The role of innovation in the analysis of
abuse of dominance in digital markets
The analysis of chosen practices of Google
Search
Paper for the ASCOLA Workshop, Leiden 30.6.-2.7.2016
Draft, please do not cite
Beata Mäihäniemi
University of Helsinki
Contents
Abstract ............................................................................................................................................................ 2
1.
Introduction .............................................................................................................................................. 3
2.
EU Competition Policy and Innovation .................................................................................................... 8
2.1.
What is innovation? .......................................................................................................................... 8
2.2.
Innovation considerations in digital markets.................................................................................. 10
2.3.
Should the EU competition policy protect disruptive over sustaining innovation? ....................... 11
3.
Strength of consideration on incentives to invest and innovate in digital markets ................................ 13
4.
Phases of competition analysis in which incentives to invest and innovate can be taken into account 18
4.1.
4.2.
5.
The role of objective justifications by digital monopolies ............................................................... 18
4.1.1.
Objective justifications for restrictions on portability of online data ......................................... 20
4.1.2.
Objective justifications that could be used by Google as regards search bias ........................ 23
Effect of practice on consumers ........................................................................................................... 26
4.2.1.
Harm to innovation in restrictions on portability of online advertising data ......................... 28
4.2.2.
Harm to innovation in search bias........................................................................................... 33
Conclusions ............................................................................................................................................. 37
Abstract
Competition investigations in digital markets are focused increasingly on future markets and for that
reason incentives to invest and innovate need to be taken into account to a greater extent than in
traditional ‘brick and mortar’ industries. In particular, competition authorities should focus on the
notion of potential competition in such markets. Therefore, products that at first glance do not seem
to pose a competitive threat on a monopolist, may be considered as future substitutes taking into
account the fast changing phase of product characteristics.
This paper aims at analysing the role of innovation in cases of abuse of dominance in digital
markets on two levels.
The first level of the analysis focuses on the strength of incentives to invest and innovate of
a digital monopolist. Would such a monopolist have less incentives to innovate and prefer to resort
to practices that foreclose his competitors or leverage his market power to adjacent markets in order
to keep its dominant position on the market? Or is he forced to innovate since he is constantly
challenged by its competitors who may, any time in the future challenge and destroy the temporary
digital monopoly? A number of theories on innovation could help in answering this question,
however, this paper focuses on two particular approaches. Firstly, it takes into account Schumpeter’
claims that larger firms will tend to innovate more than small ones as well as Arrow’s findings that
product market competition actually encourages more innovation.
The second level of the analysis focuses on the concrete phases of the competition analysis
in which innovation considerations are taken into account in digital markets. It seems that
innovation considerations are of the highest value in objective justifications (especially business
justifications), e.g. a monopolist can refuse to share an input such as IPR or trade secret to prevent
free riding on its innovation. However, innovation also plays an important role in the assessment of
the effect of the practice on consumer welfare, e.g. where the Commission notes that the practice is
question will lessen incentives to innovate of the rivals in the result of which consumers will be
deprived of new and improved products. Consequently, innovation considerations can be used both
as a defence of dominant undertaking and as the reason for antitrust intervention.
The analysis of the role of innovation in the assessment of alleged anticompetitive abuses is
conducted on the example of two concerns expressed by the European Commission in recent
investigations into practices of Google Search. These are (1) search bias, which denotes the situation
where Google is preferring its own vertical search engines over the ones of rivals in its general search
results as well as (2) restrictions on portability of advertising data to competing advertising
platforms. These two practices are ambiguous as they are based on the ‘mixed motives’ of the
dominant undertaking and one can argue that they are either product improvements or
anticompetitive practices. What is more, both of these practices concern some kind of refusal to
access to or misappropriation of information, which, due to specific characteristics of information as
an economic good further complicate the analysis.
The paper is therefore not only aiming at the identification of the relevance of innovation
considerations for abuses of dominance that arise in digital markets in general. The paper also strives
to predict the outcome of one of the most-followed EC competition investigations, the result of
which is still unknown.
2
1. Introduction
Innovation on one hand leads to an exclusion of some firms as it increases market power of an
innovating firm. On the other hand, it benefits consumers and is not more harmful than lowering
prices to get rivals’ customers on one’s side. Innovation is an essence of competition and companies
who cannot keep up end up “harmed”.1 Innovation is a common concern in digital markets, it is also
an inevitable part of every digital business’ life. Companies in digital markets need to innovate to
survive, can therefore a company be condemned on the basis of competition law for simply trying
to avoid being challenged by some disruptive innovation? Incentives to invest and innovate became
a more urgent concern in digital markets than prices. However, they do not seem to be taken
sufficiently into account by the EU competition law. In particular, in cases involving abuse of
dominance, a dominant company could possibly defend itself claiming that instead of being
involved in an exclusionary practice, it is in fact trying to compete on merits by innovating as to
improve its products or services. Nevertheless, a dominant company could be also put under the
antitrust scrutiny if it is found that the practice in question affects innovation in a negative way.
The last situation is still ambiguous as the current state of competition law still requires that the
practice which has a negative effect on consumers by depriving them of new or improved products
or limits their choice, has to lead to anticompetitive foreclosure. Therefore, the introduction of
direct considerations on innovation is still impossible under the current state of the EU
competition law.2
In digital markets competition investigations are increasingly focused on future markets
and for that reason incentives to invest and innovate need to be taken into account to a greater extent
than in traditional ‘brick and mortar’ industries. In particular, competition authorities should focus
on the notion of potential competition in such markets. Therefore, products that at first glance do
not seem to pose a competitive threat on a monopolist, may be considered as future substitutes
taking into account the fast changing phase of product characteristics.
This paper aims at analysing the role of innovation in cases of abuse of dominance in digital
markets on two levels.
The first level of the analysis focuses on the strength of incentives to invest and innovate of
a digital monopolist. Would such a monopolist have less incentives to innovate and prefer to resort
Howard Shelanski ’Information, Innovation, and Competition Policy for the Internet’ (2013) 161 University of Pennsylvania Law
Review 1663, 1692.
2 See Pablo Ibáñez Colomo, ’Restrictions on Innovation EU Competition Law’ (2016) 41 European Law Review (forthcoming), LSE
Law, Society and Economy Working Papers 22/2015, https://www.lse.ac.uk/collections/law/wps/WPS2015-22_Colomo.pdf accessed
25 April 2016.
1
3
to practices that foreclose his competitors or leverage his market power to adjacent markets in order
to keep its dominant position on the market? Or is he forced to innovate since he is constantly
challenged by its competitors who may, any time in the future, challenge and destroy the temporary
digital monopoly? A number of theories on innovation could help in answering this question,
however, this paper focuses on two particular approaches. Firstly, it takes into account Schumpeter’
claims that larger firms will tend to innovate more than small ones as well as Arrow’s findings that
product market competition actually encourages more innovation. Does this discussion make sense
in digital markets? Is digital monopoly indeed that strong and does it have more resources to support
its research and development than smaller firms? The answer to this question may lay in the
distinction between different kinds of innovation and their different effects on welfare. Therefore,
there is a need to distinguish between disruptive and sustainable innovations, of which we will learn
shortly.
The second level of the analysis focuses on the concrete phases of the competition analysis
in which innovation considerations are taken into account in digital markets. It seems that
innovation considerations are of the highest value in objective justifications (especially business
justifications), e.g. a monopolist can refuse to share an input such as IPR or trade secret to prevent
free riding on its innovation. However, innovation also plays an important role in the assessment of
the effect of the practice on consumer welfare, e.g. where the Commission notes that the practice is
question will lessen incentives to innovate of the rivals in the result of which consumers will be
deprived of new and improved products or wider consumer choice. Consequently, innovation
considerations can be used both as a defence of dominant undertaking and as the reason for antitrust
intervention on the basis of dynamic inefficiency that derives from an anticompetitive conduct. It is
unclear which of these are stronger and why and this depends largely on the circumstances of an
individual case and should be based on the assessment of effects of the practice in question. This dual
purpose of innovation considerations is especially visible in Google Search investigations.
Google, just like IBM and Microsoft can be seen as an example of a new technology case
where a monopolist does not only threaten monopoly power but also harms rivals and competition.3
Consequently, Google is the most recent example of antitrust investigations, where competition
agencies struggle with the application of conventional tools of competition law to the environment
of new technology. Google also reflects other features of new technology markets such as being a
two-sided and partly zero-price market. It also chose to go along with a business model of
behavioural advertising in which targeting advertising and gathering information on consumers
Michael A Carrier, ‘Google and Antitrust: Five Approaches to an Evolving Issue’ (2013) Harvard Journal of Law & Technology
Occasional Paper Series — July 2013, 1 <http://jolt.law.harvard.edu/antitrust/articles/Carrier.pdf> accessed 26 March 2016.
3
4
preferences are tightly connected and were this so-called “big data” may give an important advantage
of incumbent over rivals.
Based on a number of complaints filed by Google’s (often indirect) competitors both the
Federal Trade Commission (hereinafter the FTC) and the European Commission (hereinafter the
EC), opened their parallel investigations into Google’s behaviour in the sector of online search and
advertising. The FTC investigations ended with the settlement that addressed only a few of the
original concerns, while the EC investigations are still on-going and their result is yet unknown.
Nevertheless, Google case illustrates challenges that arise in new technology markets, especially in
the area of information intermediation. Since Google Search can be qualified as an online platform,
these challenges derive, among others, from the difficulty to define the relevant market as the good
in question is information (e.g. due to problems with assessing demand substitution). In Google
Search, one can observe the existence of a number of new kinds of barriers to entry that are absent
in “conventional” markets.4. These are information costs, status quo bias or advantages are seen as a
privileged access to essential inputs – e.g. trial-and-learning effects as regards to the number of
searches being conducted that also lower the cost of delivering high-quality search results.5
As complaints are coming from Google’s competitors stating that they are unable to
compete with Google, they may be unsound and Google’s behaviour could be explained by such
factors as vigorous competition, shifting consumer demand, and their own business decisions.6 Such
claims could ‘demonstrate competitors’ efforts to compete not by investing in efficiency, quality, or
innovation, but by using antitrust law to punish the successful competitor.’7
The analysis of the role of innovation in the assessment of alleged anticompetitive abuses is
conducted on the example of two concerns expressed by the European Commission in recent
investigations into practices of Google Search. These are (1) contractual restrictions on portability
of advertising data to competing advertising platforms and (2) search bias. These two practices are
ambiguous as they are based on the ‘mixed motives’ of the dominant undertaking and one can argue
that they are either product improvements or anticompetitive practices. What is more, both of these
practices concern some kind of refusal to access to or misappropriation of information, which, due
to specific characteristics of information as an economic good further complicate the analysis.
“Conventional” means current EU approach to competition law. Such wording is used throughout the entire thesis.
Mark R Patterson, “Google and Search-Engine Market Power” (2013) Harvard Journal of Law and Technology Occasional Paper
Series July 2013, 7 <http://jolt.law.harvard.edu/antitrust/articles/Patterson.pdf> accessed 7 May 2016.
6 Geoffrey A. Manne and William Rinehart, “The Market Realities that Undermined the FTC’s Antitrust Case against Google” (2013)
Harvard
Journal
of
Law
&
Technology
Occasional
Paper
Series,
2
<http://jolt.law.harvard.edu/antitrust/articles/ManneRinehart.pdf>accessed 7 May 2016.
7 Robert H. Bork and J. Gregory Sidak, ‘What Does the Chicago School Teach About Internet Search and the Antitrust Treatment of
Google?’ 8, 4 Journal of Competition Law and Economics, 663, 663.
4
5
5
Firstly, Google has been restricting the possibility to simultaneously manage campaigns on
AdWords and competing online search advertising platforms. In investigations conducted in the US,
FTC also pointed out that restrictions on portability of advertising data impaired competition in
search advertising,8 as well as restricted ability of small businesses to use tools to manage their
advertising campaigns simultaneously on Google and on other competing advertising platforms, for
example, Bing.9 Google agreed to settle and drop contractual restrictions on multihoming.
10
In a
separate letter of commitment to the FTC Commission, Google has agreed to give online advertisers
more flexibility to simultaneously manage ad campaigns on Google’s AdWords platform and on rival
ad platforms; and to refrain from misappropriating online content from so-called “vertical” websites
that focus on specific categories such as shopping or travel for use in its own vertical offerings.11
In the EU investigations, the Commission is concerned that these restrictions create
artificial switching costs which discourage advertisers using Google's AdWords from running
parallel online search advertising campaigns on competing platforms, thereby reducing consumer
choice. These restrictions do not yield any benefits for advertisers or consumers, but stifle the
development of innovative campaign management tools.12 Customers ended up with limited choice
while competitors of Google have been facing reduced incentives to innovate as Google's conduct
limited their access to customers.13 Where network effects are particularly strong we can presuppose
that the restriction on portability can be abusive.14
Secondly, “search bias” has been analysed both in the US and EU. However, while in the US
it was considered pro-competitive and “competition on the merits” (though the FTC was, as it had
Federal Trade Commission, ‘Google Agrees to Change Its Business Practices to Resolve FTC Competition Concerns In the Markets
for Devices Like Smart Phones, Games and Tablets, and in Online Search. Landmark Agreements Will Give Competitors Access to
Standard-Essential Patents; Advertisers Will Get More Flexibility to Use Rival Search Engines’ (Press Release, 3 January 2013)
<http://www.ftc.gov/news-events/press-releases/2013/01/google-agrees-change-its-business-practices-resolve-ftc> accessed 10
April 2016
9
Federal
Trade
Commission,
Google
Press
Conference
(Final
Transcript,
3
January
2013)
<http://www.ftc.gov/sites/default/files/documents/videos/google-press-conference/130103google-pc.pdf> accessed 22 March 2016.
