectives of competition
law .
Market failures, transaction costs and agreements between undertakings
In the presence of a market failure, the unrestricted activity of private operators will not lead to an efficient
allocation of resources.
8
A similar outcome may result where the costs of concluding a transaction exceed
the benefits that the parties derive from it (i.e. “transaction costs”).
9
In these circumstances, different forms
of government intervention (e.g. subsidies or taxes in the case of market failures, a bright-line rule that
eliminates the costs of bargaining) could in theory improve the functioning of markets. However, it is
well established that private co-operation routinely addresses such imperfections. Competition authorities
have come to understand that many restrictions that appear to be one-dimensional anti-competitive devices
are, upon more sophisticated scrutiny , a source of efficiency gains. V ery often, the presence of market
failures or transaction costs provides the key motivation behind these arrangements.
Market failur es and competition law
Economic literature identifies several sources of market failure, one of which—market power—would
alone justify the setting up of a competition law system.
10
Some other prominent examples of market
failures include the following:
• Externalities:
Externalities arise where a firm does not bear the full costs (negative externality) or benefits
(positive externality) of its economic activity . A negative externality is a source of allocative
7
Suffice it to examine in this regard a standard book on the economics of competition law , such as S. Bishop and
M. Walker, The Economics of EC Competition Law: Concepts, Application and Measurement, 3rd edn (London:
Sweet & Maxwell, 2010), pp.190–198. See also, inter alia, D. Carlton and J. Perloff, Modern Industrial Organization,
4th edn (Boston: Addison-Wesley , 2005), pp.395–412.
8
On the notion of market failure in general, see e.g. R. Cooter and T . Ulen, Law & Economics, 6th edn (Upper
Saddle River, NJ: Prentice Hall, 2012), pp.38–41, and G. Mankiw, Principles of Economics, 6th edn (Stamford, CT:
Southwestern, 2012), pp.135–151 and 195–229.
9
Cooter and Ulen, Law & Economics (2012), pp.84–91.
10
Market power is a source of allocative inefficiency insofar as it allows firms to raise prices above the levels that
would prevail in a competitive market and to reduce output below the social optimum. By ensuring that the conditions
of competition in a given market are not further deteriorated, competition law can be said to be guided by efficiency
and consumer welfare considerations.
Market Failures, Transaction Costs and Article 101(1) TFEU 543
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
inefficiency insofar as the good or service subject to the phenomenon will tend to be
oversupplied. T extbook examples of negative externalities include air pollution or traffic
congestion. Positive externalities, in turn, tend to be undersupplied. Classic examples in this
sense include the training of its employees by a firm.
• Public goods:
T ypically presented, for analytical purposes, as an extreme positive externality , pure public
goods are those presenting two features, that is, non-rivalry in use and non-excludability .
Goods are said to be non-rivalrous in use where consumption by one user does not prevent
it from being consumed simultaneously by other users. Think of the performance of a song,
which can be listened to by an infinite number of users. By non-excludable it is meant that
it is difficult for the producer of the good to prevent its unauthorised consumption. A classic
example of a pure public good is (was) that of terrestrial (over-the-air) television, which can
(if decoding technologies are not used) be accessed by any household.
• Information asymmetries:
This concept depicts a situation in which the information about the features of the product
subject to exchange is not evenly distributed among buyers and sellers. The archetypal
example is that of the seller of a good that has more information about its quality than does
its buyer .
11
A good example in this sense concerns the financial sector . As a rule, banks know
more about the details and implications of the contracts they conclude with end-users. The
opposite may also be true. Buyers of insurance products know more about their risk profile
than do sellers. The asymmetry of information may significantly affect the rate of transactions
in an industry .
12
Private transactions responding to market failures have regularly come to the attention of competition
authorities. V ertical restraints (as much as vertical integration) constitute an appropriate mechanism to
address the positive externalities generated by one of the parties to a distribution agreement. Exclusive
distribution agreements,
13
for instance, are effective to prevent free-riding by other resellers insofar as
they allow the appointed distributor to internalise the positive externalities generated by its investments
(e.g. promotional efforts).
14
Exclusive dealing agreements,
15
in turn, may be an adequate means to avoid
free-riding from other suppliers at the downstream level.
16
Agreements within the meaning of art.101(1) TFEU are understood to be equally effective in preserving
the value of intangible property . T ake the example of non-competition (restrictive covenant) clauses. T o
the extent that there is a goodwill attached to a business, such clauses can effectively ensure that the
acquiring party is able to fully appropriate the (intangible) assets that are being transferred.
17
Similarly , a
11
This is the example referred to by Cooter and Ulen, Law & Economics (2012), pp.41–42 and the one used in the
classic work by G. Akerlof, “The Market for ‘Lemons’: Quality Uncertainty and the Market Mechanism” (1970) 84
Quarterly Journal of Economics 488.
12
Cooter and Ulen, Law & Economics (2012), pp.48–49.
13
Guidelines on vertical restraints, para.151.
14
Guidelines on vertical restraints, para.107.
15
Exclusive dealing agreements (referred to as single branding agreements) are defined in para.129 of the Guidelines
on vertical restraints as “those agreements which have as their main element the fact that the buyer is obliged or
induced to concentrate its orders for a particular type of product with one supplier”.
16
Guidelines on vertical restraints, para.107.
17
For an economic analysis of the role of restrictive covenant clauses, see M. Trebilcock, The Common Law of
Restraint of Trade: A Legal and Economic Analysis (London: Sweet & Maxwell, 1986), pp.252–267.
544 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
selective distribution agreement
18
may be relied upon by a manufacturer to guarantee that the desired
uniform brand image of its products is preserved. The rationale behind the restrictions included in a
franchise agreement is not fundamentally different.
19
Firm co-operation alone can successfully address the negative consequences of asymmetric information.
In the Guidelines on horizontal co-operation agreements issued in 2010, the Commission acknowledges
that information exchange agreements may be a source of efficiency gains in this sense.
20
For instance,
the authority explains that the exchange of information in the insurance sector may allow competing firms
to identify the risk profile of end-users and thus avoid adverse selection.
21
Information asymmetries plague
markets for professional services.
22
As a consequence, deontological measures adopted by a professional
association may have a positive impact on allocative efficiency .
23
T ransaction costs and competition law
Transaction costs can be defined as the costs incurred by firms (and consumers) in economic exchanges
between them.
24
It is commonplace to see in literature a further distinction between three broad categories
of transaction costs, that is, between (1) search costs (i.e. the costs of finding a commercial partner selling
the good in question); (2) bargaining costs (i.e. those incurred to strike a deal); and (3) enforcement costs
(i.e. those that relate to ensuring that the terms of the deal are respected).
25
As hinted at above, these costs
may prevent the efficient reallocation of resources if some transactions are not concluded.
It has long been understood that transaction costs are essential in understanding the organisation of an
industry . Their role may explain, for instance, why suppliers are integrated with distributors in a given
sector but not in another one. Therefore it is only logical that the reduction of transaction costs is mentioned
as one of the potential efficiency gains that may result from vertical integration in the Non-Horizontal
Merger Guidelines.
26
T o the extent that vertical restraints are an imperfect substitute for a merger,
distribution agreements may lead to the same gains. For instance, vertical restraints may constitute an
effective means to induce one of the parties to the agreement to engage in the necessary relationship-specific
investments.
27
18
Pursuant to art.1(1)(e) of Commission Regulation 330/2010 on the application of art.101(3) of the Treaty on the
Functioning of the European Union to categories of vertical agreements and concerted practices [2010] OJ L102/1,
a selective distribution system is one “where the supplier undertakes to sell the contract goods or services, either
directly or indirectly , only to distributors selected on the basis of specified criteria and where these distributors
undertake not to sell such goods or services to unauthorised distributors within the territory reserved by the supplier
to operate that system”.
19
As explained in para.189 of the Guidelines on vertical restraints, franchise agreements “contain licences of
intellectual property rights relating in particular to trade marks or signs and know-how for the use and distribution of
goods or services”, and, in addition, “usually contain a combination of different vertical restraints concerning the
products being distributed, in particular selective distribution and/or non-compete and/or exclusive distribution or
weaker forms thereof”.
20
Guidelines on the applicability of art.101 of the Treaty on the Functioning of the European Union to horizontal
co-operation agreements [2011] OJ C11/1 (Guidelines on horizontal co-operation agreements), paras 95–100.
21
Guidelines on horizontal co-operation agreements, paras 95–100, in particular para.97.
22
See European Commission, Report on Professional Services (Brussels, February 9, 2004) and “Follow-up to the
Report on Competition in Professional Services” COM(2004) 83 of February 9, 2004 (Brussels, September 5, 2005).
23
See e.g. R. V an den Bergh and Y . Montangie, “Competition in Professional Services Markets: Are Latin Notaries
Different” (2006) 2 Journal of Competition Law and Economics 189.
24
Cooter and Ulen, Law & Economics (2012), p.87.
25
Cooter and Ulen, Law & Economics (2012), p.88
26
Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of
concentrations between undertakings [2008] OJ C256/6 (Non-Horizontal Merger Guidelines), para.14.
27
Guidelines on vertical restraints, para.107(d).
