EUROPEAN
COMMISSION
Brussels, XXX
(2020) XXX draft
COMMISSION STAFF WORKING DOCUMENT
IMPACT ASSESSMENT REPORT
Accompanying the document
Digital Markets Act
EN
EN
GLOSSARY
Term or acronym
Meaning or definition
AI
Artificial intelligence
B2B
Business-to-business
B2C
Business-to-consumers
B2P2C
Business-to-platform-to-consumers
CMA
Competition and Markets Authority (UK competition authority)
CFR or Charter
Charter of Fundamental Rights
DSA
Digital Services Act
EEA
European Economic Area
ECN
European Competition Network, consisting of NCAs
ECHR
European Convention on Human Rights
FTE
Full-time equivalent
GDP
Gross Domestic Product
GDPR
Regulation (EU) 2016/679 on the protection of natural persons with regard to the
processing of personal data and on the free movement of such data, and repealing
Directive 95/46/EC ( General Data Protection Regulation )
ICN
International Competition Network
OECD
Organisation for Economic Co-operation and Development
MCAD
Misleading and Comparative Advertising Directive
NCA(s)
National Competition Authority (ies) of the EEA
NCT
New Competition Tool
OECD
Organisation for Economic Co-operation and Development
OPC
Open Public Consultation
P2B Regulation
Regulation (EU) 2019/1150 of the European Parliament and of the Council of 20
June 2019 on promoting fairness and transparency for business users of online
intermediation services
SME
Small and Medium Enterprise
TFEU or Treaty
Treaty on the Functioning of the European Union
UCPD
Directive 2005/29/EC of the European Parliament and of the Council of 11 May
2005 concerning unfair business-to-consumer commercial practices in the internal
market and amending Council Directive 84/450/EEC, Directives 97/7/EC,
98/27/EC and 2002/65/EC of the European Parliament and of the Council and
Regulation (EC) No 2006/2004 of the European Parliament and of the Council
UCTD
Council Directive 93/13/EEC of 5 April 1993 on unfair terms in consumer
contracts
CONTENTS
EN
EN
5.3.
Policy options under Pillar II
EN
EN
7.2.
How do the options under Pillar II compare?
8.1.
Description of Pillar I
8.2.
Description of Pillar II
8.3.
Functioning of the feedback loop between Pillars I and II
EN
EN
1. INTRODUCTION: POLITICAL AND LEGAL CONTEXT
1 .
The digital transformation has profoundly changed the functioning of the global economy and
society. The Covid-19 crisis and the increased importance and use of digital services has only
further evidenced the importance of ensuring a borderless, fair, and contestable Single Market
for digital services where companies can thrive and where citizens have genuine choices and
control.
2 .
This Impact Assessment examines the possible policy options to ensure a competitive Single
Market for digital services and in particular fair and contestable platform markets. It combines
the assessment of two initiatives previously presented in separate Inception Impact
Assessments: (i) the Digital Services Act
package: ex ante regulatory instrument of
very large online platforms acting as gatekeepers;1 and (ii) the New Competition Tool.2
3 .
Given the breadth of the topics covered, both inception impact assessments including their
respective consultations were initially published separately. However, since the outset, both
ensur[ing] a joint analysis of the
results
with a view to exploring synergies and ensuring consistency on the policy options
pursued, in particular as regards possible remedies and enforcement. 3
4 .
As presented in both
consultations was to consult as widely as possible through various means in order to deliver
an in-depth impact assessment of the different policy options and their perceived impact on
By means of
this holistic approach, this combined approach will allow the Commission to ensure the
proper functioning of the internal market by promoting effective competition in digital
markets, in particular a fair and contestable online platform environment. This objective feeds
ensure contestability, fairness and innovation and the possibility of market entry, as well as
public interests that go beyond competition or economic considerations.
5 .
There is a broad consensus on the benefits of digital services. However, recent years have
seen a mounting global concern due to the concentration tendencies that characterise digital
markets. These concentration tendencies and the underlying market dynamics, as well as other
characteristics of digital markets can contribute to market failures and lead to inefficient
market outcomes in terms of higher prices, lower quality, less choice and innovation to the
detriment of European consumers. Over the last two decades, a small number of online
platforms have emerged as large ecosystems, playing a crucial role. Such platforms
intermediate a significant portion of transactions between consumers and businesses, and have
6 .