10 Federal Trade Commission, ‘Google Press Conference, Opening Remarks of Federal Trade Commission Chairman Jon Leibowitz’
(Press Conference 3 January 2013) 4 <https://www.ftc.gov/sites/default/files/documents/public_statements/opening-remarks-federaltrade-commission-chairman-jon-leibowitz-prepared-delivery/130103googleleibowitzremarks.pdf> accessed 22 March 2016.
11Federal Trade Commission, ‘Google Agrees to Change Its Business Practices to Resolve FTC Competition Concerns In the Markets
for Devices Like Smart Phones, Games and Tablets, and in Online Search. Landmark Agreements Will Give Competitors Access to
Standard-Essential Patents; Advertisers Will Get More Flexibility to Use Rival Search Engines’ (Press Release, 3 January 2013)
<http://www.ftc.gov/news-events/press-releases/2013/01/google-agrees-change-its-business-practices-resolve-ftc> accessed 10
December 2015
12 European Commission, ‘Commission seeks feedback on commitments offered by Google to address competition concerns –
questions and answers’ (Memo, Brussels, 25 April 2013) MEMO/13/383 <http://europa.eu/rapid/press-release_MEMO-13383_en.htm> accessed 23 March 2016.
13 ibid
14 Case T-201704 Microsoft Corp v Commission [2007] ECR II-3601, [2007] 5 CMLR 11, 657, para 562.
8
6
recently occurred, not been unanimous on this issue15), in the EU, it became the main concern of
investigations that is also based on a quite vague theory of harm – named by some scholars as the
“exclusionary discrimination”. In the EU, the EC defined preliminary search bias as an ‘unfavourable
treatment of services of search providers in Google's unpaid and sponsored search results coupled
with an alleged preferential placement of Google's own services’.16 According to the European
Commission, Google prominently displays links to its own specialised search services (in particular
to its comparison shopping product, currently called 'Google Shopping') within its web search
results and does not seem to inform users of this favourable treatment.17 Before a major shift in
investigation has occurred, and the Commission issued its recent Statement of Objections,18 Google
has been offering few sets of commitments to address these concerns, however, they were found
unsatisfying. Moreover, as new complaints were being filed by its rivals who claim that they are not
visible enough, and would require a larger space,
19
a fresh approach to the problem would be
welcomed.
The paper is therefore not only aiming at the identification of the relevance of innovation
considerations for abuses of dominance that arise in digital markets in general. It also strives to
predict the outcome of one of the most-followed EC competition investigations, the result of which
is still unknown. The European Commission has not, until Google Search case, analysed an issue
which was so deeply rooted in the Internet,20 and we can observe concerns that digital monopolies
may lead to the situation of ‘replay of the domination of monopolistic trusts in the late nineteenth
century’21 and have disastrous consequences such as entrenchment of these monopolies to other
markets.22 This is because digital monopolies may be also artificially preserved and protected from
See Brody Mullins, Rolfe Winkler and Brent Kendall, ‘Inside the U.S. Antitrust Probe of Google. Key FTC staff wanted to sue Internet
giant after finding ‘real harm to consumers and to innovation’’The Wall Street Journal (19 March 2015)
<http://www.wsj.com/articles/inside-the-u-s-antitrust-probe-of-google-1426793274>accessed 19 June 2016.
16 European Commission, ‘Antitrust: Commission probes allegations of antitrust violations by Google’ (Press release, Brussels, 30
November 2010) IP/10/1624 <http://europa.eu/rapid/press-release_IP-10-1624_en.htm?locale=en> accessed 14 March 2016.
17 European Commission, ‘Commission seeks feedback on commitments offered by Google to address competition concerns –
questions and answers’ (Memo, Brussels, 25 April 2013) MEMO/13/383 <http://europa.eu/rapid/press-release_MEMO-13383_en.htm> accessed 23 March 2016
18 European Commission, ‘Antitrust: Commission sends Statement of Objections to Google on comparison shopping service; opens
separate formal investigation on Android’ (Press Release. Brussels, 15 April 2015) <http://europa.eu/rapid/press-release_IP-154780_en.htm> accessed 15 March 2016
19 Joaquín Almunia, ‘The Google antitrust case: what is at stake?’ (European Commission, European Parliament Hearing, Brussels, 1
October 2013) < http://europa.eu/rapid/press-release_SPEECH-13-768_en.htm> accessed 24 March 2014
20 See Alexander Italianer, ‘Competition Policy in the Digital Age’ (47th Innsbruck Symposium –” Real sector economy and the Internet
- digital interconnection as an issue for competition policy”, Innsbruck, 7 March 2014)
16
<http://ec.europa.eu/competition/speeches/text/sp2014_01_en.pdf> accessed 16 April 2016.
21 Frank Pasquale, ‘Paradoxes of Digital Antitrust Why the FTC Failed to Explain Its Inaction on Search Bias’ (2013) Harvard Journal
of Law & Technology Occasional Paper Series, 19 <http://jolt.law.harvard.edu/antitrust/articles/Pasquale.pdf> accessed 20 February
2016.
22 Tim Wu, The Master Switch: The Rise and Fall of Information Empires (Atlantic Books, 2010).
15
7
the actions of regulators and politicians.23 However, it is unclear whether the “conventional”
competition law should be updated to accommodate the digital environment.
The main research question of this paper is ‘How innovation considerations are
implemented in practice in cases of abuse of dominance occurring in digital markets?.’
2. EU Competition Policy and Innovation
2.1.
What is innovation?
To start with, we need to define what innovation is. Innovation is an integral aspect of platform
competition. In digital markets innovation is a production process for many goods.24 Innovation
consists of discovery, development, and commercialisation of new and improved products and
processes.25 Innovations may be of several kinds, common typologies distinguish between (1)
discrete and complex, (2) radical and incremental, (3) disruptive and sustaining, (4) user and
manufacturer innovations.26 In this paper, we are particularly interested in sustaining and disruptive
innovations. The former denotes ‘modest changes to existing products that serve existing customer
needs’, the later, ‘technological breakthroughs that are completely different from existing products
and often render them obsolete.’ 27
This distinction is important when we try to assess which kind of innovations the EU
competition policy should be focusing on. This is because, although sustaining innovation occurs
more often and therefore brings improved products to the market, the disruptive one offers
consumers increased choice and newer products. Therefore, the challenge to competition policy is
to prevent incumbents from excluding potential competitors from the market, but at the same time,
not to discourage them from improving their products and generally, not to discourage other
companies from aiming at dominance. Competition policy should focus more on rewarding
innovation, however, dominant and successful firms should not be able to block innovation, even if
it would challenge their business in the future.28 In markets in which competition is about dynamic
circles of technological change and not price competition, intervention by antitrust authorities or
Frank Pasquale, ‘Paradoxes of Digital Antitrust Why the FTC Failed to Explain Its Inaction on Search Bias’ (2013) Harvard Journal
of Law & Technology Occasional Paper Series, 19 <http://jolt.law.harvard.edu/antitrust/articles/Pasquale.pdf> accessed 10 April 2016.
24 Howard Shelanski ’Information, Innovation, and Competition Policy for the Internet’ (2013) 161 University of Pennsylvania Law
Review 1663, 1686.
25 Michael A. Carrier, Innovation for the 21st Century, Harnessing the Power of Intellectual Property and Antitrust Law (OUP 2009),
19.
26 ibid, 26
27 ibid. 26 – 27.
28 Spencer Weber Waller and Matthew Sag, ’Promoting Innovation’ 100 Iowa Law Review, 2223, 2228.
23
8
courts may discourage innovation by preventing firms from engaging in risky Research and
Development.29
One way of promoting innovation is to allow the presence of effective competition, which
is known as the so-called ‘stick’ approach to innovation.30 Here, a legal and regulatory system
maximises the incentives and opportunities for innovation of non-dominant firms. Non-dominant
firms challenge dominant companies who aim at an exclusionary conduct which lessens competitors’
incentives to invest and innovate.
31
Moreover, competition policy should also aim at making new
products available so that consumers could affect the demand for particular products and benefit
from it.32
The role of antitrust in the face of innovation is ambiguous.33 Modern competition law is
considered unsuitable for accommodating “competition in innovation” that is one in ‘future
technology and future product markets by investing in R&D’.34 It also seems that competition law
is hostile to new and innovation. This should change as such hostility leads to an increase in
enforcement errors.35 Although “new” denotes uncertainty, a number of antitrust interventions have
been in fact undertaken under uncertainty as they considered a new product, however, they often
relied on flawed economic analysis.36 Despite the fact that it would be advisable to give innovation
more credit in competition analysis, it is still rarely so. Applying innovation paradigm would mean
that we would carefully assess the effects of behaviour and decisions on incentives to invest and
innovate.37 However, it is unclear how EU competition law will analyse the innovation-related
issues, as no criteria were introduced by the EC on how to apply innovation in practice. 38
Howard Shelanski ’Information, Innovation, and Competition Policy for the Internet’ (2013) 161 University of Pennsylvania Law
Review 1663, 1675.
30 Steven Anderman and Hedvig Schmidt, EU Competition Law and Intellectual Property Rights. The Regulation of Innovation (OUP
2011) 12.
31 Spencer Weber Waller and Matthew Sag, ’Promoting Innovation’ 100 Iowa Law Review, 2223, 2228 – 2229.
32 J Gregory Sidak and David J Teece, ’Dynamic Competition in Antitrust Law’ (2009) 5, 4 Journal of Competition Law and Economics,
581, 600.
33 Geoffrey A. Manne and Joshua D. Wright, ’Innovation and the Limits of Antitrust’, 6, 1 Journal of Competition Law and Economics,
153, 156.
34 Josef Drexl, ”Anticompetitive Stumbling Stones on the Way to a Cleaner World: Protecting Competition in Innovation without a
Market” (2012) forthcoming in Journal for Competition Law and Economics, 12-08 Max Planck Institute for Intellectual Property and
Competition Law Research Paper, 2 <http://ssrn.com/abstract=2070099> accessed 2 May 2016.
35 Geoffrey A. Manne and Joshua D. Wright, ’Innovation and the Limits of Antitrust’, 6, 1 Journal of Competition Law and Economics,
153, 169.
36 ibid, 169.
37 Rupprecht Podszun, ‘The More Technological Approach: Competition Law in the Digital Economy’ in Gintare Surblyte (ed.)
Competition on the Internet (Springer 2015), 108.
38 Pablo Ibáñez Colomo, ’Restrictions on Innovation EU Competition Law’ (2016) 41 European Law Review (forthcoming), LSE Law,
Society and Economy Working Papers 22/2015, 3. https://www.lse.ac.uk/collections/law/wps/WPS2015-22_Colomo.pdf accessed 25
April 2016.
29
9
2.2.
Innovation considerations in digital markets
Innovation in digital platforms is not an episodic activity, but a component of an industry, and R&D
is the central input of production. For example, search engines need to revise and refine their
algorithms to match consumer queries to search results.39 Consequently, no digital product or
service stays the same as its features and functionality change on the daily basis. What is more,
innovation in digital markets is conducted not only by companies offering the products or services,
but also by customers themselves as they figure out new ways of using digital platforms and
applications on their own advantage. 40
In dynamic markets an assessment of how conduct affects competition is a regular procedure.
In such markets, instead of lowering prices, companies race to introduce new or improved products
and sequential monopolies may aim at leapfrogging each other as they do not compete
simultaneously.41 In digital markets products have a short-life cycle and are technically complex,
moreover, one can also observe increased standardisation and the need for product compatibility and
interoperability. Competition is more focused on innovation instead of price competition and
tipping towards dominant position is more often observed, however, the dominant position is
usually very fragile. Finally, there are also problems with assessing product substitutability. 42 This
is because a popular product may be substituted by a totally new one that has not yet entered the
market. Therefore, products that at first glance do not seem to pose a competitive threat on a
monopolist, may be considered as future substitutes taking into account the fast changing phase of
product characteristics. To assess substitutes we need to define “potential market” that is market for
the input that has not been advertised before or a market for a hypothetical input. 43 Moreover, as
some digital companies may expand and start offering a larger variety of services the borders between
markets become blurred.44 Products in the new economy are often technologically complex. While
some products in such markets are substitutable based on price, they may not be substitutable based
on other factors. This leads to the situation where an assessment of product substitutability could
lead to a narrower market definition than it really is.45 Products can therefore differ in terms of
quality or other features such as popularity etc.
Howard Shelanski ’Information, Innovation, and Competition Policy for the Internet’ (2013) 161 University of Pennsylvania Law
Review 1663, 1685.
40 ibid.
41 ibid, 1692 -93.
42 Jonathan Faull and Ali Nikpay Faull (eds.), Faull and Nikpay: The EU Law of Competition (3rd edn, OUP 2014) 386.
43 Thorsten Käseberg, Intellectual Property, Antitrust and Cumulative Innovation in the EU and US (Hart Publishing 2012) 154.