Market Failures, Transaction Costs and Article 101(1) TFEU 545
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
Horizontal co-operation agreements and similar arrangements may also reduce transaction costs. An
oft-cited example in this regard, and one of the few in which these gains are explicitly acknowledged by
the Commission, is that of standard-setting agreements. An agreement between competitors on the “technical
or quality requirements with which current or future products, production processes, services or methods
may comply”
28
may be pro-competitive precisely because it may reduce transaction costs for both buyers
and sellers in the relevant market.
29
Similar gains can result from the joint exploitation of technologies
through a patent pool, which amounts in essence to the creation of a “one-stop shop” that reduces third
parties’ uncertainty about their ability to exploit a particular technology .
30
Other examples in this sense
include agreements for the joint licensing of television rights by football teams through a league, or the
activities of copyright collecting societies.
Efficiency gains and the object/effect divide
Recent literature reflects the struggle to identify an operational definition of what amounts to a restriction
of competition by object within the meaning of art.101(1) TFEU. If the hard core and the underpinning
logic of this notion are relatively clear, the delineation of its outer boundaries remains a distinctly complex
task. The multiplicity of methodologies adopted by EU Courts has contributed significantly to this confusion.
The resulting picture is one in which judgments are sometimes mutually contradictory , and in which the
outcome of cases is not obvious to anticipate. The disagreement between the General Court and the ECJ
concerning the status under art.101(1) TFEU of the dual pricing system to restrict parallel trade at stake
in Glaxo Spain exemplifies this reality particularly well.
31
If one accounts for the role of market failures, transaction costs, and, more generally , for
efficiency-enhancing transactions in the analysis of restrictions of competition, it becomes possible to
obtain a more clear and coherent picture of the default approach of the ECJ to this matter, which seems
to revolve around the question of whether the restrains found in an agreement can be plausibly explained
on efficiency grounds. This default approach seems to be altered where resale price maintenance clauses
and certain parallel trade restrictions are at stake.
The flaws of prevailing interpr etations
Commentators, as much as the Commission, tend to rely on the form of the restraints found in an agreement
to draw the line between restrictions by object and by effect. This approach to the question requires little
more than examining whether the agreement under examination features one of the pre-established formal
categories of “hardcore” restraints such as price fixing or market sharing. One can find traces of this
interpretation in several soft law instruments of the Commission, such as the Guidelines on the application
28
Guidelines on horizontal co-operation agreements, para.257.
29
Guidelines on horizontal co-operation agreements, paras 308, 312 and 323.
30
Guidelines on the application of art.81 of the EC Treaty to technology transfer agreements [2004] OJ C101/2
(Guidelines on technology transfer agreements), para.210. Patent pools are defined as “arrangements whereby two
or more parties assemble a package of technology which is licensed not only to contributors to the pool but also to
third parties …”. See also para.214, where the Commission clarifies that “[t]echnology pools can also produce
pro-competitive effects, in particular by reducing transaction costs and by setting a limit on cumulative royalties to
avoid double marginalisation”, and para.217.
31
Compare the analysis in GlaxoSmithKline Services Unlimited v Commission (Glaxo Spain) (C-501/06 P , C-513/06
P , C-515/06 P and C-519/06 P) [2009] E.C.R. I-9291; [2010] 4 C.M.L.R. 2; and GlaxoSmithKline Services Unlimited
v Commission (T-168/01) [2006] E.C.R. II-2969; [2006] 5 C.M.L.R. 29.
546 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
of art.101(3) TFEU,
32
but its logic is perhaps best captured by Whish, who graphically identifies an “object
box”, based on the form of the restraints alone.
33
The limits of this methodology are widely acknowledged. Agreements that do not formally contain any
restraints pertaining to the “object box” may be deemed to restrict competition by object. A clear example
in this sense is ANSEAU-NAVEW A, where the parties to the agreement had used what appeared to be a
standard-setting arrangement as a device to insulate themselves from foreign competition.
34
Likewise, in
Beef Industry Development the ECJ reiterated that there is no such thing as a closed category of object
restrictions.
35
Conversely, there are numerous examples of agreements that were not found to restrict competition by
their very nature even though they featured restraints that are, in theory , straightforward “no-nos”. In
Gøttrup-Klim, the joint purchasing agreement at stake was not deemed to restrict competition by its very
nature in spite of the fact that horizontal price fixing is a necessary ingredient of an arrangement of this
kind.
36
Similarly , the contested behaviour in T ournier (a refusal by the members of a copyright collecting
society to license parts of their repertoire) very much looks like a co-ordinated restriction of output.
37
In
Coditel II,
38
the ECJ concluded that the exclusive territorial licensing agreement at stake in the case was
not restrictive of competition by object even though it gave absolute territorial protection to the copyright
licensee and thus contradicted (on the surface) a consistent line of case law dating back to
Consten-Grundig.
39
Similar flexibility has been shown by the ECJ in cases providing for territorial protection
to the licensees of a technology such as Nungesser or Erauw-Jacquery.
40
Several policy positions taken by the Commission suggest in fact that a form-based approach to identify
object restrictions may only be meaningful, if at all, when price fixing, market sharing and similar restraints
are presented in their “naked” form. When such restraints are instead an integral and necessary ingredient
of an agreement that has redeeming virtues from an efficiency standpoint, it seems incorrect to assume
that they amount to object restrictions. According to the Guidelines on horizontal co-operation agreements,
for instance, the parties to a specialisation agreement may decide to sell their inputs jointly (which by
definition involves price fixing) without this violating art.101(1) TFEU by its very nature.
41
This is also
the case where the parties to a research and development agreements decide to exploit in common the
outcome of their joint efforts.
42
32
Guidelines on the application of art.101(3) TFEU, para.23.
33
Whish and Bailey, Competition Law (2012), in particular p.124.
34
NV IAZ International Belgium v Commission (ANSEAU-NA VEWA) (96–102, 104, 105, 108 and 110/82) [1983]
E.C.R. 3369. Likewise, in Dansk Pelsdyravlerforening, the CFI came to the conclusion that the exclusive supply
obligations at stake in the case were restrictive of competition by object. See Dansk Pelsdyravlerforening v Commission
(T-61/89) [1992] E.C.R. II-1931.
35
Beef Industry Development (C-209/07) [2008] E.C.R. I-8637 at [23].
36
Gøttrup-Klim (C-250/92) [1994] E.C.R. I-5641, in particular, at [32]. See also A. Ezrachi, “Buying Alliances and
Input Price-Fixing: In Search of a European Enforcement Standard” (2012) 8 Journal of Competition Law and
Economics 47.
37
Ministère Public v Jean-Louis T ournier (395/87) [1989] E.C.R. 2521.
38
Coditel SA, Compagnie générale pour la diffusion de la télévision v Ciné-V og Films SA (Coditel II) (262/81)
[1982] E.C.R. 3381; [1983] 1 C.M.L.R. 49, in particular at [15].
39
Établissements Consten Sàrl and Grundig-V erkaufs-GmbH v Commission (Consten-Grundig) (56/64 and 58/64)
[1966] E.C.R. 429; [1966] C.M.L.R. 418.
40
See in particular L.C. Nungesser KG and Kurt Eisele v Commission (258/78) [1982] E.C.R. 2015; [1983] 1
C.M.L.R. 278; and SPRL Louis Erauw-Jacquery v La Hesbignonne SC (27/87) [1988] E.C.R. 1919.
41
Guidelines on horizontal co-operation agreements, para.160. The US Supreme Court took the same stance in
T exaco Inc v Dagher 547 U.S. 1 (2006), where it concluded that a joint venture may set the price of its products
without this constituting a per se restraint within the meaning of s.1 of the Sherman Act.
42
Concerning research and development agreements, for instance, the Commission is careful to clarify in para.128
of the Guidelines that “R&D agreement which includes the joint exploitation of possible future results [which has a
Market Failures, Transaction Costs and Article 101(1) TFEU 547
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
Alternatively, it is regularly claimed that the category of object restrictions encapsulates a legal
presumption about the predicted anti-competitive effects of some agreements. This perspective is based
on the idea that some restraints, including price fixing and output restrictions, are so likely to have negative
effects on competition that it is not necessary to establish these on a case-by-case basis.
43
The
above-mentioned category would thus provide an administrable rule reducing the enforcement costs for
those restraints that would almost invariably restrict competition.
This approach, while intuitively appealing, is equally inconclusive. Where an agreement is deemed to
restrict competition by object, it is caught by art.101(1) TFEU even where it cannot be expected to have
significant restrictive effects on prices or output.
44
Thus a price fixing agreement aimed to sustain an
unstable cartel will be prohibited irrespective of whether the parties have the ability and the incentive to
enforce it on a sustainable basis. Similarly , a naked price fixing agreement between relatively small
competitors is prohibited outside the marginal instances where the agreement is found to be de minimis.
45
More generally , this interpretation is problematic for the same reasons that a form-based approach is
inconclusive. T o the extent that formal categories such as price fixing or market sharing are in themselves
meaningless for the purposes of identifying restrictions by object, they cannot be relied upon to formulate
a workable presumption.
Finally , it is now difficult to dispute that the subjective intent of the parties alone is insufficient to draw
the line between restrictions by object and by effect. While it is undeniable that the ECJ has frequently
relied on subjective factors to conclude that an agreement restricts competition by its very nature, it is
equally clear, as clarified in Beef Industry Development, that the absence of anti-competitive intent will
not be sufficient to assist the parties, and that the anti-competitive object of an agreement can be inferred
from objective factors alone.