A wide-range of recent studies by international as well as by National Competition
process about the role of competition policy in a fast-changing world, which included
commissioning a report a group of the independent Special Advisers to Commissioner
Vestager published in April 2019.4 These findings are further confirmed by the evidence
EN
1
EN
and its expert group5, set
up to support the Commission in monitoring and analysing the developments in the online
platform economy.
7 .
Similar reflections are taking place
the
US, Japan, the UK, and Australia, including calls for a new regulatory framework for
significant and durable market power
Majority Staff report6), substantial market power (ACCC report7), strategic market status
(Furman report8)
bottleneck power (Stigler Center9). The US report notably concludes
that each investigated platform now serves as gatekeeper; that each platform uses its
gatekeeper position to maintain its market power; and that the firms have abused their role as
intermediaries to further entrench and expand their dominance.
8 .
The need to address these concerns in digital markets was expressed in Commission President
von der Leyen mission letter for Executive Vice-President Vestager, where she stated that in
striving for digital leadership, we must focus on making markets work better for consumers,
business and society . The letter tasked Executive Vice-President with ensuring that
competition policy and rules are fit for the modern economy .10
9 .
Shaping Europe's digital
future
it is important that the competition rules remain fit for a world that is changing
fast, is increasingly digital and must become greener .11 In the same Communication, the
Commission further stated that competition policy alone cannot address all the systemic
problems that may arise in the platform economy Against this background, the Commission
also announced that it will further explore, in the context of the DSA package, ex ante rules
to ensure that markets characterised by large platforms with significant network effects acting
as gate-keepers, remain fair and contestable for innovators, businesses, and new market
entrants .12
10 . In the European Parliament, the Committee on the Internal Market and Consumer Protection
(IMCO), the Committee on Civil Liberties, Justice and Home Affairs (LIBE) and the Legal
Affairs Committee (JURI) published draft reports in April and in May 2020, as legislative
own-initiative reports.13 The final IMCO and LIBE Committees reports were adopted in
September14 and draft JURI report in October 2020.15 In parallel to these reports, the
European Parliament also adopted a resolution on competition policy on 18 June 2020, where
calls on the Commission to assess the possibility of imposing ex ante regulatory
obligations where competition law is not enough to ensure contestability in these markets 16
11 . The European Council confirmed this need to act in its New Strategic Agenda 2019-2024, by
[w]e will continue to update our European competition framework to new
technological and global market developments .17 The Council of the European Union
further explore ex ante rules to ensure that markets characterised by large platforms with
significant network effects, acting as gate-keepers, remain fair and contestable for innovators,
businesses and new market entrants 18
EN
2
EN
12 . Furthermore, the Council welcomed the public consultation on a New Competition Tool to
address structural competition problems across markets and expressed its willingness to
discussing the Co
DSA Package.19 T
that
new policy approaches for the Single Market have to be fit for the digital age [and] able to
cope with new and agile business models, especially in the digital economy 20 Finally, the
Council reiterated the importance of swift action on the DSA package in its most recent
looks forward to the
s proposal for a Digital Services
Act by the end of this year 21
2. PROBLEM DEFINITION
13 . The core problems addressed in this Impact Assessment are all linked to certain market
features of digital markets.
2.1. Market features of digital markets
14 . Digital markets for the purpose of this Impact Assessment cover both traditional information
technology markets and other markets where products or services are offered by use of digital
technologies, including products connected to the internet. This includes notably markets
where gatekeeper platforms operate, in particular B2C markets.
15 . Digital markets exhibit particular features, described in the following sections, namely:22 scale
(or scope) economies, and a high degree of vertical integration (Section 2.1.1); direct or
indirect network effects (Section 2.1.2); data dependency (Section 2.1.3); switching costs
(Section 2.1.4); asymmetric and limited information, and relatedly, behavioural biases by
consumers (Section 2.1.5); and high market concentration (Section 2.1.6).23 These features,
while not unique to digital markets, often fundamentally change the competitive process in
digital markets in particular, leading to sudden and radical decreases in competition which
may prevent markets from self-correcting.