44 European Data Protection Supervisor, Preliminary Opinion of the European Data Protection Supervisor, Privacy and competitiveness
in the age of big data: The interplay between data protection, competition law and consumer protection in the Digital Economy, March
2014, available at https://secure.edps.europa.eu/EDPSWEB/webdav/shared/Documents/Consultation/Opinions/2014/14-0326_competitition_law_big_data_EN.pdf
45 Jonathan Faull and Ali Nikpay, The EU Law of Competition (3rd edn, OUP 2014) 386.
39
10
To accommodate innovation paradigm, competition policy could rely e.g. on “innovation
economics”, which is based on the premise that policies which aim at encouraging innovation may
cause some minor “deadweight” loss to the economy, since the gains deriving from these policies –
increased productive and adaptive efficiency – will outweigh these loses. Therefore, an approach to
antitrust based on innovation economics would focus more on the impact of actions in the
marketplace on productivity and innovation, than on short-term price effects.46 Competition policy
as applied to unilateral conduct in digital markets could also borrow from the concept of “Innovation
Market Analysis”47 (hereinafter the IMA) which is usually applied to mergers. The IMA focuses on
identifying innovation competition often in the absence of current product markets. Unlike the
conventional assessment of current product markets, the IMA allows identifying direct competitors
in innovation. What is more, such an innovation market can consist of resources such as well-trained
staff, experience and know-how or intellectual property rights. 48
2.3.
Should the EU competition policy protect disruptive over sustaining innovation?
The idea of “innovative efficiencies” assumes that producers are challenged by their rivals to
continuously innovate and lower the cost of producing their technology.49 The same seems to hold
in dynamic markets. However, it may make a difference which of the kinds of technological
innovation will be of highest interest in digital markets – sustaining or disruptive? Sustaining
technological innovation is one that offers demanding, high-end consumers better performance. This
can be done in the form of incremental yearly improvements. What is more, battles to offer sustaining
innovation are usually won by incumbents as they have resources to win.50 Disruptive innovation
introduces products and services that revolutionise the direction of the market. These are usually
simpler, more convenient and less expensive and therefore attract new or less-demanding
consumers.51 Battles to offer disrupting innovation are usually won by entrants.52
Robert D. Atkinson and David Audretch, ‘Economic Doctrines and Approaches to Antitrust’ (2011) 2011-01-02 Indiana UniversityBloomington:
School
of
Public
&
Environmental
Affairs
Research
Paper
Series,
6
<http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1750259> accessed 26 May 2016.
47 Richard J. Gilbert and Steven C. Sunshine, ’Incorporating Dynamic Efficiency Concerns in Merger Analysis: The Use of Innovation
Markets,’ (1995) 63 Antitrust Law Journal 569, 771 – 574
48 Benjamin R. Kern, ’Innovation Markets, Future Markets, or Potential Competition; How Should Competition Authorities Account for
Innovation Competition in Merger Reviews?’ (2014) 37, 2 World Competition, 173, 180.
49 Steven Anderman and Hedvig Schmidt, EU Competition Law and Intellectual Property Rights. The Regulation of Innovation (OUP
2011) 14.
50 Clayton M. Christensen and Michael E. Raynor, The Innovator's Solution: Creating and Sustaining Successful Growth (Harvard
Business Review Press 2013), 46.
51 ibid.
52 ibid, 44.
46
11
Investing in sustaining innovations does not denote immunity from disruptive one. 53 Every
product will eventually be replaced by something better. We can observe the so-called “virtuous
cycle” where at first the popular product becomes valuable to users but then loses its value.54 A
number of examples illustrate such a cycle. For example, AM radio being mostly replaced by FM one
after 30 years of attempts to do so. The reason for that were the entry barriers, mostly artificial, often
legal, and created by a monopoly (AM radio)55. This process has been called by Wu “the Kronos
effect.”56
Competition policy has a larger role for disruptive innovation, as due to its radical nature,
it has a larger impact on consumer welfare and existential threat to rivals. That is incumbents have
higher incentives to block disruptive innovation.57 What is more, nowadays such disruptive
innovation can also be undergone “on the cheap” and they may enjoy network effects which can lead
to market tipping.58 Disruptors are emerging players who offer products that are cheaper and more
convenient who offer disruptive products that serve new customers who were previously not given
enough attention to.59 It is still ambiguous why some companies are more capable of engaging in
disruptive innovation than others.60
In digital markets the so-called “disruptive waves” occur more often than in traditional
brick-and-mortar industries. Therefore, innovation in such markets has to be constant and natural
response to challenges and changes in the market.61 Companies in digital markets are fully aware of
this possibility of these disruptive waves and will try to prevent them at any stake. Consequently,
competition law should be more forgiving in regards to acquisition of market power, however alert
in policing their maintenance.62 How can then competition law prevent dominant companies from
behaviour that would prevent the occurrence of creative destruction in the future as such ideas
would have a negative effect on innovation?63 What is more, competition authorities should enable
the process of disruptive innovation and ensure that it incumbents do not prevent it by their actions.
Alexandre de Streel and Pierre Larouche, ’Disruptive Innovation and Competition Policy Enforcement’, Organisation for Economic
Co-Operation and Development, Global Forum for Competition, Session III, 3 DAF/COMP/GF (2015)7, 20 October 2015.
54 Shapiro, C. and Varian, H.R., Information Rules: A Strategic Guide to Network Economy (Harvard Business School Press, Boston,
Massachusetts, 1999) 180.
55Tim Wu, The Master Switch. The Rise and Fall of Information Intermediaries (Atlantic Books 2010) 125 – 135.
56 ibid, 25
57 Alexandre de Streel and Pierre Larouche, ’Disruptive Innovation and Competition Policy Enforcement, Organisation for Economic
Co-Operation and Development, Global Forum for Competition, Session III, DAF/COMP/GF(2015)7, 20 October 2015, 6.
58 ibid, 3.
59 Richard Karlgaard, The Soft Edge: Where Great Companies Find Lasting Success (Jossey-Bass 2014) 7.
60 J Gregory Sidak and David J Teece, ’Dynamic Competition in Antitrust Law’ (2009) 5, 4 Journal of Competition Law and Economics,
581, 597.
61 Richard Karlgaard, The Soft Edge: Where Great Companies Find Lasting Success (Jossey-Bass 2014) 1.
62 Spencer Weber Waller and Matthew Sag, ’Promoting Innovation’ 100 Iowa Law Review, 2223, 2224.
63 ibid, 2227.
53
12
This however not an easy task as assessing a threat to disruptive innovation is prone to risk of Type
I and Type II errors.64
3. Strength of consideration on incentives to invest and innovate in digital markets
In this part we assess which of the two famous theories on innovation will have stronger implications
in digital markets – Schumpeter’s or Arrow’s. We are particularly interested in incentives to invest
and innovate of an online platform, such as Google. ‘Online platforms’ include companies such as
search engines, social media, e-commerce platforms, app stores, and price-comparison websites,
which play a central role in social and economic life.65 Moreover, as many of online platforms also
operate as intermediaries and take a form of two-sided markets, the conventional methods of market
power analysis are insufficient to address problems arising in such an environment. 66 Due the fact
that in such markets competition occurs more in future markets than in current ones, competitors’
incentives to invest and innovate as well as the notion of potential competition gain an increased
importance.
The discussion, Schumpeter v. Arrow, although unresolved and widely commented on is
still very useful as it sheds a light on different factors that need to be taken into account when
analysing an environment in which innovation may arise. This discussion should be continued as to
new technology markets, but it will always be a discussion “in progress”. Most of the discussion
focuses on situations where information is protected by IPR, however e.g. trade secrets could be
taken into account as well. The strength of incentives to invest and innovate of a digital monopolist
is however difficult to assess. Would such a monopolist have less incentives to innovate and prefer
to resort to practices that foreclose his competitors or leverage his market power to adjacent markets
in order to keep its dominant position on the market? Or is he forced to innovate since he is
constantly challenged by its competitors who may, any time in the future challenge and destroy the
temporary digital monopoly?
Alexandre de Streel and Pierre Larouche, ’Disruptive Innovation and Competition Policy Enforcement, Organisation for Economic
Co-Operation and Development, Global Forum for Competition, Session III, DAF/COMP/GF(2015)7, 20 October 2015, 5.
65 See European Commission (2015), ‘A Digital Single Market Strategy for Europe’, Communication to the European Parliament, the
Council, the European Economic and Social Committee and the Committee of the Regions, COM (2015) 192 final.
66 Monopolkommission. Competition policy: The challenge of digital markets. Special Report by the Monopolies Commission pursuant
to Section 44(1) (4) of the Act against Restraints on Competition (Summary of the Special Report 68: Competition policy: The challenge
of digital markets, Special Report by the Monopolies Commission pursuant to Section 44(1)(4) ARC, 1 June 2015) 1
<http://www.monopolkommission.de/images/PDF/SG/SG68/S68_summary.pdf> , accessed 21 February 2016
64
13
According to Schumpeter, a monopolist has higher incentives and also funds to innovate
than a small firm.67 Moreover, what keeps the monopoly going are ‘new consumers’ goods, the new
methods of production or transportation, the new markets, the new forms of industrial organization
that capitalist enterprise creates.’68 For example, Google is investing in its own vertical search
engines. Nevertheless, as Schumpeter’s thesis is applicable to product not technology markets, it
does not take into account complements, and this significantly narrows the choice of substitution.
This is problematic as innovation in fact enhances the value of complements.69 Not taking
complements into account would narrow the choice of substitutes and would be inacceptable in new
technology markets. This seems to be a major flaw of the Schumpeterian approach. Still, as noted
earlier, a monopolist is more willing to invest in sustainable innovation than in disruptive one. He
will do anything to stop rivals from dethroning him. To protect his position, digital monopoly may
become involved in the so-called “defensive leveraging”.70 He may be keeping its position in the
market in order to drive competitors out of it and impede innovation so that new products and
services could not enter the market.
Nevertheless, Schumpeter,71 also introduces the idea of “creative destruction” where new
technologies repeatedly displace older ones which leads to higher economic growth than the
ordinary forces of price competition.72 “Creative destruction” therefore ‘incessantly revolutionises
the economic structure from within, incessantly destroying the old one, incessantly creating a new
one.’73 In the process of “creative destruction” an incumbent tries to resist new forms of innovation
which affect the nature of competition. However, even these newly successful firms, that creatively
destruct, will eventually be replaced as this cycle is indefinite.74 Therefore, according to Schumpeter,
technological innovation occurs in waves by different technologies, as he calls it - “perennial gales of
creative destruction”. Antitrust law should aim at promoting creative destruction by reducing
barriers to entry.75
Joseph A. Schumpeter, Capitalism, Socialism and Democracy (Routledge 2003) 83 see also J Gregory Sidak and David J Teece,
’Dynamic Competition in Antitrust Law’ (2009) 5, 4 Journal of Competition Law and Economics, 581, 598 – 599.
68 Joseph A. Schumpeter, Capitalism, Socialism and Democracy (Routledge 2003) 82 – 83.
69 J Gregory Sidak and David J Teece, ’Dynamic Competition in Antitrust Law’ (2009) 5, 4 Journal of Competition Law and Economics,
581, 595.
70 Tim Wu, The Master Switch: The Rise and Fall of Information Empires, First Vintage Book Edition (2011).
71 Joseph A. Schumpeter, Capitalism, Socialism and Democracy 81 (Routledge, 1942, reprint edition 1976).
72 Hebert J. Hovenkamp, Competition for Innovation, University of Iowa, Legal Studies Research Paper 13-26, 2013, pp. 4 – 5. Note
to Schumpeter, Joseph A. Schumpeter, “Capitalism, Socialism and Democracy 81 (1942, reprint edition 1976).
73 Joseph Schumpeter, Capitalism, Socialism and Democracy (3rd edn 1950), 97 – 98.
74 ibid 84,, after Spencer Weber Waller and Matthew Sag, ’Promoting Innovation’ 100 Iowa Law Review, 2223, 2226.
75 Spencer Weber Waller and Matthew Sag, ’Promoting Innovation’ 100 Iowa Law Review, 2223, 2229.
67
14
Arrow claims that it is more likely that innovation, especially the disruptive one, would
come from small outsiders that could come up with disruptive innovations.76 Therefore, the incentive
to invent is less under monopolistic than under competitive conditions.77 Rivalry, especially
aggressive one will offer consumers new services and better terms.78
The problem that arises here is that for disruptive innovation to succeed, it has to draw to
itself consumers that are either new or unserved by the disrupted product or services or that are
drawn away from that product or service in the first place. This may be difficult to achieve where
consumers become locked-in into existing product or standard. Due to Google’s popularity, or also
because some of the users believe that other horizontal search engines do not appeal to them, some
users may be reluctant to switch to alternative search engines. This can be partly explained by status
quo bias, which denotes that ‘doing nothing or maintaining one’s current or previous decision is
almost always a possibility.’79 Therefore, although it seems that in new technology markets barriers
to entry are either non-existent or very weak, consumers may prefer to stick to the “old” standards
instead of incurring significant learning costs. This is mostly done for convenience reasons. Status quo
bias therefore explains why consumers prefer to stick with the default option, even though superior
alternatives can be easily accessed.80 What is more, a new network may be killed before it is even
started as consumers who join the new network early may fear being stranded if everyone else
chooses a different network. Therefore, the new network has to convince its new customers that it
will be successful, this can be done e.g. by offering a better price etc.81 In order to draw customers to
itself a newcomer should offer them a better deal. Even though switching costs and network benefits
may be aiming at preventing new service or products from succeeding, it may still be more attractive
to consumers than the older one. However, an incumbent, to protect its dominant position the may
invest in improvements that a rival has, which will strengthen its dominant position even further.82
Even though the new entrant may be offering a better product or service, it will still lack
the scale of this trial-and-error effects. They denote a situation in which search results are adapted
to previous searchers of consumers as well as their preferences. Trial-and-learning effects are then
See Kenneth J. Arrow, Economic Welfare and the Allocation of Resources for Invention in the Rate of and Direction of Inventive
Activity: Economic and Social Factors, (Richard Nelson, 1962), 609.
77 Kenneth Arrow, ‘Economic Welfare and the Allocation of Resources for Invention’ in The Rate and Direction of Inventive Activity:
Economic and Social Factors, Universities-National Bureau 1962, 619.