46
A fr esh economic perspective on the object/effect divide
The approach followed by the ECJ to identify object restrictions and distinguish them from agreements
requiring a case-by-case assessment is generally not guided by the formal features of the agreement under
examination. This becomes apparent when one expands the universe of the cases considered beyond those
where the relevant restraints were found to restrict competition by their very nature. The ECJ has
consistently insisted on the need to take account of the nature of the agreement, of the objectives it pursues
and, in some cases, of the economic context in which it is concluded, before establishing that it is indeed
restrictive of competition by its very nature.
47
price fixing ingredient by definition] is not necessarily restrictive of competition”. See also art.5 of Commission
Regulation 1217/2010 on the application of art.101(3) of the Treaty on the Functioning of the European Union to
certain categories of research and development agreements [2010] OJ L335/36.
43
Guidelines on the application of art.101(3) TFEU para.21. See also, e.g. A. Jones and B. Sufrin, EU Competition
Law — T ext, Cases and Materials (Oxford: Oxford University Press, 2011), p.203 and, more recently , R. Whish,
“Object Restrictions”, New Frontiers of Antitrust Conference (February 10, 2012), http://www .concurr ences.com
[Accessed September 5, 2012].
44
T-Mobile (C-8/08) [2009] E.C.R. I-4529 at [30]–[31] and [36].
45
For a discussion, see J. Faull and A. Nikpay (eds), The EC Law of Competition, 2nd edn (Oxford: Oxford University
Press, 2007), pp.227–228. One can rightly presume that such an agreement is unlikely to be observed in practice,
insofar as it cannot be expected to achieve little more than harming the position of the parties themselves (see H.
Hovenkamp, The Antitrust Enterprise (Cambridge, MA: Harvard University Press, 2005), p.112: “price-fixing is
profitable only if the participants collectively occupy enough of the market to make their price increase stick”) but
this fact does not mean they will not be prohibited if concluded.
46
See in particular Beef Industry Development (C-209/07) [2008] E.C.R. I-8637 at [21].
47
See in particular ANSEAU-NAVEWA (96–102, 104, 105, 108 and 110/82) [1983] E.C.R. 3369 at [23]–[25];
Compagnie Royale Asturienne des Mines and Rheinzink GmbH v Commission (29/83 and 30/83) [1984] E.C.R. 1679;
548 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
This legal principle can be readily formalised as meaning that a given restraint will be deemed to restrict
competition by object only to the extent that, and in those instances where, it is not a plausible source of
efficiency gains. The true question does not seem to be whether the restraint in question can be presumed
to have anti-competitive effects, or whether it bears a particular form; but whether, in the light of the
nature of the agreement, and the context in which it is concluded, it is a convincing means to enhance
efficiency and not simply a means to extract wealth from customers or suppliers. Put differently , the crucial
factor is not that the agreement can be presumed to deteriorate the conditions of competition on the relevant
market(s), but the fact that it cannot be expected to improve them.
It is precisely where the relevant restraint has no redeeming virtues, as is the case of those sustaining
a “naked” price fixing cartel, that anti-competitive intent can be safely presumed. Where the agreement
cannot be reasonably explained on efficiency grounds, or where there is not a clear link between the
efficiency claims and the relevant restraints (as was the case, for instance, in Beef Industry Development),
48
one can rightly assume that the primary motivation of the parties is to restrict competition. This may help
explain why the ECJ seems to give relevance to subjective factors in some cases.
An overview of the case law suggests that, where the parties to the agreement are in a vertical (e.g.
supplier/distributor; licensor/licensee) relationship, an analysis of the nature of the agreement in the abstract
seems to be sufficient to draw conclusions about the plausibility of the efficiency explanation for the
restraints found therein. Where the agreement is horizontal, on the other hand, the ECJ seems to put more
emphasis on the careful assessment of the “context” in which it is concluded, which involves considering
the features of the relevant market, the relative position of the parties or the nature of the products concerned
by the agreement. This pattern seems to apply irrespective of the source of the efficiency gains. The
paragraphs below examine, in turn, instances in which market failures, productive efficiencies and
transaction costs were relevant in the analysis.
Consider, first, the example of landmark cases examining the compatibility of vertical restraints with
art.101(1) TFEU, including Société T echnique Minière,
49
Metro I and Pronuptia. In all of these judgments,
the ECJ started its analysis under art.101(1) TFEU by identifying the (efficiency enhancing) reasons why
the parties would agree to the restraints under consideration. The fact that these were, in the abstract, a
plausible source of efficiency gains, was sufficient to conclude that they were not restrictive by object.
In Société T echnique Minière and Nungesser, the ECJ explored the reasons why exclusive territorial
protection may be a necessary means to induce the distributor or licensee, respectively , to engage in the
necessary investments to resell or manufacture a product.
50
In this sense, the ECJ clearly grasped the
essence of the free-rider argument. The role of externalities was also prominent in other contexts, such as
the one at stake in Remia, which is a notable judgment in that the ECJ overtly acknowledged the
output-expansion virtues of restrictive covenant clauses.
51
[1985] 1 C.M.L.R. 688 at [26]; and General Motors BV v Commission (C-551/03 P) [2006] E.C.R. I-3173; [2006] 5
C.M.L.R. 1 at [66].
48
In this case, the parties to the agreement claimed that co-ordinated capacity reductions in the relevant industry
were a source of efficiency gains (see [13] of the judgment). As the ECJ pointed out in the judgment, however, such
a reduction would have been the result of the normal play of market forces and thus in the absence of the agreement
in question, the main point of which was, it would seem, to share the burden in the transition to profitability . See Beef
Industry Development (C-209/07) [2008] E.C.R. I-8637 at [34]–[35].
49
Société T echnique Minière v Maschinenbau Ulm GmbH (56/65) [1966] E.C.R. 235; [1966] C.M.L.R. 357.
50
Nungesser is very explicit in this regard. See in particular Nungesser (258/78) [1982] E.C.R. 2015 at [57] of the
judgment.
51
Remia (42/84) [1985] E.C.R. 2545, para.19: “Should the vendor and the purchaser remain competitors after the
transfer , it is clear that the agreement for the transfer of the undertaking could not be given effect. The vendor, with
his particularly detailed knowledge of the transferred undertaking, would still be in a position to win back his former
customers immediately after the transfer and thereby drive the undertaking out of business. Against that background
non-competition clauses incorporated in an agreement for the transfer of an undertaking in principle have the merit
Market Failures, Transaction Costs and Article 101(1) TFEU 549
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
Pronuptia and Metro I, in turn, are suggestive of a solid understanding of the importance of some
restraints to preserve the value of the intangible property that is associated with the products covered by
the selective distribution and franchise agreements. It is probably in Pierre Fabre, however, that this
understanding transpires most clearly . The ECJ held that the restraints found in selective distribution
agreements amount, in the absence of an “objective justification” (in the absence of a clear link between
the restraints and the market failure identified, one would understand), to a restriction of competition by
object.
52
The position of the ECJ in Coditel II is equally illuminating about its approach to restrictions by object.
The Court clarified that the rules that apply to those instances where the intellectual property is embodied
in a tangible good, in which case the intellectual property rights they incorporate are subject to the
exhaustion principle, cannot be extended as such to those in which it is exploited in the form of a public
good, such as a broadcast.
53
The value of a broadcast (which, as mentioned in s.2, is non-rival in use) to
a licensee may depend on its ability to prevent other operators from exploiting it at the same time (that is,
on its excludability). As a result, the licensing agreement may provide for absolute territorial protection
without it being restrictive by object.
Consider now the case of information sharing agreements. Because these agreements are concluded
between competitors, efficiency may not always be the most compelling explanation for the observed
conduct. An analysis of the context in which the agreement is concluded is thus deemed necessary . In
John Deere, the ECJ and the GC did not deny that, as a matter of principle, information exchanges may
improve the conditions of competition on a particular market.
54
It is clear from Asnef-Equifax that, where
this is the case, they may fall outside the scope of art.101(1) TFEU altogether .
55
On the other hand, these
agreements may be restrictive of competition by object where, as confirmed in T-Mobile, they afford the
parties the possibility of “removing uncertainties concerning the intended conduct of the participating
undertakings”,
56
or, if one prefers, where the behaviour of the parties is not a proportionate response to
potential market failures (typically related, as seen above, to information asymmetry concerns).
When the above is considered, it appears that it no longer seems appropriate to think of, and interpret,
the judgment in W outers as a rara avis, as is very often done.
57
The ECJ examined the deontological rule
of ensuring that the transfer has the effect intended. By virtue of that very fact they contribute to the promotion of
competition because they lead to an increase in the number of undertakings in the market in question.”
52
Pierre Fabre (C-439/09) [2011] 5 C.M.L.R. 31 at [39]. The need for an “objective justification” for the restraints
found in selective distribution systems is not new . See Groupement d’achat Édouard Leclerc v Commission (T-88/92)
[1996] E.C.R. II-1961 at [112]. More recently , in Auto 24, the ECJ considered that the supplier is under no obligation
to make public the criteria on the basis of which selective distributors are selected, as these may compromise “business
secrets”. See Auto 24 Sàrl v Jaguar Land Rover France SAS (C-158/11) [2012] 5 C.M.L.R. 3 at [31].