2.1.1. Scale (or scope) economies, and high degree of vertical integration
16 . Whenever the fixed costs of entering an industry or operating a business are particularly high
compared to the variable costs of producing a particular good or service, e.g. a smartphone
operating system, or a search engine for internet searches, firms that operate at a large scale
always have a material cost (or quality) advantage over smaller rivals.24 Similarly, there may
be strong synergies between different but related business activities (such as a search engine
and a comparison shopping site) which make it difficult for rivals offering only one such
product to compete against multi-product firms covering a whole range of different products,
or even offering entire ecosystems. This problem is exacerbated where the multi-product firm
is also vertically integrated, i.e. it operates activities at both upstream and downstream levels
of the supply chain. In this case, the integrated firm may not have sufficient incentives to
grant its downstream rivals access to the inputs (such as intellectual property, or data)
produced by its own upstream affiliate. Where such inputs are crucial for success on the
downstream market, smaller, non-integrated rivals may be put at a competitive disadvantage.
EN
3
EN
Extreme returns to scale, along with network effects and the role of data (see Sections 2.1.2
and 2.1.3), have been identified as the three key characteristics of the digital economy.25
17 . There has been broad consensus among the NCAs with relevant experience26, as well as
among the respondents to the Open Public Consultation
27
extreme economies of
scale and scope
high start-up costs
high fixed operating costs and high degree of
vertical integration are an important or very important source or part of the reasons for
emerging and existing market failures. Moreover, according to the International Competition
28), 51% of the respondents indicated that
economies of scale played an important role in the assessment of market power in digital
markets in the cases that they have investigated.
18 . In its enforcement practice under Article 102 of the Treaty on the Functioning of the
European Union TFEU
Commission has found the aforementioned market features to
constitute, to varying degrees, barriers to entry in the digital sector. For instance, in Google
Shopping, the Commission based its dominance assessment for the market for general search
services among other things on the existence of barriers to expansion and entry, notably the
significant investments in terms of time and resources required to establish a fully-fledged
general search engine.29 Likewise, in its Google Android decision, the Commission found that
developing a smart mobile OS [operating system] is a costly and time-consuming process 30
As regards economies of scope, the Commission made particular reference in its Preliminary
Assessment in the Amazon e-
he ability of e-book readers to drive
sales and lock-in customers: that with its Kindle e-book reader, Amazon operates a closed
"ecosystem" (or "walled garden"). Customers who own a Kindle can use that e-book reader
only for ebooks purchased in Amazon's Kindle store. 31 In the same case, the Commission
also found substantial economies of scale for e-book retailing, in particular because of the
need to construct a sufficiently large catalogue of available titles (which requires agreements
with a large number of E-book Suppliers), and because of the scale and scope of investments
needed to set up a viable e-book distribution platform.
2.1.2. Network effects
19 . Where network effects are present, the value of a service increases with the number of users.
For instance in case of a social network, a greater number of users increases the value of the
network for each user. Such network effects, which arise among users of the same user group,
are called direct network effects. Indirect network effects, also known as cross-side effects,
typically occur in case of platforms which link at least two user groups and where the value of
a good or service for a user of one group increases according to the number of users of the
other group. For instance, the more users join a particular mobile operating system, the more
apps will be developed for this particular operating system, which in turn makes the operating
system more attractive for its users.
20 . Network effects tend to favour large providers over smaller rivals, and provide them with
market power, in a similar way as scale and scope economies. Both the NCAs with relevant
EN
4
EN
experience32, as well as the respondents to the OPC33 broadly agreed that strong
direct/indirect network effects are an important or very important source or part of the
reasons for emerging and existing market failures.
21 . One reason why network effects tend to favour large incumbents is that even aggressive initial
price offers will not necessarily allow smaller rivals to challenge incumbents and steal market
shares from them. This is so because in many industries characterised by network effects,
prices are often zero for consumers, with firms monetising their services through advertising
and/or access to consumer data (e.g. Google, Facebook).34 Therefore, there is no room for
entrants to undercut incumbents when the going price for a particular service is already zero.
There was some consensus among the NCAs with relevant experience35, and strong support
among the respondents to the OPC36, that zero-pricing markets are an important or very
important source or part of the reasons for emerging and existing market failures. Moreover,
according to an ICN Report on digital markets, 77% of the respondents indicated that network
effects played an important role in the assessment of market power in digital markets in the
cases that they have investigated.