78 Howard Shelanski ’Information, Innovation, and Competition Policy for the Internet’ (2013) 161 University of Pennsylvania Law
Review 1663, 1668.
79 William Samuelson and Richard Zeckhauser, ‘Status Quo Bias in Decision Making’ (1988) 1, 1 Journal of Risk and Uncertainty, 7,
8.
80 Maurice E. Stucke, ‘Behavioural Antitrust and Monopolization’ (2012) 8, 3 Journal of Competition Law and Economics 545, 552.
More on behavioural antitrust in Amanda P Reeves and Maurice E Stucke, ‘Behavioural Antitrust’ 86 Indiana Law Journal, 1527 –
1586.
81 Suzanne Scotchmer, Innovation and Incentives (The MIT Press 2004), 295.
82 ibid, 296.
76
15
considered as a main competitive disadvantage in attracting consumers and advertisers. 83 The trialand-error learning effects have to be however conducted on the sufficient scale.84 These effects may
give Google an insurmountable advantage over its rivals, however, this does not denote that they
could not be achieved by newcomers as regards to a disruptive product or service.
According to Arrow, an incentive to invest can exist even under perfect competition in the
product markets, however not in the “market” for the information contained in the invention. This
observation holds especially when an invention reduces cost. However, monopoly would denote
more incentives to innovate in situations where the firm is able to retain added value from its
products and services – e.g. by charging a royalty for a product improvement.85 Therefore, the only
case where we can say that monopoly denotes more incentives to innovate is that appropriability
may be greater under monopoly than under competition.86
Breakthrough innovations do not generate wealth to the same extent as sustaining, followon innovations do. This is because follow-on innovations in fact adapt the breakthrough innovation
to accommodate social needs. They bring more income and create job opportunities. This wealth
cannot be achieved without accessing the knowledge about breakthrough innovation by creator’s
competitors.87 The aim of disruptive innovation is to attract consumers away from the current
market. The aim of the disruption is to displace the incumbent from its position 88
Therefore, the fact that a company is an incumbent with a large market power does not
denote that it will innovate as innovation is not connected to the firm size.
89
What is more, the
opposite is also true as a small company may be able to find funding for its disruptive innovation.
This can be found from inside sources, such efficient sale of products but also outside ones. For
example, capital markets could provide financing for future projects. Moreover, some product
development goals can be achieved by collaborative organizational agreements, such as joint
ventures or co-production and co-marketing agreements. 90
One source of revenue that Google has and that allows it to fund its innovations is the
possession of huge amount of data for the most accurate search results, that can be later on sold to
Maurice E. Stucke, ‘Behavioural Antitrust and Monopolization’ (2012) 8, 3 Journal of Competition Law & Economics, 545, 555.
ibid, 548.
85 Kenneth J. Arrow, ‘Economic Welfare and the Allocation of Resources for Invention’, National Bureau of Economic Research, The
Rate and Direction of Inventive Activity: Economic and Social Factors, Universities-National Bureau Committee for Economic
Research, Committee on Economic Growth of the Social Science Research Council, Princeton University Press, 1962, 609, 619.
86 ibid, 622.
87 Björn Lundqvist, Standardization under EU Competition Rules and US Antitrust Laws. The Rise and Limits of Self-Regulation
(Edward Elgar 2014) 95.
88 Alexandre de Streel and Pierre Larouche, ’Disruptive Innovation and Competition Policy Enforcement, Organisation for Economic
Co-Operation and Development, Global Forum for Competition, Session III, DAF/COMP/GF(2015)7, 20 October 2015, 6.
89 ibid, 598.
90 J Gregory Sidak and David J Teece, ’Dynamic Competition in Antitrust Law’ (2009) 5, 4 Journal of Competition Law and Economics,
581, 590.
83
84
16
advertisers. Big data can be then considered as an insurmountable barrier to entry, but control over
personal data can be used to entrench monopoly power, since owning data allows targeting users
with ads posted on Google by advertisers, which leads to increase in Google’s revenues. In analyses
of Google’s behaviours, surprisingly little attention is being paid to the control Google entrenches
over personal data of its users. This observation could be used as a counterargument for claims that
Google owns its dominant position to innovations it has invested in.91 Consequently, Google may be
trying to leverage is market power to earn more revenues by moving into selling big data. Still,
nothing prevents newcomers from gathering such data by themselves, although this process could
be lengthy.
To sum up, as it is not easy for small companies to take over the market with disruptive
innovation, however, it is also not entirely impossible, especially in digital markets which often do
not require expensive pre-investments and long prior phases of Research and Development. We can
therefore conclude that in slow-moving markets concentration is more conductive to innovation,
however, disruptive innovations are more likely to be introduced by newcomers, not incumbents.
Therefore, in digital markets monopolists will aim at making their products or services better, but
will be less likely to work on new and revolutionary technology if they are not challenged by their
competitors. This situation is problematic as in digital markets revenues actually comes from
sustaining innovation.
Nevertheless, all kinds of monopolies, whether ones exhibiting network effects or not, will,
eventually end, or lose some of their market share. This can occur either because of antitrust
investigations, divestiture or by a process of destructive innovation, were a new product or service
is about to enter the market and take it over.92 Therefore, the relevant question is ‘how long
monopoly might last on the Internet, but when information is the commodity, a monopoly’s
longevity is potentially great and always hard to predict.’93 The digital monopoly is still usually
temporary, nevertheless where it limits innovation significantly, competition authority should
intervene by promoting disruptive innovation. It is however unclear how competition authority
would assess that innovation is limited – either directly or indirectly. Moreover, digital monopoly
can also defend itself by means of objective (business) justifications based on dynamic efficiencies.
Both of these concerns are analysed in a further part of this paper.
Nathan Newman,’ Search, Antitrust and the Economics of the Control of User Data’ (2014) 31, 2 Yale Journal on Regulation, 401,
401.
92 Tim Wu, ‘Taking Innovation Seriously: Antitrust Enforcement if Innovation Mattered Most’ (2012) 78, 2 Antitrust Law Journal, 316,
318.
93 Ibid, 320.
91
17
4. Phases of competition analysis in which incentives to invest and innovate can be taken
into account
Innovation considerations gain more prominence in antitrust policy due to the shift in competition
authorities’ enforcement priorities. It is more common for the European Commission to support its
involvement in a case on the basis of innovation considerations. 94 However, in digital markets special
responsibility should not be entrenched. What is more, it should be counterbalanced by objective
justification defence. It also seems that when assessing the effects of the practice on consumers, the
Commission may aim at proving that the practice in question affects innovation in an
anticompetitive way. However, should the Commission also make sure that its intervention will not
affect innovation in a negative way?
It seems that innovation considerations are of the highest value in objective justifications
(especially business justifications), e.g. a monopolist can refuse to share an input such as IPR or trade
secret to prevent free riding on its innovation. However innovation also plays an important role in
the assessment of the effect of the practice on consumer welfare, e.g. where the Commission notes
that the practice is question will lessen incentives to innovate of the rivals in the result of which
consumers will be deprived of new and improved products or their choice of goods will be limited.
Consequently, innovation considerations can be used both as a defence of dominant undertaking and
as the reason for antitrust intervention.
4.1.
The role of objective justifications by digital monopolies
The Article 102, unlike 101 TFEU, does not state explicitly that a dominant undertaking could try to
justify its allegedly anticompetitive behaviour.95 However, the European Commission allows the use
of justification, as it had stressed in its Guidance on Article 82 [now 102].96 Consequently, although
some practices are considered as abusive, they may be justified on the basis of objective arguments.
In Irish Sugar it has been stressed that
even if the existence of a dominant position does not deprive an undertaking placed in that
position of the right to protect its own commercial interests when they are threatened (…),
the protection of the commercial position of an undertaking in a dominant position with
the characteristics of that of the applicant at the time in question must, at the very least,
in order to be lawful, be based on criteria of economic efficiency and consistent with the
Pablo Ibáñez Colomo, ’Restrictions on Innovation EU Competition Law’ (2016) 41 European Law Review (forthcoming), LSE Law,
Society and Economy Working Papers 22/2015, 2- 3. https://www.lse.ac.uk/collections/law/wps/WPS2015-22_Colomo.pdf accessed
25 April 2016
95 Tjarda van der Vijver,’Objective Justification and Artilce 102 TFEU’ (2012) 35,1 World Competition, 55, 55.
96 Guidance on the Commission’s Enforcement Priorities in Applying Article 82 of the EC Treaty to Abusive Exclusionary Conduct by
Dominant Undertakings’ [2009] OJ C45/02,
94
18
interests of consumers. In this case, the applicant has not shown that those conditions
were fulfilled.97
Objective justification can play an important role in an assessment of possible effects of a practice in
question. In particular, one should analyse business justifications as well as the importance of assets
such as business secrets. Objective justification can be submitted by a dominant undertaking upon
receiving of which the Commission is assessing whether such a justification can be in fact used. 98
Business decisions made at early stages of development and production will affect later on
competition in the relevant markets, however this impact is often indirect and of remote nature. 99
The dominant company introduces objective justification by demonstrating that its
conduct is objectively necessary or it produces substantial efficiencies which outweigh any
anticompetitive effects on consumers. 100 In this context, the Commission will assess whether the
conduct in question is indispensable and proportionate to the goal allegedly pursued by the
dominant undertaking.101 The conduct may be objectively necessary for health or safety reasons
which are related to the nature of the product in question.102 This denotes explanations other than
efficiencies. It can also take into account technical or commercial requirements relating to the
product or service. Moreover, demonstrating that the product is objectively necessary does not
require balancing between the negative and positive effects.103
In this context, the dominant undertaking will generally be expected to demonstrate, with
a sufficient degree of probability, and on the basis of verifiable evidence, that the following
cumulative conditions are fulfilled:
— the efficiencies have been, or are likely to be, realised as a result of the conduct. They
may, for example, include technical improvements in the quality of goods, or a reduction
in the cost of production or distribution,
— the conduct is indispensable to the realisation of those efficiencies: there must be no
less anti-competitive alternatives to the conduct that are capable of producing the same
efficiencies,
— the likely efficiencies brought about by the conduct outweigh any likely negative
effects on competition and consumer welfare in the affected markets,
— the conduct does not eliminate effective competition, by removing all or most existing
sources of actual or potential competition. Rivalry between undertakings is an essential
driver of economic efficiency, including dynamic efficiencies in the form of innovation. In
its absence the dominant undertaking will lack adequate incentives to continue to create
and pass on efficiency gains. Where there is no residual competition and no foreseeable
Case T-228/97 Irish Sugar plc v Commission ECR 1999 II-02969, 189.
European Commission, “Communication from the Commission – Guidance on the Commission’s Enforcement Priorities in Applying
Article 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings,”, OJEU C45/02 (2009), para 31.
99 Josef Drexl, ”Anticompetitive Stumbling Stones on the Way to a Cleaner World: Protecting Competition in Innovation without a
Market” (2012) forthcoming in Journal for Competition Law and Economics, 12-08 Max Planck Institute for Intellectual Property and
Competition Law Research Paper, 8 <http://ssrn.com/abstract=2070099> accessed 2 October 2014.
100 European Commission, ‘Communication from the Commission – Guidance on the Commission’s Enforcement Priorities in Applying
Article 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings’ [2009] OJ C45/02, para 30.
101 ibid., para 28.
102 ibid. para 29.
103 Jonathan Faull and Ali Nikpay Faull (eds.), Faull and Nikpay: The EU Law of Competition (3rd edn, OUP 2014) 395.
97
98
19
threat of entry, the protection of rivalry and the competitive process outweighs possible
efficiency gains. In the Commission's view, exclusionary conduct which maintains,
creates or strengthens a market position approaching that of a monopoly can normally
not be justified on the grounds that it also creates efficiency gains.104
Efficiencies have to outweigh the negative effect of the practice is question. For example, in
the British Airways105, it has been stressed that such efficiencies can only be proved based on the whole
of the circumstances of the case. Moreover, the exclusionary effect of the practice can be
counterbalanced or outweighed, by advantages in terms of efficiency which also benefit the
consumer. Where the exclusionary effect has nothing to do with the efficiencies, the practice must
be regarded as an abuse.106 In Intel, it was required for the conduct to be capable of achieving the
legitimate goal, that it had no equally effective alternative in achieving the legitimate goal with a less
restrictive or less exclusionary effect and that the conduct was “proportionate” Proportionality
denotes that ‘the legitimate objective pursued by Intel should not be outweighed by the exclusionary
effect.’107
Business justifications should be used on a wider scale as efficiency justifications. This can
be enhanced by a number of possible policy options such as more comprehensive guidelines,
precedent cases, lowering standard of proof, obligation on the Commission to review efficiency
justifications, integrating efficiency assessment into the competitive assessment or even a change of
the welfare standard that is applied in competition procedures – from consumer to total welfare. 108
In cases involving unilateral conduct, competition on the merits may lead to exclusion of inefficient
competitors from the market, and consequently, illegal monopolisation or abuse of dominance.109
What is more, business decisions made at early stages of development and production will later on
impact competition in the relevant markets, however this impact is often indirect and of remote
nature.110
4.1.1.
Objective justifications for restrictions on portability of online data
One of the main reasons for the growing importance of advertising in Google’s business model lays
in the merger between the search giant and Double Click.111 The merger has allowed Google to apply
ibid.
Case C-95/04 British Airways v Commission [2007] ECR I-02331
106 ibid, para 86.
107 Intel Corporation COMP/37.990 European Commission, [2009] OJ C227/07, para 1624.
108 Hans W Friederiszick and Linda Gratz, ‘Hidden Efficiencies. On the Relevance of Business Justifications in Abuse of Dominance
Cases’ (2013) 0 (0) Journal of Competition Law & Economics, 1, 22 -23.