53
In Coditel II (262/81) [1982] E.C.R. 3381; [1983] 1 C.M.L.R. 49 at [10], the ECJ carefully explained that “the
problems involved in the observance of a film producer’ s rights in relation to the requirements of the Treaty are not
the same as those which arise in connection with literary and artistic works the placing of which at the disposal of
the public is inseparable from the circulation of the material form of the works, as in the case of books or records,
whereas the film belongs to the category of literary and artistic works made available to the public by performances
which may be infinitely repeated …”.
54
John Deere Ltd v Commission (C-7/95 P) [1998] E.C.R I-3111; [1998] 5 C.M.L.R. 311 at [67]. See also John
Deere v Commission (T-35/92) [1994] E.C.R. II-957 at [51].
55
Asnef-Equifax, Servicios de Información sobre Solvencia y Crédito, SL and Administración del Estado v Asociación
de Usuarios de Servicios Bancarios (C-238/05) [2006] E.C.R. I-11125; [2007] 4 C.M.L.R. 6 at [55]–[56].
56
John Deere (C-7/95 P) [1998] E.C.R I-3111 at [90]. See also T-Mobile (C-8/08) [2009] E.C.R. I-4529 at [35].
57
Just to mention some positions, see I. Forrester, “Where Law Meets Competition: Is Wouters Like a Cassis de
Dijon or a Platypus?” in C.-D. Ehlermann and I. Atanasiu (eds), European Competition Law Annual 2004: The
Relationship Between Competition Law and the (Liberal) Professions (Oxford: Hart Publishing, 2006), pp.271–294;
G. Monti, “Article 81 EC and Public Policy” (2002) 39 C.M.L. Rev. 1057 (who talks about a “European rule of
reason”); and the general discussion in Whish and Bailey, Competition Law (2012), pp.130–133.
550 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
at stake in that case and concluded that, in the specific economic and, in particular, regulatory context in
which it was implemented (it must be noted that, again, it involved competing undertakings),
58
it was a
plausible and proportionate means to ensure that lawyers in the Netherlands perform their task independently
and in the sole interest of the client.
59
In economic parlance, these concerns clearly relate to the information
asymmetries that, as mentioned above, are typically present in professional services markets. This same
position was already sketched by the GC in the EPI judgment of 2001, in which virtually identical
vocabulary was used in relation to a deontological rule prohibiting comparative advertising.
60
A similar methodology is followed where the agreement is understood to be a source of productive
efficiency gains. It is useful to illustrate these ideas by reference to both a vertical and a horizontal
agreement. In Delimitis,
61
for instance, the ECJ explicitly referred to the fact that exclusive dealing allows
producers to secure supplies (and thus plan their production more effectively and expand output) and the
retailer to obtain better contractual conditions (e.g. in the form of rebates). It is on the basis of these insights
that it reached a conclusion on the inappropriateness of deeming exclusive dealing agreements restrictive
by object.
62
The position of the ECJ in Gøttrup-Klim, in turn, can be formalised as meaning that, in light
of the context in which the (horizontal) joint purchasing agreement was concluded—a market in which
the price paid for supplies depends on the volume of orders and in which purchasers are in a relatively
weak bargaining position—the restraints examined could be convincingly explained by factors other than
the desire to gain market power and extract rents from suppliers.
63
These same conclusions can be extended to those instances in which the agreement leads to significant
gains in terms of transaction cost reductions. In T ournier, the ECJ held that the refusal by a copyright
collecting society to license only part of its repertoire does not necessarily restrict competition within the
meaning of art.101(1) TFEU even though it can be said to amount, as mentioned above, to a co-ordinated
reduction of output.
64
A reading of the judgment makes it clear that the search and enforcement cost gains
that result from the activity of collecting societies were not only acknowledged and understood by the
Court, but determined the outcome of the case.
65
Similarly , the US Supreme Court examined the licensing
activities of collecting societies in Broadcasting Music Inc v Columbia Broadcasting System. Even though
the joint licensing of intellectual property rights to end-users through a single point of sale is by definition
tantamount to price fixing, Justice White found it inappropriate to treat such an arrangement as a “naked”
price fixing cartel. The primary reason for this is that such a form of horizontal co-operation seems
indispensable to enable copyright holders to enforce their rights, and thus to offer what could be seen as
a new product on the market.
66
58
This context was relevant in several respects. For instance, the ECJ noted at [91] of Wouters v Algemene Raad
van de Nederlandse Orde van Advocaten (C309/99) [2002] E.C.R. I-1577; [2002] 4 C.M.L.R. 27 that the market for
accounting services was highly concentrated, both in absolute terms and in relation to lawyers. At [103], in turn, the
ECJ took account of the fact that accountants were not subject to a similar deontological regime.
59
Wouters (C309/99) [2002] E.C.R. I-1577 at [105]–[110].
60
Institute of Professional Representatives before the European Patent Office v Commission (T-144/99) [2001]
E.C.R. II-1087; [2001] 5 C.M.L.R. 2 at [70]–[79].
61
Stergios Delimitis v Henninger Bräu AG (C-234/89) [1991] E.C.R. I-935; [1992] 5 C.M.L.R. 210.
62
Delimitis (C-234/89) [1991] E.C.R. I-935 at [10]–[12].
63
Gøttrup-Klim (C-250/92) [1994] E.C.R. I-5641 at [31]–[34].
64
T ournier (395/87) [1989] E.C.R. 2521 at [31].
65
T ournier (395/87) [1989] E.C.R. 2521, in particular at [29] and [31].
66
Broadcast Music, Inc v Columbia Broadcasting System, Inc 441 U.S. 1 (1979): “[The] substantial lowering of
costs, which is of course potentially beneficial to both sellers and buyers, differentiates the blanket license from
individual use licenses. The blanket license is composed of the individual compositions plus the aggregating service.
Here, the whole is truly greater than the sum of its parts; it is, to some extent, a different product.”
Market Failures, Transaction Costs and Article 101(1) TFEU 551
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
Parallel trade and intent
There are two well-defined instances in which the starting point (and crucial factor) in distinguishing
between restrictions by object and by effect tends not to be the availability of an efficiency explanation
for the agreement. One of them concerns agreements restricting parallel trade among EU Member States.
These are agreements that provide for absolute territorial protection to a distributor or that expressly
provide for export prohibitions. As pointed out above, there is a consistent line of case law pursuant to
which they are deemed restrictive of competition by their very nature, and this, irrespective of the efficiency
claims made by the parties. The reasons behind this position are well established. In Glaxo Spain, the ECJ
held that the integration of markets is an objective of EU competition law worthy of being protected in
and of itself, and thus not merely as a proxy for other aims.
67
When establishing restrictions to market integration, the ECJ tends to follow an alternative methodology
that seems to give particular weight to the intent of the undertakings. In Consten-Grundig, the analysis
focused on the fact that the parties had conceived a device to insulate the buyer from competition coming
from distributors established in other Member States.
68
Similarly , in Miller International, the ECJ took
the view that agreements providing for export prohibitions are restrictive of competition by their very
nature, and this, insofar as the “agreed purpose of the contracting parties is the endeavour to isolate a part
of the market”.
69
In Glaxo, where the doctrine is reiterated, the ECJ referred, in more general terms, to
agreements “aimed at prohibiting or limiting parallel trade”.
70
Intent alone does not seem to be a reliable indicator to establish violations of art.101(1) TFEU, insofar
as it fails to capture the complexity of the case law . Think in this sense of the reasoning of the ECJ in
Coditel II or Erauw-Jacquery. Had it not followed the default, efficiency-based methodological approach
sketched above, the ECJ would have concluded in the two cases (and rightly so!) that the aim of the
contentious restraints was to restrict parallel trade. Suffice it to recall that one of the clauses examined in
the latter judgment was an export prohibition, which was understood to amount to an object restriction in
Miller International.
The ECJ was careful to point out in Glaxo Spain that the rule whereby agreements restricting parallel
trade are restrictive by their very nature applies only “in principle”. The crucial question for the purposes
of this work is therefore whether economic analysis can be usefully relied upon to identify in advance the
instances in which such agreements will not be found to restrict competition by object on the basis of the
intent of the parties. In this regard, it would seem that the sui generis, intent-based methodology followed
in some cases is no more than merely the formal manifestation of a rebuttable presumption whereby
parallel trade restrictions are a disproportionate means to achieve the potential efficiency gains entailed
by the agreement (and this on account of the role of market integration as an autonomous objective of EU
competition law).
Conversely, EU Courts seem to rely on the default methodology described above in those cases where
the efficiency explanations behind a given restraint are sufficiently compelling (i.e. where an analysis of
the nature and the context of the agreement suggests that the efficiency gains can only materialise in the
presence of explicit parallel trade restrictions). For instance, one may infer from the judgment in Société
67
Glaxo Spain (C-501/06 P , C-513/06 P , C-515/06 P and C-519/06 P) [2009] E.C.R. I-9291 at [61]–[64]. The GC
had instead considered in its judgment that market integration was protected insofar as it is likely to lead to consumer
welfare improvements. See GlaxoSmithKline Services Unlimited v Commission (T-168/01) [2006] E.C.R. II-2969;
[2006] 5 C.M.L.R. 29 at [114]–[136].
68
The ECJ held in Consten-Grundig (56/64) [1966] E.C.R. 429 that the agreement was aimed “at isolating the
French market for Grundig products and maintaining artificially , for products of a very well-known brand, separate
national markets within the community”.
69
Miller International Schallplatten GmbH v Commission (19/77) [1978] E.C.R. 131 at [7].