22 . In its enforcement practice under Article 102 TFEU, the Commission has found network
effects to constitute, to varying degrees, a barrier to entry in the digital sector. For instance, in
the Microsoft case, the Commission found that network effects represented a relevant barrier
media player would not meet with significant consumer demand if there
was no or no significant amount of corresponding digital content which this player could play
back 37 In Google Android
network effects arise because, when
deciding which licensable smart mobile OS to develop for, app developers consider the
revenue potential of that OS
s,
mobile OSs with a large user base are considered mo
38
2.1.3. Data dependency
23 . Data dependency refers to scenarios where the operations of companies are largely based on
big datasets. As the Special Advisors to Commissioner Vestag
ata is a core input
factor for production processes, logistics, targeted marketing, smart products and services, as
well as Artificial Intelligence (AI). It drives interoperability in interconnected environments
and will revolutionise sectors such as mobility and healthcare. The competitive relevance of
data is consequently very important. The competitiveness of firms will increasingly depend on
timely access to relevant data and the ability to use that data to develop new, innovative
applications and products. Against this background, an important debate has emerged on
whether, and if so under which conditions and on which legal basis, public intervention is
needed to ensure sufficient and timely access. 39 Data is particularly relevant as an input for
digital services and products.
24 . There has been broad consensus among the NCAs with relevant experience40, as well as
among the respondents to the OPC41 that lack of access to a given input/asset which is
necessary to compete on the market (e.g. access to data)
data dependency
conjunction with data protection issues, are an important or very important source or part of
EN
5
EN
the reasons for emerging and existing market failures. Moreover, according to an ICN Report
on digital markets, 49% of the respondents indicated that dependency on data played an
important role in the assessment of market power in digital markets in the cases that they have
investigated.
25 .
evance in a
number of cases in the digital sector. For instance, in Apple/Shazam, the Commission found
bility to use the Customer Information to put competitors at a competitive disadvantage
42 In the Google Shopping case,
the Commission identified the availability of data in the form of user search queries, paired
-home on Google for their general searches, as an important
ecause a general search service uses search data to refine the relevance
of its general search results pages, it needs to receive a certain volume of queries in order to
compete viably. The greater the number of queries a general search service receives, the
quicker it is able to detect a change in user behaviour patterns and update and improve its
relevance. 43
2.1.4. Switching costs
26 . Switching costs are onetime expenses a consumer or business incurs or the inconvenience it
experiences when switching between products or service providers. Examples include the cost
of learning how to use a different mobile operating system, or the cost of exporting digital
libraries (such as images, music files, book files etc.) into another operating system (or else,
firms to have market power over those consumers who were successfully attracted to a given
product in the past. Like for network effects, the presence of switching costs makes it more
difficult for entrants to contest the market position of firms which have already acquired a
large customer base. The NCAs with relevant experience44, as well as the respondents to the
OPC45, generally showed that high customer switching costs are an important or very
important source or part of the reasons for emerging and existing market failures.
27 . One implication of switching costs in a platform context is that either one (or both sides) of
the platform tend to single-home for specific purposes, i.e. users only use one platform, rather
than using several platforms simultaneously.46 For instance, the vast majority of smartphone
users owns either an iPhone or an Android phone, but not both at the same time, and they tend
to be very loyal to their operating system. The fact that customers typically use one platform
(i.e. they predominantly single-home) and cannot easily switch
NCAs with relevant experience47, as well as by the respondents to the OPC48, as an important
or very important source or part of the reasons for emerging and existing market failures.
Moreover, according to the ICN Report on digital markets, 41% of the respondents indicated
that switching costs played an important role in the assessment of market power in digital
markets in the cases that they have investigated.
EN
6
EN
28 . Research on the basis of agent-based simulations also found evidence of biases that
reinforce consumer lock in, such as escalation of commitment , and availability bias 49. In
escalation of commitment , users commit themselves to one platform, even when switching
may provide higher user utility. Hence, those users never switch platforms. For instance, a
consumer purchasing on a large e-commerce marketplace offering a range of products, would
not switch to one or several other platforms even if the latter are specialised in the specific
type of goods the consumer is interested in. Convenience and user habits would prevail over
the benefit (e.g. higher quality) potentially resulting from the use of a more specialised
platform. Users subject to an availability bias may make platform choice decisions using a
heuristic that relies on vivid or recent data. For example, users may easily recall a platform
that has many users, as media and social media will be mentioning such a platform. Social
norming (e.g. follow friend
comes to high coordinated switching costs.