109 Joseph Drexl, ’Real Knowledge is to Know the Extent of One’s Own Ignorance: On the Consumer Harm Approach in InnovationRelated Competition Cases., Max Plank Institute for Intellectual Property, Competition and Tax Law Research Paper Series No. 0915, 2.
110 ibid, 8.
111 See Google/ DoubleClick (Case COMP/M.4731), Regulation (EC) No 139/2004 Merger Procedure [2008]
<http://ec.europa.eu/competition/mergers/cases/decisions/m4731_20080311_20682_en.pdf>accessed 15 March 2015.
104
105
20
an algorithm of a much higher quality and gave it a competitive advantage, since Google and
DoubleClick both operate on the same market that is online advertising - Google provides online
advertising space while DoubleClick is a seller of ad serving services.112
In order to provide advertising services, Google Search has developed a special AdWords
platform. In AdWords advertisers are able to select which words and phrases are relevant to their
business and use them as their keywords. When Google’s users conduct a search for these particular
words, sponsored links are displayed alongside with search results. Each time a user clicks one of
those sponsored links, the advertiser pays Google.113 Therefore, nowadays, Google’s primary revenue
derives from its provision of advertising services.
Few important particularities make Google’s advertising system particularly successful.
Firstly, this is due to the fact that Google’s ads are contextualised and personalised to match
searchers’ queries. Secondly, also the fact that Google operates worldwide draws a number of
advertisers to Google. Thirdly, Google does not fix prices of “sponsored links” as they are organised
according to the bidding system. Finally, advertisers are charged a price according to the so-called
“cost per click”, so if users are not directed to the advertisers’ website, no additional cost need to be
incurred, which is why online advertising is better tailored to the needs of advertisers than
traditional one.114
Although the European Commission has noted that search advertising, due to its
profitability, is an important part of every search engine's business,115 it did not put this observation
sufficiently into practice. In fact, online advertising is able to target consumers as to their
preferences, previous searches or even location,116 which leads to an increasing importance of
behavioural factors in the analysis of product market. Since such advertising is taking into account
the behaviour of users, it is known as behavioural and is the most valuable source of revenue of online
platforms. Behavioural advertising is the most popular revenue model in online search and
advertising markets.117
Sophie van Loon, ‘The Power of Google: First Mover Advantage or Abuse of a Dominant Position?’ (2012), in Aurelio LopezTarruella (ed), Google and the Law, Empirical Approaches to Legal Aspects of Knowledge-Economy: Business Models, Series, 22
Information Technology and Law Series (Springer 2012) 24.
113 Aurelio Lopez-Tarruella,’ Introduction: Google Pushing the Boundaries of Law’ in Aurelio Lopez-Tarruella (ed), Google and the Law,
Empirical Approaches to Legal Aspects of Knowledge-Economy: Business Models, 22 Information Technology and Law Series, 3.
114 ibid, 3 – 4.
115 European Commission, ‘Antitrust: Commission seeks feedback on commitments offered by Google to address competition
concerns, European Commission’ (Press release 25 April 2013) IP/13/371 <http://europa.eu/rapid/press-release_IP-13-371_en.htm>
accessed 10 March 2016.
116 Jonathan Levin and Paul Milgrom, ‘Online Advertising: Heterogeneity and Conflation in Market Design’ (2010) American Economic
Review: Papers & Proceedings 100 , 603, 603
117 Pamela Jones Harbour and Tara Isa Koslov, ‘Section 2 in a Web 2.0 World: an Expanded Vision of Relevant Product Markets’
(2010) 76 Antitrust Law Journal, 769, 780.
112
21
Google could try to defend itself by claiming that information on how ‘Ad Words’ works
and how to port advertising data from its advertising platform to other is a trade secret. Trade secret
is valuable and confidential information that gives the enterprise a competitive advantage and can
be therefore easily misappropriated. One cannot qualify as a trade secret an information that could
be obtained from other sources on the market.118 Trade secret is not a form of IP right but it is closely
analogous to patents, however it is chosen by the firms that prefer not to disclose the information. 119
As the protection of trade secrets is not yet officially regulated in the whole EU, for now there are
only national laws to be applied to such information.120 The directive on trade secrets in preparation
and aims at protection of undisclosed know-how and business information (trade secrets) against
their lawful acquisitions, use and disclosure.121
A clear link between trade secrecy and innovation can be identified namely
a firm that holds, as a secret, information that is valuable and that it makes reasonable
efforts to protect can sue anyone who misappropriates the information. Firms tend to elect
trade secrecy rather than patents when they do not wish to disclose information. But
unlike the patent, copyright, and antitrust laws, the law of trade secrets has not erected
innovation roadblocks.122
Google has already tried to justify the API’s restriction on the grounds that it ensures that
third-party intermediaries take advantage of unique features available on AdWords.123 By restricting
portability Google is therefore protecting itself from revealing how AdWords works and is trying to
prevent free-riding. Where Google may be preventing free-riding on its investment, sharing this
information would deprive it from incentives to invest and innovate. This is because, although
confidential information is uneconomical to duplicate, it is also easy to share and such sharing may
be disastrous for Google’s business. Moreover, we should be careful using “convenience criterion”
instead of factual and legal criteria for its application. Imposing a duty to share information has been
also criticised by one of the FTC Commissioners, Rosh, who, in his (dissenting) opinion, claimed
that such an action ‘could discourage innovation, especially in the software industry’ and there are
Jonathan Faull and Ali Nikpay, Faull and Nikpay: The EU Law of Competition (3rd edn, OUP 2014) 155
ibid, 1448.
120 See Parliament and the Council Proposal for a Directive on the protection of undisclosed know-how and business information (trade
secrets) against their unlawful acquisition, use and disclosure, COM/2013/0813<http://eur-lex.europa.eu/legalcontent/EN/TXT/?uri=CELEX:52013PC0813> accessed 15 March 2016.
121 European Commission, ‘Trade Secrets’ <http://ec.europa.eu/growth/industry/intellectual-property/trade-secrets/index_en.htm>
accessed 25 March 2016.
122 Michael A. Carrier, Innovation for the 21st Century, Harnessing the Power of Intellectual Property and Antitrust Law (OUP 2009),
21.
123 J. Thomas Rosh, ‘Concurring and Dissenting Statement of Commissioner J. Thomas Rosch Regarding Google's Search Practices
- In the Matter of Google Inc., dissenting and concurring opinion’, In the Matter of Google Inc., (FTC File No. 111-0163, January 3,
2012) 6, <https://www.ftc.gov/sites/default/files/documents/public_statements/concurring-and-dissenting-statement-commissionerj.thomas-rosch-regarding-googles-search-practices/130103googlesearchstmt.pdf>accessed 19 July 2015.
118
119
22
also other alternative options for dealing with interoperability issues such as e.g. private antitrust
actions.124
4.1.2. Objective justifications that could be used by Google as regards search bias
In the EU, the EC defined preliminary search bias as an ‘unfavourable treatment of services of search
providers in Google's unpaid and sponsored search results coupled with an alleged preferential
placement of Google's own services’.125 Google has been accused of manipulating its search results
so that they would prefer their own vertical search services, in particular Google Shopping. Google
has been accused by its competitors of not delivering the results that would serve consumers best
and ones that alter search results in such a way so that Google can serve its own interests instead. 126
This has been classified by some scholars as exclusionary and anticompetitive conduct, however
some claim that Google should be free to provide whatever results they like.127
Nevertheless, it is also in Google’s own interest to constantly improve its search and
innovate. Innovation adds value to Google Search from the perspective of both consumers and
advertisers.128 Moreover, where Google, as a search engine, is designed to be subjective that is to
‘point you to destination sites that have the information you are seeking, not to send you to other
search engines’,129 the fact that Google offers nowadays also vertical search engines can be a proof of
it trying to create a better search experience.130
Google has in fact tried to defend itself by responding to the Statement of Objections
claiming that its business increases choice for European consumers and offers valuable opportunities
for businesses of all sizes. Therefore, in Google’s view, the Commission does not seem to recognise
the benefits to consumers and advertisers that Google’s business creates. Moreover, the Statement
of Objections does not, in Google’s view, back the claim up or provide a clear legal theory to connect
its claims with its proposed remedy’.131 Moreover, the SO does not recognise presence of other major
shopping services like Amazon or eBay in the market for comparison shopping or the fact that Google
ibid.
European Commission, ‘Antitrust: Commission probes allegations of antitrust violations by Google’ (Press release, Brussels, 30
November 2010) IP/10/1624 <http://europa.eu/rapid/press-release_IP-10-1624_en.htm?locale=en> accessed 14 March 2016.
126 Mark R Patterson, “Google and Search-Engine Market Power” (2013) Harvard Journal of Law and Technology Occasional Paper
Series July 2013, 2 <http://jolt.law.harvard.edu/antitrust/articles/Patterson.pdf> accessed 7 October 2015.
127 ibid, 3.
128 Robert H. Bork and J. Gregory Sidak, ‘What Does the Chicago School Teach About Internet Search and the Antitrust Treatment of
Google?’ 8(4) Journal of Competition Law and Economics, 663, 666.
129 The Incredible Stupidity Of Investigating Google For Acting Like A Search Engine, Danny Sullivan on November 30, 2010 at 7:52
am, http://searchengineland.com/the-incredible-stupidity-of-investigating-google-for-acting-like-a-search-engine-57268 accessed on
December 4, 2014.
130 ibid.
131
Google,
‘Improving
quality
isn’t
anti-competitive,
(Google
Europe
Blog,
27
August
2015)
<http://googlepolicyeurope.blogspot.fi/2015/08/improving-quality-isnt-anti-competitive.html> accessed 23 March 2016 .
124
125
23
has succeeded to develop new ways to organize and rank product information and to present it to
users in useful formats in search and ads.
132
However, according to the European Consumer
Organisation, there is no evidence of Google offering superior services than its competitors. What is
more, Google places its own search results in a hot spot where most users click.133
Without a doubt, Google has been constantly improving its search services in order to
match preferences of both searchers and advertisers. In particular, the introduction of AdWords
represents a major shift in Google’s business model, which is primarily based on competition for
eyeballs and advertising-based internet that includes vertical search engines.134 In fact, some of the
strategic moves of Google, such a switching from “10 blue links”-model to Universal Search
(promoting own vertical search services)135 or fighting with ban link farms, spandexes and websites
that do not add any content,136 gave a rise to antitrust problems. In particular, in recent years,
Google’s business model has changed significantly as regards offering search services, since it has
moved into other areas of search such as weather predictions, flight tracking, currency conversion
etc. These specific areas of search offered by Google have moved from search results page up to above
other organic search results. Google now keeps searchers on its own page instead of directing users
to third-party websites.137 It is now in competition with other websites which also offer specific
areas of search.138 Google offers a number of complementary services that improve search
functionality and attract searchers to its search engine. These include e.g. Google News, Google
Images, Google Patents, Google Scholar, etc. On one hand, offering such complementary services can
be seen as a “co-investment” which means that “the more you use it, the more it can tailor its offerings
to you”.139 On the other hand, these services are offered to keep users within services offered by
Google.
ibid.
The European Consumer Organisation, ‘Google case. Questions and Answers’ (2014) BEUC-X-2014-025<http://www.beuc.org/publications/beuc-x-2014-025_ama_google_questions_and_answers_april_2014.pdf> accessed 10 December
2014.
134 Andrea Renda, ‘Searching for harm or harming search? A look at the European Commission’s antitrust investigation against Google’
(2015) 118 CEPS Special Report, 6 <https://www.ceps.eu/system/files/AR%20Antitrust%20Investigation%20Google.pdf> accessed
21 March 2016
135 ibid. 8
136 ibid, 9.
137 Eric Goldman, ’Revisiting Search Engine Bias’ (2011) 38, 1 William Mitchell Law Review, 96, 103.
138 ibid, 104.
139 Frank Pasquale, ‘Paradoxes of Digital Antitrust Why the FTC Failed to Explain Its Inaction on Search Bias’ (2013) Harvard Journal
of Law & Technology Occasional Paper Series, <http://jolt.law.harvard.edu/antitrust/articles/Pasquale.pdf> accessed 20 November
2014.
132
133
24
Algorithm require a constant adaptation to users’ search needs, if Google wished to profit
from online search. Innovation is a better tool for addressing algorithms than lengthy regulatory
process.140
What is more, Google itself is trying to justify its involvement in search bias by the use of the socalled “convenience” argument. It is assuming that consumers are better off when they e.g. find a
desired address in a few seconds; as such information is ready and available in Google’s Maps. Google
may be right to assume that consumers do prefer to find the desired information faster. This is
because many of the consumers in their status quo-state most probably do not even wonder or
complain why it is so convenient to conduct such a search. In this matter, Google is using the fact
that consumers are often biased in their decisions and that their decisions are rather intuitive than
conscious.141 The fact that Google is the largest search engine and the most popular one too, may also
have much to do with the fact that it provides search results much quicker and they are also more
accurate.142 Because Google constantly improves the design of its algorithm, impairing such
improvements could be more harmful than useful. Therefore, going back to “ten blue links”, would
be desired by Google’s competitors but have a negative effects on its users.143
According to FTC, Google’s primary reason for changing the look and feel of its search
results to highlight its own products was to improve the user experience. Similarly, changes to
Google’s algorithm that had the effect of demoting certain competing websites had some plausible
connection with improving Google’s search results, especially when competitors often tried to game
Google’s algorithm in ways that benefited those firms, but not consumers looking for the best search
results. According to Google, its rivals engaged in many of the same product design choices that
Google did, suggesting that this practice benefits consumers.144
Although search bias is thought to affect innovation, it seems that in fact it is in itself the
result of innovation incurred by Google and consequently a product improvement. Improvements to
Google’s algorithm lead to increased quality of search results that are difficult to compete with, even
by means of innovative efforts. No other search engine is prevented from competing with Google,
and challenging it by offering new or improved products. However, to be able to “dethrone” Google
Howard Shelanski ’Information, Innovation, and Competition Policy for the Internet’ (2013) 161 University of Pennsylvania Law
Review 1663, 1695 – 1696.