70
Glaxo Spain (C-501/06 P , C-513/06 P , C-515/06 P and C-519/06 P) [2009] E.C.R. I-9291 at [59]–[60].
552 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
T echnique Minière a principle whereby absolute territorial protection might be necessary in some cases
to overcome free-riding concerns (and would thus not amount to an object restriction).
71
According to the
Guidelines on vertical restraints, this may indeed be the case where “substantial investments by the
distributor to start up and/or develop [a] new market are necessary”, and absolute territorial protection is
provided for a limited period of time (i.e. two years or less).
72
An analysis of case law shows that restrictions to parallel trade are typically examined under the default,
efficiency-based methodology where they seek to preserve the value of a public good. This idea is
particularly apparent in Coditel II, where the ECJ referred explicitly to the nature and purpose of the
intellectual property right at stake (communication to the public). In Erauw-Jacquery, the activities covered
by the agreement were part of the core of the exclusive rights covered by intellectual property legislation.
It is thus only logical (and sound from an economic perspective) that art.101(1) TFEU was interpreted in
a way that the recourse to a licensing agreement was not made less advantageous for the right holder than
its in-house exploitation. This is the crucial difference with other cases such as Glaxo Spain and
Consten-Grundig, which merely concerned the distribution of the tangible goods incorporating the intangible
property of the supplier .
Efficiency gains and the vertical price fixing fetish
When the ECJ examined the status of vertical price fixing under art.101(1) TFEU in Binon, it did not deny
that such a restraint is a potential source of substantial efficiency gains, but accounted for them in a manner
that, again, departed from the default methodology outlined above. In a succinct paragraph, from which
a clear rationale is difficult to draw , the ECJ concluded that vertical price fixing agreements are restrictive
of competition by their very nature, seemingly on account of the fact that they limit the freedom of the
distributor to set prices in its dealings with third parties.
73
This same position would be confirmed in
Pronuptia,
74
and then in Erauw-Jacquery, where the ECJ relied on a literal reading of art.101(1) TFEU.
75
As can be seen, if resale price maintenance was understood to amount to an object restriction in these
cases, this is, first and foremost, due to the alternative methodological approach followed by the ECJ. T o
the extent that other vertical agreements such as selective distribution have a comparable impact on prices,
that this impact has been openly and repeatedly acknowledged by EU courts,
76
and that selective distribution
agreements limit the freedom of action of retailers in their dealings with third parties in the same way
vertical price fixing does, there seems to be no compelling reason why the ECJ could not have reached a
different conclusion in the above-mentioned cases.
It would seem that the current legal status of vertical price fixing cannot be seen as an unavoidable
manifestation of the peculiarities of EU competition law , or as a univocal rule deriving logically from the
letter of the Treaty . If anything, this status is at odds with the flexibility displayed in the vast majority of
cases dealing with vertical restraints. Arguably , the position expressed in Binon and Pronuptia is, more
than anything, expressive of a powerful and long-standing fetish that can be observed on both sides of the
Atlantic. Even though the potential efficiency explanations for vertical price fixing have been identified
71
In Société T echnique Minière (56/65) [1966] E.C.R. 235, the ECJ held that “it may be doubted whether there is
an interference with competition if the said agreement seems really necessary for the penetration of a new area by an
undertaking”. For a discussion of the potential meaning and practical application of this idea, see e.g. Faull and Ali
(eds), The EC Law of Competition (2007), pp.240–243.
72
Guidelines on vertical restraints, para.61.
73
SA Binon & Cie v SA Agence et messageries de la presse (243/83) [1985] E.C.R. 2015 at [44].
74
Pronuptia (161/84) [1986] E.C.R. 353 at [25], cited above.
75
Erauw-Jacquery (27/87) [1988] E.C.R. 1919 at [15].
76
In Metro I, the ECJ acknowledged and did not dispute that selective distribution agreements significantly reduce
the scope for price competition for the products offered within the network of operators. See in particular Metro I
(26/76) [1977] E.C.R. 1875 at [21] and [22]. See also Pierre Fabre (C-439/09) [2011] 5 C.M.L.R. 31 at [40].
Market Failures, Transaction Costs and Article 101(1) TFEU 553
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
for over 50 years, and that the consensus among economists is that this restraint cannot be equated in any
way to “naked” cartels,
77
only in 2007 did the US Supreme Court reverse a century-old precedent in
Leegin.
78
The assessment of restrictive effects on competition
The restrictive effects of agreements addressing market failur es
Where an analysis of the nature and/or the context of an agreement suggests that the restraints contained
therein are a plausible source of efficiency gains, it is necessary to establish, on a case-by-case basis, its
restrictive effects on competition within the meaning of art.101(1) TFEU. In this regard, it can no longer
be disputed that the mere fact that an agreement curtails the contractual freedom of the parties is as such
insufficient to conclude that it is caught by the prohibition. It is equally difficult to dispute that the ECJ
has regularly balanced the efficiency enhancing and restrictive effects of the agreement within art.101(1)
TFEU. This is so in spite of the bold assertion to the contrary made in Metropole T elevision, where the
GC held that such balancing must be confined to art.101(3) TFEU.
79
W outers and Gøttrup-Klim are cases
that immediately come to mind when these issues are raised, but are by no means the only ones. Suffice
it to mention, among the most explicit examples, those of Remia, already mentioned above, or of Metro
II.
80
The complexity of case law on this point is currently understood and formalised only imperfectly . As
mentioned in the introduction, the Commission considers in its policy documents that restrictive effects
are likely to be manifested where the agreement increases market power, that is, where the agreement
may lead to increased prices and reduced output. The most sophisticated attempt to develop a theoretical
framework based on this same idea is probably the work by Odudu, who associates the legal notion of
restriction of competition with the economic one of allocative inefficiency .
81
From this perspective, an
agreement will be deemed to restrict competition when it can be presumed to lead to higher prices or to
restrict output. This theoretical framework seems appropriate for those agreements that are likely to lead
to (1) productive efficiency gains (clear examples of which include joint production and joint purchasing),
but needs refinement, it is submitted, to reflect the nature of (2) those that address market failures. These
two instances are examined in turn.
77
For an expression of this consensus, see in particular the report submitted by the European Advisory Group on
Competition Policy , “Hardcore restrictions under the Block Exemption Regulation on vertical agreements: An economic
view” (September 2009), http://ec.europa.eu [Accessed September 6, 2012]. See also the views of Hovenkamp, who
is associated with the Harvard tradition, and who criticised the vertical price fixing fetish eloquently (and this before
Leegin): “The courts need to jettison the per se rule for RPM and adopt a form of analysis that is more concerned
about powerful dealers or dealer collusion, and less concerned about whether the challenged restraint requires a dealer
to set a specific price”. See Hovenkamp, The Antitrust Enterprise (2005), p.188.
78
Leegin Creative Leather Products, Inc v PSKS, Inc 551 U.S. 877 (2007).
79
Métropole télévision SA v Commission (T-206/99) E.C.R. II-1057; [2001] 4 C.M.L.R. 39.
80
Metro SB-Großmärkte GmbH & Co KG v Commission (Metr o II) (75/84) [1986] E.C.R. 3021 at [45]: “Furthermore,
the Court recognized in that connection in its judgment in Metro I that some limitation in price competition was to
be regarded as inherent in any selective distribution system, because the prices applied by specialist dealers necessarily
remained within a much narrower margin than would be expected if there were competition between specialist dealers
and non-specialist dealers. It stated that that limitation was counterbalanced by competition as regards the quality of
the services supplied to customers , which was not normally possible in the absence of an adequate profit margin
covering the higher costs entailed by such services” (emphasis added).
81
O. Odudu, The Boundaries of EC Competition Law: The Scope of Article 81 (Oxford: Oxford University Press,
2006), in particular pp.102–127. The general claim made by the author is that “the substantive meaning of restriction
of competition within Article [101(1) TFEU] is allocative inefficiency”.
554 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
Where productive efficiency gains are at stake, the assessment of restrictive effects on competition
seems to revolve around the question of whether the efficiency gains generated by the agreement are likely
to be passed on to consumers. In other words, it is necessary to balance, under art.101 TFEU, the alleged
total welfare gains against the allocative inefficiency (consumer welfare losses) that may result from the
increase in market power . In this context, the purpose of art.101(1) TFEU is, it would seem, to mark the
point at which it is not possible to presume that productive efficiency gains will benefit consumers. Thus,
where market power reaches a certain degree, any efficiency claims gains will have to be carefully
established and quantified under art.101(3) TFEU. Conversely, where the increase in market power is
relatively modest, the agreement will not be deemed to restrict competition within the meaning of art.101(1)
TFEU.
This analytical framework follows the principles sketched by the ECJ in Gøttrup-Klim, where the Court
took the view that, depending on the economic conditions prevailing on the relevant market, a joint
purchasing agreement may “make way for more effective competition”.
82
The same would be true where,
as in O2, the agreement facilitates market entry , in which case it cannot be truly said to lead to an increase
in market power .
83
The most recent policy instruments of the Commission, of which the Guidelines on
horizontal co-operation agreements is an example, formalises these insights by setting market share
thresholds below which one can presume that the efficiency gains are likely to outweigh a moderate
increase in market power .
A market power-based framework is, however, not appropriate (or , more precisely , it is incomplete)
where the restraints found in an agreement constitute a reaction to a market failure. As explained in s.2,
such restraints are in themselves a source of allocative efficiency gains. As a consequence, a sole focus
on the observed ex post impact of the agreement on prices and output can only lead to misleading outcomes.
For instance, it is incorrect to claim that an exclusive distribution or an exclusive territorial licensing
agreement leads to higher prices insofar as they restrict intra-brand competition. In such cases, it is necessary
to consider that the reseller may not have distributed the products in the absence of mechanisms allowing
it to capture the positive externalities generated by its activity; or that the holder of an intellectual property
right would not have accepted to license the rights without it being able to preserve the value of the public
good in question.
84
The genius of the ECJ in the cases discussed above (which is not captured in a market power-based
framework) was to understand early in time that agreements addressing market failures require a balancing
of the ex ante allocative efficiency gains against potential ex post losses. In cases where an agreement has
been found to constitute a proportionate response to a market failure, the ECJ has shown its readiness to
leave the agreement outside the scope of art.101(1) TFEU altogether without it being necessary to assess
its impact on the market in which it is implemented.
The proportionality assessment undertaken by the ECJ under art.101(1) TFEU can be deconstructed
into two steps. First, the ECJ seems to isolate the restraints that are related to the market failures identified.
82
Gøttrup-Klim (C-250/93) [1994] E.C.R. I-5641 at [32].
83
O2 (Germany) GmbH & Co OHG v Commission (T-328/03) [2006] E.C.R. II-1231; [2006] 5 C.M.L.R. 5 at
[68]–[73].
84
Korah has long insisted on this fundamental point. See V . Korah, “Draft block exemption for technology transfer”
(2004) 25 European Competition Law Review 247, 261: “Once the investment has been made in successful r & d, its
cost is water under the bridge and the market would be more competitive if free riders could exploit the results. More
would be produced and the competition might bring prices down. Similarly, once a licence has been granted, the
market would be more competitive if no restraints were imposed on the parties’ conduct. Seen ex ante, however, the
inability to impose restrictions is likely to reduce the incentive to grant licences for production within the Community
and even to result in r & d moving elsewhere”. Interestingly , the US Department of Justice considered in its Guidelines
on the Licensing of Intellectual Property (http://www .justice.gov/atr/ [Accessed September 6, 2012]) that “an exclusive
license may raise antitrust concerns only if the licensees themselves, or the licensor and its licensees, are in a horizontal
relationship”.
Market Failures, Transaction Costs and Article 101(1) TFEU 555
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
In Pronuptia and L’Or eal, for instance, the ECJ distinguished between the restraints relating to the
protection of the intangible property of the supplier and those protecting resellers against free-riding.
85
The second step of the analysis examines the proportionality stricto sensu of the restraints that relate to
the market failure in question. The length of non-compete obligations may be found to be excessive, as
in Remia
86
; vertical price fixing may be deemed to go beyond what is necessary to preserve the uniformity
of a franchise system, as in Pronuptia
87
; or the prohibition of the parallel trade of decoding equipment
may be found to be a disproportionate means to protect the value of the rights subject to the exclusive
territorial licensing agreement, as in Premier League.
88
By excluding the need to assess the restrictive effects on competition on a case-by-case basis in judgments
like Metro I, Remia and Pronuptia, the ECJ essentially laid down a presumption whereby the allocative
efficiency gains weigh more than the negative impact of the agreement on prices and output. What is
notable about these cases is that the applicability of art.101(1) TFEU does not depend on the context of
the agreement, but on the existence of an adequate link between the restraints and the protection of the
intangible property in question. The presumption can be reversed in the presence of additional factors. As
acknowledged in Metro I and Metro II it may be that the multiplication of parallel networks exacerbates
the price rigidities that selective distribution systems entail by definition.
89
Elements of a similar presumption can be found, regarding exclusive territorial licensing agreements,
in Erauw-Jacquery. The essential criterion to establish a restriction under art.101(1) TFEU was, again,
not the context in which it was implemented but the financial effort deployed by the licensor to develop
a new technology . It would therefore seem that, where there is evidence that this effort has been undertaken,
the agreement will be presumed to have a net positive impact on competition. The judgment is particularly
valuable in that it hints at the award of an intellectual property right as a predictable and practicable proxy
that these efforts have been undertaken.
90
Consistency would dictate that Coditel II, which also concerns public goods but is less explicit in this
regard, is interpreted as laying down the same presumption. This is all the more so if one reads the case
in conjunction with A.G. Reischl’s Opinion, who clearly understood that absolute territorial protection
was necessary to ensure that the intellectual property rights exploited can be appropriated by the licensee
85
Pronuptia (161/84) [1986] E.C.R. 353, in particular at [23]. L’Oréal and SA L’Oréal v PVBA “De Nieuwe AMCK”
NV (31/80) [1980] E.C.R. 3775 at [17].
86
Remia (42/84) [1985] E.C.R. 2545 at [25]–[27].
87
Pronuptia (161/84) [1986] E.C.R. 353 at [25].
88
In the latter case, the ECJ was careful to clarify that the principles of Coditel II were not being questioned. See
Premier League [2012] Bus. L.R. 1321 at [141]: (“In the main proceedings, the actual grant of exclusive licences for
the broadcasting of Premier League matches is not called into question. Those proceedings concern only the additional
obligations designed to ensure compliance with the territorial limitations upon exploitation of those licences that are
contained in the clauses of the contracts concluded between the right holders and the broadcasters concerned, namely
the obligation on the broadcasters not to supply decoding devices enabling access to the protected subject-matter with
a view to their use outside the territory covered by the licence agreement”).
89
Metro I (26/76) [1977] E.C.R. 1875, para.22; and Metro II (75/84) [1986] E.C.R. 3021 at [35]–[42]. In Metro II,
the complainant claimed that the Commission had not adequately considered the cumulative impact of parallel networks
of selective distribution systems when granting an exemption under art.101(3) TFEU. The reference to networks of
parallel agreements reminds of the conditions laid down in Delimitis for the assessment of exclusive dealing obligations.
To the extent that it does, it shows that the outcome of the latter case would not have changed dramatically had the
ECJ given more prominence in the analysis to the market failure related and the transaction cost related gains entailed
by such arrangements.
90
Erauw-Jacquery (27/87) [1988] E.C.R. 1919 at [10], cited above. The ECJ refers to a scenario in which a person
“has made considerable efforts to develop varieties of basic seed which may be the subject-matter of plant breeders’
rights” (emphasis added).
556 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
and are not destroyed by the import of foreign signals.
91
It is in fact difficult to see how a single exclusive
television licence (e.g. a licence for a sporting event, or a Hollywood output deal) can restrict upstream
or downstream competition “to an appreciable degree”.
92
Accordingly, it is logical that factors external to
the agreement are considered before a restriction of competition is established. As is true concerning
selective distribution agreements, downstream competition may be significantly affected in this context,
for instance, as a consequence of the cumulative effect of several exclusive licences concluded by one
and the same television operator .
93
These insights help make sense of W outers and show that there is arguably nothing exceptional in the
reasoning displayed by the ECJ. The ruling can be read as laying down the principle whereby proportionate
reactions to information asymmetry concerns are not restrictive of competition, and this, again, irrespective
of the market power enjoyed by the parties. The position of the ECJ in Asnef-Equifax and the treatment
of information exchange agreements in the Guidelines on horizontal co-operation agreements are based
on this same logic. Depending on the nature of the exchange, an agreement of this kind covering the whole
of an industry may fall outside the scope of art.101(1) TFEU altogether .
94
Extending the analysis to transaction costs
T o the extent that agreements leading to transaction costs reductions lead to allocative efficiency gains,
there is every reason to assess their effects on competition in a way that is consistent with that followed
in cases in which market failures were relevant in the analysis. As explained above, the ECJ held in
T ournier that the activities of collecting societies fall outside the scope of art.101(1) TFEU if they pursue
a “legitimate aim”.
95
The Commission has taken the same view in several policy instruments. Standard
setting, the status of which is examined in detail in the Guidelines on horizontal co-operation agreements,
provides one such example. If the relatively strict conditions set out in paras 277 to 286 are fulfilled, such
91
Opinion of A.G. Reischl delivered on September 14, 1982 in Coditel II (262/81) [1982] E.C.R. 3381, in particular
at 3412. In particular, the Advocate General notes that distributors play a fundamental role in the financing of new
film productions, which is typically offered in exchange for some degree of territorial protection. A strict stance on
these arrangements could have a negative impact on the production of films. In addition, the Advocate General found
it inappropriate to refer in this case to a restriction in parallel imports insofar as the protected works were not embodied
in a tangible good (in the words of the Advocate General, “it would impoverish the market and depress competition”).
92
Suffice it to take the example of the first decision taken by the Commission in the field (Commission Decision
89/536 of relating to a proceeding under art.85 of the EEC Treaty (IV/31.734 — Film purchases by German television
stations) [1989] OJ L284/36). While the agreement covered an important number of films (around 1,500), it could
only have a limited impact on competition as such (this library represented 4.5% of the total number of films available
to television operators). Against this background, it is not surprising that the Decision, which imposed amendments
to the agreements to secure third-party access to some rights, was heavily criticised at the time. See e.g. W. Rothnie,
“Commission Re-runs same old Bill (Film Purchases by German Television Stations)” [1990] E.I.P .R. 72.