141 See Daniel Kahneman, Thinking Fast and Slow, (Penguin Books 2012).
142 Daniel Zimmer, ‘Digital Markets. New Rules for Competition Law’ (2015) Journal of Competition Law and Practice (editorial) 1 <
doi: 10.1093/jeclap/lpv049>accessed 20 March 2016.
143 Marina. Lao, ‘Neutral Search as a Basis for Antitrust Action?’ (2013): 26, 2 Harvard Journal of Law and Technology, 11
<http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2245295> accessed 1 June 2013
144
Federal
Trade
Commission,
‘Google
Press
Conference’
Final
Transcript
(3
January
2013)
5<http://www.ftc.gov/sites/default/files/documents/videos/google-press-conference/130103google-pc.pdf>accessed on 5 May 2014
140
25
rival search engine should compete by means of disruptive innovation as incremental changes are
insufficient to lure Google’s customers to alternative search engines.
What is then being affected by innovation is quality of search results that is to a large extent
based on Google’s algorithm.145 There is therefore little doubt that Universal Results and One Boxes
improve quality of results for users compared with the display of a list of blue links. Results are
displayed in a more user-friendly and organised way. Consequently, although a number or vertical
search engines complain that their prominence and visibility in internet searches has worsened, this
situation may be the result of competition on the merits and not an anticompetitive practice that
negatively affects innovation.146’
Competitors may be also foreclosed due to strong network effects, namely trial-andlearning effects. Google has an advantage in the form of an algorithm, which gets better every time
someone conducts a search. It therefore seems that the learning effect of an algorithm in combination
with large amounts of personal data owned by Google may be a source of a dominance. Network
effect can also be treated as a “buzz” word for “scale” of search. Due to these network effects, search
results provided by Google may be still, even despite being biased, of higher quality than the ones
offered by its competitors, such as Bing or Yahoo!. This is because competitors lack user data that
make search more accurate since Google can offer search results that are adjusted to users’
preferences. Each search feeds into Google’s algorithm by providing specific information on what is
being searched, by whom and how often. This information is given by searchers for free in exchange
for free search. The so-called “big data” give Google a significant advantage over rivals that is based
on economies of scale and scope. For a new entrant it may be difficult to predict search terms as he
will lack the scale of trial-and-errors experimentation. The incumbent has then a competitive
advantage over entrants.147
4.2.
Effect of practice on consumers
Another sphere where innovation comes into play when assessing possible anti-competiveness of an
abuse in question is when the effect of the practice on consumers is being analysed. When analysing
exclusionary abuses the Commission recommends focusing on an abuse that leads to a foreclosure
which hinders competition and thereby harms consumers.148 However, the current ordo-liberal
145
ibid.
146 See Renato Nazzini, ‘Google and the (Ever-stretching) Boundaries of Article 102 TFUE’ (2015) 6, 5 Journal of European Competition
Law and Practice, 301, 308.
147 Maurice E. Stucke, ‘Behavioural Antitrust and Monopolization’ (2012) 8, 3 Journal of Competition Law & Economics, 545, 555.
148 European Commission, ‘DG Competition Discussion Paper on the Application of Article 82 [now 102 TFEU] of the Treaty to
exclusionary abuses’ (Staff Discussion Paper) [2005] para 56, see also European Commission, ‘Communication from the Commission
— Guidance on the Commission's Enforcement Priorities in Applying Article 82 of the EC Treaty to Abusive Exclusionary Conduct by
Dominant Undertakings’ (2009) OJ C 45/02, para 19.
26
approach to the Article 102 TFEU still does not require that an investigation includes the effects that
the possible foreclosure would have on the overall efficiency as well as on consumer welfare.149 Even
though the courts do not require consumer harm as a proof of foreclosure, the Commission
recommends considering it in the analysis.
According to the EC, to foreclose, the dominant undertaking should be able to influence to
its own advantage and to the detriment of consumers at least some parameters of competition such
as e.g. prices, output, innovation, the variety or quality of goods or services.150 Therefore, while
assessing the impact of the abuse on consumer welfare, we look whether it leads to any inefficiencies.
Antitrust is based on a premise that competitive markets are ones where not only ‘allocative
efficiencies’ and ‘productive efficiencies’ occur, but also innovative and dynamic efficiencies. 151 In
digital markets, we should increasingly focus on dynamic efficiencies that is harm innovation. It
could be visible in lessened consumer choice and or limiting the occurrence of new or better
products. However, it is ambiguous how would this harm to innovation be measured.
In fact, “consumer choice” may well be a much better criterion for identifying a restraint of
competition in innovation-related cases than the more general “consumer harm”. The consumer
choice criterion leaves the decision of what should be preferred to the market to consumers
themselves.152 As competition agencies are not able to correctly predict the future preferences of
consumers, they may be better off letting consumers decide for themselves.153
Dynamic inefficiency of the practice in question therefore denotes that it affects innovation,
however, under the EU competition law, there is yet no possibility of claiming that the direct harm
to innovation has occurred. Instead, one has to first prove that the practice indirectly leads to
anticompetitive foreclosure that has also affected innovation.154 This requirement, that to be a
competitive concern the harm to innovation, should be secondary to exclusion makes much sense.
This is because exclusionary practices are considered an alternative to innovation, and therefore may
Roger van den Bergh, Peter D. Camesasca, European Competition Law and Economics: A Comparative Perspective, (2nd edn,
Sweet and Maxwell 2006) 250.
150 European Commission, “Communication from the Commission – Guidance on the Commission’s Enforcement Priorities in Applying
Article 82 of the EC Treaty to Abusive Exclusionary Conduct by Dominant Undertakings,”, OJEU C45/02 (2009): para 11.
151 Steven Anderman and Hedvig Schmidt, EU Competition Law and Intellectual Property Rights (OUP 2011, 2nd edn), 14.
152 Joseph Drexl, ’Real Knowledge is to Know the Extent of One’s Own Ignorance: On the Consumer Harm Approach in InnovationRelated Competition Cases’, 09-15 Max Plank Institute for Intellectual Property, Competition and Tax Law Research Paper Series, 32
– 33.
153 ibid, 3 – 4.
154 Pablo Ibáñez Colomo, ’Restrictions on Innovation EU Competition Law’ (2016) 41 European Law Review (forthcoming), LSE Law,
Society and Economy Working Papers 22/2015, 3. https://www.lse.ac.uk/collections/law/wps/WPS2015-22_Colomo.pdf accessed 25
April 2016 .
149
27
cause competitive problems in industries which are primarily concerned with innovation. Exclusion
would be considered as the supreme evil if innovation would be taken into account.155
According to Colomo, introduction of indirect innovation considerations would be
welcomed in the EU competition law,
156
however, the direct introduction of innovation
considerations in the analysis, at least in the current state of EU competition law, is impossible. This
is, among others, due to high standard of proof the claimant should produce to show that the practice
in question harms innovation. What is more, the aim of the EU competition law is not to ‘fine-tune
markets and achieve optimal outcomes’.157 It may be difficult to prove the direct harm to innovation
as regards consumer welfare. It would be more evitable to introduce these claims indirectly, for
example by showing that the behaviour in question lessens consumer choice etc. or impedes
disruptive innovation by aiming at the exclusion of rivals.
4.2.1. Harm to innovation in restrictions on portability of online advertising data
To port online advertising data used in Google’s AdWords campaigns to other advertising platforms,
advertisers would need to get an access to Application program interfaces (hereinafter APIs). API
can be defined as ‘a user interface for machines to access machines,’158 and allow communication
between computer programmes and hardware In order for others to use APIs, it has to be either
published or licensed, otherwise the use of such APIs would be considered as a breach of copyright. 159
Examples of APIs include interfaces between an operating system and the programs that run on it
(to allow the programs access to the computer’s screen, keyboard and file system, for example); the
interface allowing a program to accept plug-ins providing additional functionality, such as
Photoshop or Mozilla Firefox; the interface allowing software programs to interact with Google
Maps, Gmail, Facebook and Twitter.160 In Google’s advertising platform, APIs would allow
advertisers to develop their own software programs to automatically manage and optimize their
advertising campaigns.161
Tim Wu, ‘Taking Innovation Seriously: Antitrust Enforcement if Innovation Mattered Most’ (2012) 78, 2 Antitrust Law Journal, 313,
316.
156 Pablo Ibáñez Colomo, ’Restrictions on Innovation EU Competition Law’ (2016) 41 European Law Review (forthcoming), LSE Law,
Society and Economy Working Papers 22/2015, 4 https://www.lse.ac.uk/collections/law/wps/WPS2015-22_Colomo.pdf accessed 25
April 2016 .
157 ibid, 19.
158 Andrew Katz, ‘Google, APIs and the Law. Use, Reuse and Lock-In’, in Aurelio Lopez-Tarruella, (ed) Google and the Law. Empirical
Approaches to Legal Aspects of Knowledge-Economy - Business Models, Series: (22 Information Technology and Law Series,
Springer 2012) 287.
159 ibid, 289– 290.
160 ibid, 288.
161 Federal Trade Commission, ‘Google Agrees to Change Its Business Practices to Resolve FTC Competition Concerns In the Markets
for Devices Like Smart Phones, Games and Tablets, and in Online Search Landmark Agreements Will Give Competitors Access to
Standard-Essential Patents; Advertisers Will Get More Flexibility to Use Rival Search Engines’ (Press Release, 3 January 2013)
155
28
Do restrictions on portability of advertising data reduce consumer choice or prevent the
occurrence of new or better products?
In investigations conducted in the US, FTC pointed out that restrictions on portability of
advertising data impaired competition in search advertising,162 as well as restricted ability of small
businesses to use tools to manage their advertising campaigns simultaneously on Google and on
other competing advertising platforms, for example, Bing.163 Google agreed to settle and drop
contractual restrictions on multihoming.
164
In a separate letter of commitment to the FTC
Commission, Google has agreed to give online advertisers more flexibility to simultaneously manage
ad campaigns on Google’s AdWords platform and on rival ad platforms; and to refrain from
misappropriating online content from so-called “vertical” websites that focus on specific categories
such as shopping or travel for use in its own vertical offerings.165
In the EU investigations, the Commission is concerned that these restrictions on porting
the data create artificial switching costs which discourage advertisers using Google's AdWords from
running parallel online search advertising campaigns on competing platforms, thereby reducing
consumer choice. These restrictions do not yield any benefits for advertisers or consumers, but stifle
the development of innovative campaign management tools.166 Customers ended up with limited
choice while Google’s competitors face reduced incentives to innovate as Google's conduct limited
their access to customers.167 According to the European Commission, where a publisher (the website
owner) is only allowed to place advertisement that has been provided by Google and cannot display
at all, or only to a limited extent, advertisements from other companies, web users are limited in their
choice. As Google is also a major search advertising intermediator, this reduces the availability of
services to consumers and accordingly inhibits innovation.168
<http://www.ftc.gov/news-events/press-releases/2013/01/google-agrees-change-its-business-practices-resolve-ftc>accessed
20
March 2016
162 ibid.
163
Federal
Trade
Commission,
Google
Press
Conference
(Final
Transcript,
3
January
2013)
<http://www.ftc.gov/sites/default/files/documents/videos/google-press-conference/130103google-pc.pdf> accessed 22 March 2016
164 Federal Trade Commission, ‘Google Press Conference, Opening Remarks of Federal Trade Commission Chairman Jon Leibowitz’
(Press Conference 3 January 2013) 4 <https://www.ftc.gov/sites/default/files/documents/public_statements/opening-remarks-federaltrade-commission-chairman-jon-leibowitz-prepared-delivery/130103googleleibowitzremarks.pdf> accessed 22 March 2016.
165 Federal Trade Commission, ‘Google Agrees to Change Its Business Practices to Resolve FTC Competition Concerns In the Markets
for Devices Like Smart Phones, Games and Tablets, and in Online Search. Landmark Agreements Will Give Competitors Access to
Standard-Essential Patents; Advertisers Will Get More Flexibility to Use Rival Search Engines’ (Press Release, 3 January 2013)
<http://www.ftc.gov/news-events/press-releases/2013/01/google-agrees-change-its-business-practices-resolve-ftc> accessed 10
December 2015
166 European Commission, ‘Commission seeks feedback on commitments offered by Google to address competition concerns –
questions and answers’ (Memo, Brussels, 25 April 2013) MEMO/13/383 <http://europa.eu/rapid/press-release_MEMO-13383_en.htm> accessed 23 March 2016
167 ibid
168 European Commission, MEMO/13/383, Brussels, 25 April 2013, Commission seeks feedback on commitments offered by Google
to address competition concerns – questions and answers. http://europa.eu/rapid/press-release_MEMO-13-383_en.htm
29
Are restrictions on portability of advertising data indeed affecting innovation? To answer
this question we should analyse whether it would be too expensive on impractical for Google’s
competitors to create similar campaigns on competing advertising platforms and consequently,
increase consumer choice. Similar campaigns available on different platforms would probably
increase the consumer choice as to the number of services, but not necessarily the quality.
What is more, are there other methods of porting the data?169 Creating a second system like
AdWords would not lead to large duplication costs.170 Moreover, the FTC’s Commissioner Rosh
noted that the restrictions in question prevented licensees from using software developed or used by
a third party (but not the advertiser itself) to comingle data and simultaneously manage an ad
campaign involving both Google and the rivals. The restriction did not prevent users from exporting
AdWords data onto a rivals’ platforms, which would need to done with their own software.171 These
alternative options of porting data are definitely less advantageous, however, they are available.
However, one should point out that advertisers usually do use programmes of third parties to port
the data and rarely can do that by themselves.
172
Nevertheless, we cannot deny the existence of
alternative ways available for rivals to create own advertising campaigns. According to FTC, faced
with restrictions on portability advertisers could develop their own software programmes that
would allow them to move the data from one campaign to another, however it is quite costly for them
to develop own programmes. Moreover, as advertisers are more acquainted with using third parties’
software to port the data instead, in order for the software programmers to create such a software,
they would need to be provided with interoperability information in the form of application
programming interfaces (APIs).173
As we have pointed out before, the EU competition law cannot use the theory of harm based
solely on harm to innovation, here, lessened consumer choice, without the prove of foreclosure. It is
enough that, in the absence of actual foreclosure the risk of it is possible in the long-run. Therefore,
although at the moment advertising platforms are still able to compete with Google and provide
See Case C-7/97 Oscar Bronner GmbH Co KG v Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co. KG, Mediaprint
Zeitungsvertriebsgesellschaft mbH & Co. KG and Mediaprint Anzeigengesellschaft mbH & Co. KG [1998] ECR I -7791, [1999] 4 CMLR
112, para 43.
170 ibid, para 46, see also Case C-418/01 IMS Health GmbH & Co OHG v NDC Health GmbH & Co KG [2000] ECR I – 5039, [2004] 4
CMLR 28, para 28.
171 Concurring and Dissenting Statement of Commissioner J. Thomas Rosch Regarding Google’s Search Practices In the Matter of
Google
Inc.,
FTC
File
No.
111-0163
January
3,
2012,
4
<
https://www.ftc.gov/sites/default/files/documents/public_statements/concurring-and-dissenting-statement-commissioner-j.thomasrosch-regarding-googles-search-practices/130103googlesearchstmt.pdf> accessed 22 March 2016
172 Case C-418/01 IMS Health GmbH & Co OHG v NDC Health GmbH & Co KG [2000] ECR I – 5039, [2004] 4 CMLR 28, para 44.
173 Federal Trade Commission, ‘Google Agrees to Change Its Business Practices to Resolve FTC Competition Concerns In the Markets
for Devices Like Smart Phones, Games and Tablets, and in Online Search. Landmark Agreements Will Give Competitors Access to
Standard-Essential Patents; Advertisers Will Get More Flexibility to Use Rival Search Engines’( Press release) (3 January 2013), 3
<http://www.ftc.gov/news-events/press-releases/2013/01/google-agrees-change-its-business-practices-resolve-ftc>accessed
4
October 2014
169
30
advertisers with advertising campaigns, a potential threat exists that due to Google’s restrictions on
portability, combined with exclusive obligations and Google’s possession of users’ data, other
advertising platforms will not be able to compete on equal terms.
Moreover, potential foreclosure may also derive from the fact that Google’s advertising
platform is considered by advertisers as a de facto standard and unavoidable trading partner.
Standards can have a number of positive effects, such as obvious welfare effects, ensuring
compatibility and interoperability between products from different vendors. In the result of allowing
compatibility and interoperability products offered by different firms can be combined and used
together and this increases consumer choice and convenience and reduces cost.174 Conversely, when
such interoperability is denied, these positive effects would not occur, to the detriment of consumers.
Therefore, large platforms such as Google often choose either to provide open access to the platform
or to make interoperability difficult. This may lead to occurrence of such negative effects of
standards.
To compete with Google on equal foot, its rivals would have to offer a completely new way
of advertising on the Internet that would be so much better and efficient that it could overcome pathdependence and allow consumers to switch to that new service without significant switching costs.
Alternatively, Google’s rivals would be able to compete if they could gather as significant volumes of
users’ data as Google has. This is highly unlikely because Google as a fist-mover enjoys strong
network effects that tipped the market towards Google’s standard. Finally, it also seems that Google
has managed to develop a very successful business model that is difficult to compete with, therefore,
to address this issue EC could impose on Google an order to allow interoperability from Google so
that competitors can in fact compete as regards to improved products.
It is however ambiguous which consumers’ choice is being limited. Google operates as a
two-sided market, where it brings together two interdependent groups of consumers, namely
advertisers and searchers. Therefore, is it enough that restrictions on portability harm only one group
of Google’s consumers or do we need to prove harm to both sides? Advertisers are clearly limited in
their choice as they can only place advertisements with Google, or mostly with Google. They will
most probably not switch to competing platforms as these do not have as high volumes of personal
data as Google and are therefore unable to target searchers with advertisements that would be as
accurately matching their previous searches and preferences as the ones offered by Google. Moreover,
due to limited choice, advertisers also end up paying higher prices than they would if they could
move their AdWords-based ads elsewhere. What is more, searchers do not seem to care about
174
Miguel de la Mano, ‘Competition policy, the “o-ring” of the Patent System?’ (Annual CRA Conference, Brussels, 7 December 2011).
31
advertisements that much, as they would even prefer search results that do not link to ads at all. They
are however tolerating the ads as this is the cost of search being offered for free.
It seems that the practice in question does in fact limit the choice of advertisers and may
potentially exclude them from the market. The practice prevents “multihoming” which practically
means that consumers use more than one website or application for specific tasks.175 When users
multihome, they could be targeted with the same ads on different online platforms. Multihoming is
a popular practice on the Internet and may be highly beneficial for its users. It also lessens the chance
that the market will tip towards one dominant standard and ensures more competition on the
market. 176 Since multihoming can be encouraged by an increased interoperability between different
platforms it invites competition from smaller networks, therefore prevents tipping towards one
dominant standard. However, it can also diminish the significance arising from cross-platform
network effects.177 Therefore, if special responsibility of a dominant advertising platform could be
entrenched so that publishers could use not only multiple platforms, but also one single interface for
managing multiple advertising campaigns, allowing portability of advertising data between Google
and its competitors could lead to an increased competition in the search advertising market and
therefore increased consumer choice.
As at this moment it is highly unlikely that any other search engine can gain as many users
as Google, this may change once portability is allowed. However, this does not seem to hold where
we take into account the fact that users multihome even without interoperability information and
the range of substitutes for AdWords is wider than excepted. For example, advertisers can now put
their advertisements on vertical search engines, such as Instagram, Facebook etc. Moreover, although
from consumer welfare’s point of view it would be better that the system of open standards and full
interoperability is implemented, it would only be feasible if the dominant firm would not chill
innovation as to result in a net welfare loss.178 Imposing a duty to share information on how
advertising data is ported could discourage innovation.179
As we have pointed out in the part on objective justifications available to Google as to
restrictions on portability, Google could try to defend itself on the basis that it has heavily invested
Oxera, ‘Benefits of online platforms, Report Prepared for Google’ (2015), 5 <http://www.oxera.com/getmedia/84df70f3-8fe0-4ad1b4ba-d235ee50cb30/The-benefits-of-online-platforms-main-findings-%28October-2015%29.pdf.aspx?ext=.pdf> accessed 23 March
2015.
176 ibid.
177
Andres V. Lerner, ‘The Role of “Big Data” in Online Platform Competition’ (2014), 60
<http://awards.concurrences.com/IMG/pdf/big.pdf>accessed 23 March 2016.
178 Marina Lao, ‘Networks, Access, and ”Essential Facilities”: from Terminal Railroad to Microsoft’ (2009) 62, 2 SMU Law Review, 557,
561 <http://ssrn.com/abstract=1365934> accessed 20 May 2016.
179 ibid.
175
32
in AdWords and information on how to port the advertising data could be considered as a trade
secret.
4.2.2. Harm to innovation in search bias
Recent developments prove that search bias is a more complex problem and can be seen as both an
anticompetitive practice and competition on the merits. According to the European Commission,
Google prominently displays links to its own specialised search services (in particular to its
comparison shopping product, currently called 'Google Shopping') within its web search results and
does not seem to inform users of this favourable treatment.180 Before a major shift in investigation has
occurred, and the Commission issued its recent Statement of Objections,181 Google has been offering
few sets of commitments to address these concerns, however, they were found unsatisfying.
Moreover, as new complaints were being filed by its rivals who claim that they are not visible enough,
and would require a larger space, 182 a fresh approach to the problem would be welcomed.
The SO expresses the concern that the practice may have a negative impact on rivals’
incentive to innovate. The reduced incentives would not come from their exclusion from the market,
but from the fact that their efforts to improve their services would not be rewarded in the form of a
more prominent display in the search engine. Google’s own services, irrespective of their merits,
would still benefit from preferential access.183 The Commission therefore explicitly stated that this
behaviour
(…)may therefore artificially divert traffic from rival comparison shopping
services and hinder their ability to compete on the market. The Commission is
concerned that users do not necessarily see the most relevant results in response
to queries - this is to the detriment of consumers, and stifles innovation.184
The EC has pointed out that since Google is an important source of traffic for competing
specialised search services, search bias may also reduce competitors' incentives to innovate in
specialised search.185 Search bias may lead to the harm to innovation because it reduces incentives of
European Commission, ‘Commission seeks feedback on commitments offered by Google to address competition concerns –
questions and answers’ (Memo, Brussels, 25 April 2013) MEMO/13/383 <http://europa.eu/rapid/press-release_MEMO-13383_en.htm> accessed 23 March 2016
181 European Commission, ‘Antitrust: Commission sends Statement of Objections to Google on comparison shopping service; opens
separate formal investigation on Android’ (Press Release. Brussels, 15 April 2015) <http://europa.eu/rapid/press-release_IP-154780_en.htm> accessed 15 March 2016.
182 Joaquín Almunia, ‘The Google antitrust case: what is at stake?’ (European Commission, European Parliament Hearing, Brussels,
1 October 2013) <http://europa.eu/rapid/press-release_SPEECH-13-768_en.htm> accessed 24 March 2014
183 Pablo Ibáñez Colomo, ’Restrictions on Innovation EU Competition Law’ (2016) 41 European Law Review (forthcoming), LSE Law,
Society and Economy Working Papers 22/2015, 16 https://www.lse.ac.uk/collections/law/wps/WPS2015-22_Colomo.pdf accessed 25
April 2016
184 European Commission, ‘Antitrust: Commission send Statement of Objections to Google on comparison shopping service; opens
separate formal investigation on Android’, (Press Release, Brussels, 15 April 2015) IP/15/4780 <http://europa.eu/rapid/pressrelease_IP-15-4780_en.htm>accessed on 16 April 2015.
185 ibid.
180
33
existing and potential vertical search engines to invest in ‘innovative and disruptive technologies of
specialised search186 and it also causes harm by slowing evolution of search technology in general.187
The EC pointed out that due to Google's systematic favouring of its subsequent comparison
shopping services "Google Product Search" and "Google Shopping", both experienced higher rates of
growth, to the detriment of rival comparison shopping services.188 This is especially relevant as in the
market for comparison shopping where Google is facing competition form a number of alternative
providers,189 such a behaviour could possibly be highly exclusionary. What is more, Google does not
apply to its own comparison shopping service the system of penalties, which it uses in relation to
other comparison shopping services on the basis of defined parameters, and which can lead to the
lowering of the rank in which they appear in Google's general search results pages. Froogle, Google's
first comparison shopping service, did not benefit from any favourable treatment, and performed
poorly.190
According to the EC since users do not necessarily see the most relevant comparison
shopping results in response to their queries, Google rivals’ incentives to innovate are lowered as
they know that however good their product, they will not benefit from the same prominence as
Google's product.191 The Commission is of an opinion that many of vertical services that are
potentially foreclosed by Google might be innovative and Google's practices could be therefore
limiting opportunities of European consumers to benefit from such innovative services. At the same
time, it is for users to decide whether they wish to visit these sites based on their merits.192 What is
more, also the Federal Trade Commission who has analysed search bias in its own investigations,
was concerned that condemning search bias as anticompetitive may have negative effect on
incentives to invest and innovate of all search providers.193
It also seems that the role of Google’s vertical search services is not to generate revenue, but
rather to (1) keep users within Google’s network so the internet giant can secure its continuous flow
Michael Luca, Tim Wu, Sebastian Couvidat, Daniel Frank and William Seltzer, ‘Does Google Content Degrade Google Search?
Experimental Evidence’, (2015) Harvard Business School NOM Unit Working Paper 16-035, 20 <http://ssrn.com/abstract=2667143>
accessed 18 November 2015
187 ibid, 21
188 ibid.
189 European Commission, ‘Antitrust: Commission send Statement of Objections to Google on comparison shopping service; opens
separate formal investigation on Android’, (Press Release, Brussels, 15 April 2015) IP/15/4780 <http://europa.eu/rapid/pressrelease_IP-15-4780_en.htm>accessed on 16 April 2015.
190 ibid.
191 European Commission, ‘Fact Sheet, Antitrust: Commission sends Statement of Objections to Google on comparison shopping
service’ (Memo, Brussels, 15 April 2015), MEMO/15/4781 <http://europa.eu/rapid/press-release_MEMO-15-4781_en.htm >accessed
20 April 2015
192 European Commission, ‘Commission seeks feedback on commitments offered by Google to address competition concerns –
questions and answers’ (Memo, Brussels, 25 April 2013) MEMO/13/383 <http://europa.eu/rapid/press-release_MEMO-13383_en.htm> accessed 23 March 2016
193 Robert H. Bork and Gregory Sidak, ‘What Does the Chicago School Teach about Internet Search and the Antitrust Treatment of
Google?’ 8, 4 Journal of Competition Law & Economics, 663, 666.