93
This is in fact the scenario that has triggered intervention by competition authorities in the pay-TV markets. It
can be observed that all premium television rights, typically offered in subscription-based channels (these includes
top sports events and recent film releases), end up in the hands of a single broadcaster, which is the true competition
concern. See in this sense the Decision of the AGCM of June 14, 2000 in Case A274, Stream/T elepiù, in which the
competition authority found that the cumulative effect of exclusivity agreements concluded by the incumbent pay-TV
operator amounted to an abuse of dominance. The concern with the cumulative effect is also clearly manifested in
Commission Decision of March 22, 2006 relating to a proceeding pursuant to art.81 of the EC Treaty (Case
COMP/38.173 — Joint selling of the media rights to the F A Premier League) [2008] OJ C7/18.
94
See Asnef-Equifax (C-238/05) [2006] E.C.R. I-11125 at [55]–[56] (referred to above) and the Guidelines on
horizontal co-operation agreements, paras 75–94. See also M. Bennett and P . Collins, “The Law and Economics of
Information Sharing: The Good, the Bad and the Ugly” (2010) 6 European Competition Journal 311. These authors
suggest, in line with the principles that derive from the above-mentioned Guidelines, that the exchange of aggregated
historic data are unlikely to restrict competition within the meaning of art.101(1) TFEU.
95
T ournier (395/87) [1989] E.C.R. 2521 at [31].
Market Failures, Transaction Costs and Article 101(1) TFEU 557
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
agreements are deemed not to run counter to art.101(1) TFEU.
96
The same is true of agreements providing
for the joint licensing of a technology through a pool of “essential” patents.
97
These examples could be interpreted as an expression of a general principle whereby agreements leading
to substantial transaction cost reductions, and that do not go beyond that deemed necessary to achieve
these reductions, can be presumed to fall outside the scope of art.101(1) TFEU. The relevant question is
then whether, and how , this presumption can be made operational. If one considers the example of the
joint licensing of technologies or that of collecting societies, one may claim that the relevant factor is
whether the agreement allows for the emergence of a new product that the parties to the agreement alone
would not have been able to offer .
One can find factual illustrations of these ideas in the administrative practice of the Commission. Suffice
it to think of the joint selling of television rights, one of the few areas in which transaction costs have been
explicitly considered by the authority .
98
In UEF A Champions League, the Commission concluded, in
contradiction with well-established case law , that the agreement was restrictive of competition insofar as
it restricted the freedom of action of the parties.
99
In the light of the above, however, there is every reason
to believe that the authority could have concluded that the agreement allowed for the emergence of a new
product and thus did not infringe art.101(1) TFEU. In fact, it was overtly acknowledged in the decision
that the joint selling of the rights to a sports competition is an arrangement that leads to significant gains
in terms of transaction cost reductions for television operators (the alternative being an individual negotiation
by the operator with each of the teams taking part in the competition).
100
In addition, the Commission was
96
It is explained in paras 277–286 of the Guidelines that standard setting agreements are unlikely to raise concerns
from a competition law perspective where “participation in standard-setting is unrestricted and the procedure for
adopting the standard in question is transpar ent, standardisation agreements which contain no obligation to comply
with the standard and provide access to the standard on fair , reasonable and non-discriminatory terms will normally
not restrict competition within the meaning of art.101(1)” (emphasis in the original).
97
Guidelines on technology transfer agreements, para.220: “When a pool is composed only of technologies that
are essential and therefore by necessity also complements, the creation of the pool as such generally falls outside
Article [101(1)] irrespective of the market position of the parties. However, the conditions on which licences are
granted may be caught by Article [101(1)]”. In para.226, the Commission requires, in particular that the pool licenses
its technology on a fair , reasonable and non-discriminatory basis.
98
Transaction costs have been explicitly considered only sparingly in the practice of the Commission. The author
has been able to find just three Commission decisions in which transaction costs were a clearly relevant factor in the
substantive analysis under art.101(1) TFEU. In IFPI, the Commission accepted that the notified agreements led to
substantive gains in terms of transaction costs. Commission Decision 2003/300 of October 8, 2002 relating to a
proceeding under art.81 of the EC Treaty and art.53 of the EEA Agreement (Case No COMP/C2/38.014 — IFPI
Simulcasting) [2003] OJ L107/58 at [90]. This was also the case in Commission Decision 2003/778 of July 23, 2003
relating to a proceeding pursuant to art.81 of the EC Treaty and art.53 of the EEA Agreement (COMP/C.2-37.398 —
Joint selling of the commercial rights of the UEFA Champions League) [2003] OJ L291/25. See in particular
[148]–[149] of the Decision, where the Commission accepts that, in the absence of the joint selling agreement,
television operators interested in acquiring the rights for the event in question would have to incur substantial transaction
costs. A reference to transaction costs is also made in Commission Decision 92/204 of February 5, 1992 relating to
a proceeding pursuant to art.85 of the EEC Treaty (IV/31.572 and 32.571 — Building and construction industry in
the Netherlands) [1992] OJ L92/1. Although this Decision concerns a straightforward cartel case (more precisely , it
concerns a formal mechanism for collusive bid rigging among building firms), the parties to the agreement had sought
to justify the collusive mechanism by resorting to what the Commission called the “theory of transaction costs” (see
[74]). In essence, the parties improbably claimed in essence that, to the extent that participating in a tender is costly ,
incurring these costs is only justified if the building company can pass these costs to the purchaser.
99
The Commission did not explain whether the agreement was restrictive by object or by effect. As in many other
cases, the authority disregarded case law principles and concluded that it was caught by art.101(1) TFEU because it
restricted the freedom of action between the football teams exploiting the rights through the UEFA. See UEFA
Champions League [2003] OJ L291/25 at [113]–[116].
100
UEFA Champions League [2003] OJ L291/25 at [148].
558 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
willing to concede that, given the particular nature of the UEFA Champions League as a competition, the
agreement was probably the only workable way in which the television rights could have been licensed.
101
Implications of the analysis
The respective scopes of Articles 101(1) and 101(3) TFEU
T wo fundamental ideas stem from the analysis above. The first one is that the assessment of restrictive
effects of competition under art.101(1) TFEU requires some form of balancing. This is particularly apparent
when the agreement is a source of allocative efficiencies (insofar as the ECJ does not even require an
assessment of the context in which the agreement is concluded), but it is also true where the agreement
leads to productive efficiency gains. The second idea is that such efficiency gains, while acknowledged
in the assessment, are not expressly quantified, but merely presumed. This is in line with the position
taken by the GC in MasterCard, where it held that the analysis of the “objective necessity” of a restraint
under art.101(1) TFEU “cannot be but relatively abstract”.
102
These ideas constitute a valuable starting point to define the role and purpose of art.101(3) TFEU.
Contrary to what has often been argued,
103
the fact that the balancing of the pro- and anti-competitive
effects of an agreement takes place in some form within art.101(1) TFEU does not mean that the former
is devoid of purpose. Article 101(3) TFEU is the appropriate forum to give the parties the chance to
quantify efficiency gains explicitly and show that the agreement is on the whole pro-competitive. The
bold statement found in Metropole T elevision can be understood in this light. If the Court meant in that
case, as it held in MasterCard, that the pro- and anti-competitive effects of agreements are not explicitly
weighed against one another under art.101(1) TFEU but are the result of an abstract assessment based on
presumptions, there is no reason to disagree with the judgment.
It may be necessary to quantify the efficiency gains of an agreement under art.101(3) TFEU where it
creates or strengthens market power beyond a certain degree, or where the context in which it is concluded
suggests that the negative impact on prices and output may weigh more than the allocative efficiency gains
achieved. More generally , it may be necessary to resort to art.101(3) TFEU where the restraints are
understood to go beyond what is deemed necessary to achieve the efficiency gains identified in the first
place (as is the case where a franchising agreement provides for vertical price fixing).
One can think of several concrete examples, drawn from the cases discussed above, in which art.101(3)
TFEU could be of relevance. Beef Industry Development is a clear one. As already explained, the agreement
as such was not a credible source of efficiency gains, as the capacity reductions would have, in all likelihood,
materialised as a result of the normal play of market forces. It was therefore for the parties to show that
the agreement was necessary to achieve these gains. Similarly , it has been shown that agreements restricting
parallel trade and relating to the distribution of tangible goods are presumed to restrict competition by
object. That the benefit of art.101(3) TFEU is open to these agreements can no longer be disputed after
Glaxo Spain.
104
By the same token, the ECJ made it clear in Binon (as the Commission concedes in the
101
UEFA Champions League [2003] OJ L291/25 at [145].
102
MasterCard, Inc and MasterCard Europe v Commission (T-111/08) [2012] 5 C.M.L.R. 5 at [80].
103
See in particular Monti, EC Competition Law (2007), pp.26–27. According to the author, the bifurcated nature
of art.101 TFEU “is incompatible with a neoclassical definition of competition”. For an comprehensive discussion
of art.101(3) TFEU, see B. Sufrin, “The Evolution of Article 81(3) of the EC Treaty” (2006) 51 Antitrust Bulletin
915.
104
This is all the more so if one considers, as pointed out above, that the Commission itself is willing to accept that
absolute territorial protection may in some instances fall outside the scope of art.101(1) TFEU altogether . The position
of the ECJ in Glaxo Spain case illustrates very well the respective roles of art.101(1) and (3) TFEU.