186
34
of users’ personal information and (2) leverage Google’s dominance to an adjacent market of vertical
search and therefore ensure its future position on the market that does in fact have a (at least future)
tendency to conduct search by the use of specialised, vertical search services. Where Google
attempts to offer a number of vertical search services, it aims at building a “minimum spanning
tree”.194 Therefore, when Google is expanding its offer, for example by offering Books or YouTube
services, and thereafter chooses to place them on its launch pages, this makes Google a more
attractive place to start the search, in the result of which users continue to search within Google due
to low cognitive cost of using it.195
However, searchers look for information on the Internet by increasingly using alternative
sources which are not characterised as general search engines.196 This is especially relevant in regards
to vertical search engines that offer information on travelling, music, books, ones that are specialising
in sharing pictures, music etc. Therefore, it does not seem that the use of Google is essential in a strict
sense of essentiality, as alternatives do exist.197 Moreover, it is also in Google’s own interest to
constantly improve its search and innovate. Innovation adds value to Google Search from the
perspective of both consumers and advertisers.198 Where Google as a search engine, is designed to be
subjective that is to ‘point you to destination sites that have the information you are seeking, not to
send you to other search engines’,199 the fact that Google offers nowadays also vertical search engines
can be a proof of it trying to create a better search experience. 200Although search bias is thought to
affect innovation, it seems that in fact it is in itself the result of innovation incurred by Google and
consequently a product improvement. Improvements to Google’s algorithm lead to increased quality
of search results that are difficult to compete with, even by means of innovative efforts. No other
search engine is prevented from competing with Google, and challenging it by offering new or
improved products. However, to be able to “dethrone” Google rival search engine should compete by
means of disruptive innovation as incremental changes are insufficient to lure Google’s customers to
alternative search engines.
194 Adam Candeub, ‘Behavioural Economics, Internet Search, and Antitrust’ (2014) 9 I/S: A Journal of Law and Policy for the Information
Society;
MSU
Legal
Studies
Research
Paper
No.
12-03,
410
<http://digitalcommons.law.msu.edu/cgi/viewcontent.cgi?article=1506&context=facpubs> accessed 10 December 2015.
195 ibid.
196 Marina. Lao, ‘Neutral Search as a Basis for Antitrust Action?’ (2013): 26, 2 Harvard Journal of Law and Technology, 7
<http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2245295> accessed 1 June 2013.
197 ibid, 8.
198 ibid.
199 Danny Sullivan, ‘The Incredible Stupidity Of Investigating Google For Acting Like A Search Engine’ (Search Engine Land, 30
November 2010) <http://searchengineland.com/the-incredible-stupidity-of-investigating-google-for-acting-like-a-search-engine57268>accessed 4 December 2014.
200 ibid.
35
What is then being affected by innovation is the quality of search results that is to a large
extent based on Google’s algorithm.201 There is therefore little doubt that Universal Results and One
Boxes improve quality of results for users compared with the display of a list of blue links. Results
are displayed in a more user-friendly and organised way. Consequently, although a number or
vertical search engines complain that their prominence and visibility in internet searches has
worsened, this situation may be the result of competition on the merits and not an anticompetitive
practice that negatively affects innovation.202 What is more, perhaps Google has no incentive to
exclude better services. Where specialised search-segment does not seem to be competitive enough,
it may be better for Google to become involved in such services itself, especially as the market for
specialised search services is expanding. Integration of general search with vertical one can also be
a strategy aiming at avoidance of a double-marginalisation problem.203 However, search bias may
deter better services for searchers/buyers, as well as monopolise matchmaking of searchers/buyers
and advertisers/sellers.204
Reduced choice of consumers may also derive from the strong status quo bias in the result
of which consumers are continuing to use Google over other search engines. It may come from the
fact that Google is in possession of large amounts of big data which allows it offer search results that
are tailored to customers’ preferences and previous searches. Therefore, searchers could switch to
alternative search engines if they were not under the influence of a status quo bias.
Google may be able to provide more accurate search results because of the design of its
algorithm. Technical information is gathered by Google in the form of data when users are browsing
and clicking on particular links and websites, and this information affects search results they receive.
Consequently, results obtained by users while browsing for unspecified and general search inquiries,
such as “restaurants in Helsinki,” may be based on searchers’ preferences and therefore, biased.
Sometimes information can be biased “on purpose”, as for example during the occurrence of tsunami
in a specific part of the world, users typing in the word “tsunami” using Google Search, may first get
hits that relate to that particular tsunami, even though they would only wish to get some general
information instead. At this point, it may be difficult to assess whether such a solution is good or
bad, it is definitely subjective. However, one may come to the conclusion that the more general the
search inquiry is, the more biased it could be if users are only wishing to find out some information
ibid.
Renato Nazzini, ‘Google and the (Ever-stretching) Boundaries of Article 102 TFUE’ (2015) 6, 5 Journal of European Competition
Law & Practice, 301, 308.
203 Lars Wiethaus, ‘Google’s Favouring of Own Services: Comments from an Economic Perspective’ (2015) 6, 7 Journal of European
Competition Law and Practice, 506, 509.
204 ibid, 510.
201
202
36
that does not have instrumental, further use, but, instead, it is information which has a value in itself
– it can be classified as an intrinsic good.
A solution to this problem could be that searchers are offered search results in two forms:
(1) a blank, unadvertised form (that would not be offered for free anymore as it would be
unsubsidised by advertisements) and (2) the biased one with advertisements (offered for free). This
situation reminds of White and Yellow Pages that used to be distributed in a paper form. In White
Pages companies were all listed in unbiased way, while in Yellow Pages only companies that actually
paid for advertising were listed in. Therefore, quality dissemination could be possibly explained by
this distinction in listings. Alternatively, search results that link to advertisements or Google’s own
content should be distinguished in a more visible way. In particular, it would be advisable that
Google’s own links to its vertical search services such as e.g. Google Shopping should be marked as
clearly as advertisements of its rivals. For now, it seems that Google’s own vertical are kind of
blended into the general search results and although they could be considered as some kind of
advertisement, they are not marked as such. It should be at least mentioned that the query is based
on advertising. However, the attempt to mark that the top search results are offered by Google has
come to a halt. Google has offered three sets of commitments, all of which were found unsatisfying
by the European Commission and also, by Google’s competitors.
5. Conclusions
Innovation in digital markets is a constant process and also a specific input into the industry.
Therefore, innovation considerations should be and luckily increasingly are taken into account in
competition investigations. However, the tools by which innovation could be embraced are still
unspecified. Competition law could borrow from the area of economics and aim at implementing
ideas such as innovation economics into its analysis. What is more, in the area of unilateral conduct
some tools applied to mergers could be of an increased use.
As we have concluded in part 3 of this paper, digital monopolist is more likely to invest in
sustaining innovation due to revenues it is able to collect from his main area of operation as well as
from the possible revenues it earns by leveraging its dominant position to adjacent markets.
Nevertheless, even the digital monopoly is only temporary and is not immune to disruptive
innovations that can come from other (or outside) market players. Therefore, it is the disruptive
innovation that the EC should be promoting and it should monitor whether incumbents do not aim
at excluding its rivals from the market to prevent this particular kind of technological innovation.
The discussion between Schumpeter v. Arrow is still up-to-date, however the later should prevail in
37
digital markets as the idea that a monopolist is more eager to innovate has to be confronted with
technological, fast-moving markets which are also based on complements. What is more, also the
European Commission favours rivalry over monopoly as the environment to promote innovation.
Without a doubt, Google has incentives to invest based on the fact that it is a network
which exhibits networks effects and therefore tips towards a monopoly. Google has also more
revenue to invest as due to the possession of large amounts of big data, it can earn more profits from
advertising that can be later on used to sponsor its leveraging attempts to other, adjacent markets.
Google, as an example of a digital monopoly may be more willing to invest in its current services, but
it is not about to invest significantly in a revolutionary products or services. Moreover, it may aim at
excluding its rivals from the market to prevent disruptive innovation.
In part we asked the following questions (a) whether Google could defend by the use of
objective justifications based on innovation or (b) is the European Commission able to prove that,
besides (or absent the requirement of) anticompetitive exclusion, the practice in question also causes
dynamic inefficiency in the form of limited consumer choice or deterring the development of new
and improved products? It seems that introduction of innovation considerations is much easier done
when it comes to objective justifications. When assessing the impact of an allegedly anticompetitive
abuse, we, in the current state of the EU law, cannot use direct harm to innovation as a working
proxy, as this harm may be only transitory or not so clear-cut. Instead, we can show that the practice
in question, besides excluding competitors, also indirectly leads to dynamic inefficiencies such as
lessened incentives to invest and innovate or lessened consumer choice. It is ambiguous whether EU
competition law should be updated to invite direct application of harm to innovation as a possible
theory of harm. Perhaps such an approach would better suit digital markets.
We have concluded in part 4.1., that digital monopolies, even despite their special
responsibility, should be able to defend themselves by the use of objective justifications. In particular,
Google could claim that information on how to port advertising data is a trade secret that it does not
wish to reveal to prevent free-riding on its investment. Moreover, as regards search bias, Google
could claim the popularity of its search engine derives from increased investment efforts and an
attempt to promote its vertical services is aimed at improving the user experience.
As regards using innovation consideration as a proof of consumer harm, we have concluded
in part 4.2. that innovation is tightly connected to dynamic efficiency. This, in digital markets would
denote increased consumer choice and new and better products. Therefore, where the behaviour of
the digital monopolist leads to exclusion and the mentioned form of consumer harm, it could be
considered as anticompetitive. As regards restrictions on portability of online advertising platforms,
we have concluded that they are not total restrictions and advertisers could design ways to port data
38
by themselves, without the help of 3rd parties, but this is highly unlikely. Nevertheless, the presence
of other vertical search engines, who also invested in advertising tools, such as e.g. Facebook, speaks
against treating this practice as anticompetitive especially when Google is able to defend itself by
showing that the practice is in fact an attempt to prevent free-riding. What is more, as regards search
bias, searchers look for information on the Internet by increasingly using alternative sources and
Google has been increasingly investing in the design of its algorithm and constantly improving its
search.
Finally, we should also accept that decisions relating to innovation are undertaken under
uncertainty, and they may lead to either false positives (classifying a legal behaviour as an
anticompetitive practice) or false negatives (classifying an anticompetitive practice as legal). As
noted by Hovenkamp,
antitrust is too blunt an instrument to detect and remedy every
anticompetitive act. Some practices will effectively be immune because our
institutions are not up to the task of identifying them without producing an
unacceptable number of false positives. (…) the social cost of false negatives is
very likely much less that the social cost of false positives. The antitrust case
law is filled with examples of practices that may be very well have been
anticompetitive, but one can never be sure enough to risk the public costs of
condemnation that might deter beneficial conduct.205
Therefore, the enforcement of competition policy is imperfect and we cannot avoid decision errors.
This premise holds especially in dynamic markets based on innovation because in such an
environment it may be difficult to decide whether it is more advisable to intervene or not. However,
in accordance to this error-cost theory, false positives are more costly than false negatives. This is so,
as markets’ self-correction mechanism is able to fix false negatives, but not false positives.
Consequently, deciding what is anticompetitive and what is not is a difficult task, as both kinds of
errors are inevitable206
Errors are unavoidable in antitrust and especially where new abuses are involved. False
negatives lead to losses which result from an anti-competitive conduct, such as ones deriving from
strategies which raise rivals’ costs or reduce rivals’ profits which may, force competitors to exit the
market or to produce on a lower scale.207 False positives, may, on the other hand, cause opportunity
losses as they prevent the incumbent from implementing a strategy that would in fact lead to an
increased allocative or productive efficiency. However, false positives can also lead to dynamic
inefficiencies.208 What is more, false positives generally lessen the incentive to strive for dominance,
Herbert Hovenkamp, The Antitrust Enterprise. Principle and Execution (Harvard University Press 2005) 50.
Geoffrey A. Manne and Joshua D. Wright, ’Innovation and the Limits of Antitrust’, 6,1 Journal of Competition Law and Economics,
153, 157
207 Thorsten Käseberg, Intellectual Property, Antitrust and Cumulative Innovation in the EU and US (Hart Publishing 2012) 82.
208 ibid, 83.
205
206
39
which would lead to loss in dynamic efficiency, which would be especially damaging in markets in
which firms compete through innovation. However, the negative effects on dynamic innovation
would be overrun by the benefits of market power.209 Still, in dynamic markets the effects of false
positives can be more damaging to consumers than in conventional antitrust cases210
Many decisions in digital markets are still undertaken under uncertainty and we cannot
avoid it. As we have pointed out, digital markets are based on potential competition which involves
hypothetical assessment of product substitutes as well as consumer preferences. Similarly, also with
Google’s behaviour, we can only hypothetically assess whether any of the two allegedly
anticompetitive practices introduced in here, are in fact harmful to innovation or whether Google
could objectively justify its behaviour. It seems that as the direct effect on innovation cannot be
proved, or is challenging to prove, the Commission should not base its theory of harm on evidence
that is not clear-cut. This is especially so, where Google is able to introduce quite convincing
objective justifications for its actions. Nevertheless, the case in question constitutes a good
background for the analysis of innovation considerations in cases of unilateral conduct that occurs
in digital markets.
ibid. 84.
Geoffrey A. Manne and Joshua D. Wright, ’Innovation and the Limits of Antitrust’, 6,1 Journal of Competition Law and Economics,
153, 195.
209
210
40