Market Failures, Transaction Costs and Article 101(1) TFEU 559
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
current version of the Guidelines on vertical restraints), that resale price maintenance may satisfy the
conditions set out in art.101(3) TFEU.
105
This interpretation of art.101(3) TFEU is all the more desirable if one considers that it would lead to
efficiency gains being considered in a consistent manner across competition law provisions. A reading of
the Horizontal Merger Guidelines, for instance, shows that the pro- (typically , productive efficiency gains)
and anti-competitive (allocative inefficiency) effects of a concentration between competing firms are
rarely ever balanced explicitly under art.2 of Regulation 139/2004. As under art.101(1) TFEU, the
Commission proceeds by means of presumptions. Accordingly, it is prima facie excluded that a horizontal
merger will be anti-competitive where the market share of the new entity is below 25 per cent,
106
and the
opposite is true where the market share exceeds 50 per cent.
107
In fact, it is only where a careful assessment
of the relevant market confirms that the negative effects of the operation are likely to prevail that the
merging parties are invited to identify and quantify the expected efficiency gains.
108
The role of objective
justifications under art.102 TFEU does not seem to be fundamentally different, insofar as they come into
play where the prima facie abusive nature of the behaviour is established.
109
The role of public policy considerations
Several commentators have explored the role of public policy and, more generally , of non-economic
considerations in art.101 TFEU.
110
Accounting for public policy objectives in the analysis of agreements
may conflict with the core objectives of competition law , it is claimed, insofar as it may lead to outcomes
differing from those that would have resulted from a decision based on the latter alone. The fundamental
question in this debate seems to be whether such diverging outcomes are acceptable, or whether, as the
Commission holds in the Guidelines on the application of art.101(3) TFEU, “goals pursued by other
Treaty provisions can be taken into account to the extent that they can be subsumed under the four
conditions” laid down in that provision.
111
T ownley has used an appealing and very concrete example to illustrate how conflicts between what
may be termed economic and non-economic considerations may arise in practice. The author explores the
status under competition law of an agreement between supermarket chains setting minimum prices for
alcoholic drinks to prevent teenage consumption. In spite of the laudable objectives pursued by such an
agreement, it could not be validated under art.101 TFEU, this author suggests, unless public health, which
is in principle foreign to competition law analysis, is considered.
112
It is submitted that it is not necessary to resort to a sui generis analytical approach based on the balancing
between economic and non-economic objectives in situations such as that described by T ownley . An
analytical framework revolving around efficiency is sufficiently robust to capture the supposed peculiarities
105
Binon (243/83) [1985] E.C.R. 2015 at [46].
106
Horizontal Merger Guidelines, paras 17–18.
107
Horizontal Merger Guidelines, paras 17–18.
108
Horizontal Merger Guidelines, paras 76–84.
109
Guidance Paper, paras 28–31. See, in general, A.A. Llorens, “The Role of Objective Justification and Efficiencies
in the Application of Article 82 EC” (2007) 44 C.M.L. Rev. 1727.
110
In addition to Monti, “Article 81 EC and Public Policy” (2002) 39 C.M.L. Rev. 1057, see, in particular, H.
Schweitzer, “Competition Law and Public Policy: Reconsidering an Uneasy Relationship: The Example of Art. 81”,
EUI Law Working Papers No.2007-30, SSRN, http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1092883 [Accessed
September 6, 2012]; C. Townley , Article 81 EC and Public Policy (Oxford: Hart Publishing, 2009); as well as Odudu,
The Boundaries of EC Competition Law (2006), pp.159–174.
111
Guidelines on the application of art.101(3) TFEU, para.42. The Commission refers in support of its position to
Matra Hachette SA v Commission (T-17/93) [1994] E.C.R. II-595.
112
C. T ownley , “Is Anything more Important than Consumer Welfare (in Article 81 EC)?: reflections of a Community
lawyer” (2007–2008) 10 Cambridge Y earbook of European Legal Studies 345.
560 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
of these situations. V ery often, intervention allegedly based on non-economic grounds is in fact a response
to a market failure and thus not necessarily in conflict with (allocative) efficiency . For instance,
environmental protection measures are by and large a reaction to the negative externalities generated by
some economic activities. The same can be said of the issues at stake in W outers, which can be seen as a
reaction to information asymmetries plaguing the relevant market.
The Commission has in fact already examined alleged non-economic concerns under a prism that is
similar to the default methodology sketched above. T ake CECED, an omnipresent example in these
discussions. The Commission carefully quantified the positive impact on consumer and total welfare of
the agreement in a fully orthodox way.
113
From the perspective taken in this work, the only true question
that arises in relation to this case is whether the agreement should have been left outside the scope of
art.101(1) TFEU altogether . T o the extent that the allocative efficiency gains deriving from the agreement
clearly exceeded the negative impact of the observed restriction of output, the Commission could have
concluded, in line with Metro I or Pronuptia, that it did not restrict competition.
With this background in mind, it seems relatively straightforward to re-examine the hypothetical
advanced by T ownley . It is beyond dispute that excessive alcohol consumption is a source of negative
externalities.
114
Accordingly, one could argue that, at the very least, an agreement setting recommended
prices for alcoholic drinks could be a proportionate response to the market failure identified and thus
would not be caught by art.101(1) TFEU. The same conclusions could extend to similar agreements.
Suffice it to think of a code of practice whereby the adhering restaurants, following the advice of
cardiologists, agree not to use trans fats, limit the amount of salt in their dishes or reduce the size of the
portions served.
Conclusions
The ECJ has consistently displayed a remarkable grasp of the motivations behind firm co-operation. This
is manifested in the flexibility with which art.101(1) TFEU has long been interpreted and applied. The
fundamental claim made in this piece is that positive economic analysis helps provide a neat and coherent
picture of some rulings that are still seen as puzzling in many respects. It is submitted that it is possible
to make sense of, and formalise, the principles deriving from case law by reference to the notions of market
failure and transaction costs. Moreover, these notions provide a benchmark on the basis of which actual
and alleged contradictions between judgments can be examined systematically .
Policy action no longer ignores that private transactions can effectively address the consequences of
market failures and reduce transaction costs. However, the potential of these (nowadays obvious) insights
to illuminate case law is still largely ignored. Much could be gained if the Commission, in future soft law
instruments, continued the path opened in the 1990s with the Guidelines on vertical restraints and
considered how these factors influence the outcome of cases under art.101(1) TFEU (and not simply under
art.101(3) TFEU). Similarly , future generations of students would very much benefit if the scope of
art.101(1) TFEU were defined on the basis of standard, well-established tools used in mainstream
economics, and not with the aid of legal categories that lack sufficient explanatory power . The fundamental
lessons that can be drawn from the analysis above can be summarised as follows.
First, the crucial question marking the dividing line between restrictions by object and by effect seems
to be that of whether the restraints found in an agreement can be plausibly explained on efficiency grounds.
113
Commission Decision 2000/475 of January 24, 1999 relating to a proceeding under art.81 of the EC Treaty and
art.53 of the EEA Agreement (Case IV .F .1/36.718 — CECED) [2000] OJ L187/47 at [55]–[57], in particular [56],
where the Commission estimates that the “benefits to society brought about by the CECED agreement appear to be
more than seven times greater than the increased purchase costs of more energy-efficient washing machines”.
114
See e.g. J. Miron and J. Zwiebel, “The Economic Case Against Drug Prohibition” (1995) 9 Journal of Economic
Perspectives 175.
Market Failures, Transaction Costs and Article 101(1) TFEU 561
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors
Contrary to what is sometimes assumed, the ECJ seems to focus on the fact that the agreement lacks
redeeming virtues, and not on its potential to restrict competition. The formal features of the agreement
(the fact that, for instance, it provides for price fixing or market sharing) are not decisive. Secondly , the
ECJ tends to adopt a stricter stance, which ignores efficiency claims, when the agreements restrict parallel
trade. It is only where such claims are sufficiently compelling (as is the case, in particular, where an
agreement seeks to preserve the value of a public good) that the Court will revert to its efficiency based,
default methodology. EU courts have also departed from this methodology where an agreement provides
for vertical price fixing.
Agreements addressing market failures lead to allocative efficiency gains. The ECJ seems to acknowledge
this fundamental insight, as evidenced by its approach to the assessment of their restrictive effects under
art.101(1) TFEU. Where the response to a market failure is deemed proportionate, the analysis of the ECJ
is, it would seem, based on the presumption that the ex ante allocative efficiency gains are sufficient to
outweigh any ex post losses. This same analytical framework can be extended to those agreements that
lead to substantial transaction cost reductions.
Efficiency gains are merely presumed under art.101(1) TFEU, as confirmed by the GC in MasterCard.
It is only under art.101(3) TFEU that these gains are explicitly quantified and balanced against the restrictive
effects established in accordance with the first paragraph of the provision. Such explicit balancing may
be necessary where the agreement creates or strengthens the market power of the parties beyond a certain
degree, or where the response to a market failure is deemed disproportionate.
Non-economic considerations (concerns with issues such as public health, environmental protection or
deontology) can often be seen as reactions to perceived market failures. T o the extent that this is the case,
there is no reason why agreements providing for such considerations should be evaluated in the light of
a sui generis analytical framework. From an economic perspective, the logic of W outers or CECED is not
fundamentally different from that found in, for instance, Metro I or John Deere.
562 European Law Review
(2012) 37 E.L. Rev. October © 2012 Thomson Reuters (Professional) UK Limited and Contributors