29 . In its enforcement practice under Article 102 TFEU, the Commission has found switching
costs to constitute, to varying degrees, a barrier to entry in the digital sector. For instance, in
the Amazon e-book MFNs
customers that have already
purchased Kindle e-books may face costs in switching to another e-book platform, due to the
need to acquire an additional e-book reader and the inability to transfer the library of e-
books purchased in Amazon's Kindle store to a different e-book reader. 50 In an internal
document, Microsoft
The Windows API
is so deeply embedded in the
source code of many Windows apps that there is a huge switching cost to using a different
operating system instead. 51 Switching costs are also relevant where customers are
businesses, not final consumers. This is demonstrated by the Google Android case, where the
OEMs wishing to switch to other licensable smart mobile OSs face
switching costs.
implement the Android OS on our devices was approximately 50 million Euro, with lead time
of 1.5-2 y
52
2.1.5.
30 . Information asymmetry on the customer side occurs when customers (consumers or
businesses) in an economic transaction possess substantially less knowledge than the other
party does, so that they cannot make informed decisions. For instance, a lack of information
on alternative price offers in the market for a particular good or service may prevent
direct price competition with their rivals, allowing firms to charge higher prices than they
otherwise could. Similarly, a lack of meaningful transparency about all relevant product
characteristics in view of the complexity of the terms and conditions of the service for an
average user, e.g. the data usage of different social networks, prevents users from comparing
different offers against each other, which in turn reduces the incentives of suppliers to
improve their offerings along those dimensions that are not observable to users.
31 .
or exacerbated by so-called behavioural biases. For instance, consumers may systematically
EN
7
EN
misjudge their own preferences, or the prices or attributes of products, leading to suboptimal
consumption choices, such as buying excessive amounts of add-on warranties, insurance
products, or gym memberships. Another example is default bias , i.e. the tendency to favour
the default option when given a choice between several options; this bias discourages
switching to different alternatives (such as a different browser, different search engine, etc.)
whenever certain software products come pre-
has similar adverse effects on competition as would limited information about the existence of
these alternatives.
32 . The feature of information asymmetry on the customer side has been recognised as
important or somewhat important source or part of the reasons for emerging and existing
market failures by the NCAs with relevant experience53, as well as among the respondents to
the OPC54. Moreover, according to the ICN Report on digital markets, 44% of the respondents
indicated that consumer bias played an important role in the assessment of market power in
digital markets in the cases that they have investigated.
33 . In its enforcement practice under Article 102 TFEU, the Commission has found behavioural
biases to constitute, to varying degrees, a barrier to entry in the digital sector. For instance, in
the Google Android
users that find apps pre-installed and
55 In other
default bias
status quo bias
-
installation of operating systems, app stores, search engines, etc, a powerful tool to lock in
In 2016, approximately 260 million smartphones were sold
in Europe, of which approximately 197 million smartphones or 76% were Google Android
devices. Practically all of these Google Android smartphones had the Google Search app pre-
installed with the rest of the GMS bundle. 56
2.1.6. High market concentration
34 . A market is considered highly concentrated if one or at most two firms hold a market share of
close to 100%. The structural market features discussed above all favour the emergence of
few large firms, or even just one super-dominant firm. There is evidence for a trend of
growing concentration (and, relatedly, growing mark-ups) at the industry level, which has
been documented both for the US and for the EU, although there is an ongoing debate
whether growth in concentration may have been slower in the EU than in the US.57
35 . NCAs with relevant experience58, as well as the respondents to the OPC59 broadly agreed that
one or few large players on the market (i.e. concentrated market)
important or important source or part of the reasons for market failures. From a competition
perspective, it may be very difficult to maintain competition in the market , and even if
possible, it may not be desirable, in a situation in which having only one network is the most
beneficial outcome for consumers. However, in such a situation it is essential to keep
competition for the market open. Any successful attempt by a firm to lock-in a group of
consumers, so that the market is no longer contestable for a new entrant, will prevent such
EN
8
EN
competition for the market , with possible adverse consequences for prices, product variety,
and innovation.60
36 . In light of the above, ensuring contestability is of utmost importance, as
n these markets,
the issues frequently arise from a combination of complex interleaving of firm conduct,
consumer behaviour, economic characteristics, technological factors, and various aspects of
regulation. Promoting competition in this sector will therefore not be purely about limiting
anti-competitive conduct, important as that is. It will also require more proactive market-
opening measures. 61
2.2. What are the problems?
37 . The market features explained in Section 2.1 lead to two main problem groups:
38 . First, in many of these digital markets, in particular B2C markets, large online platforms have
gatekeeper platforms
ome of these gatekeeper platforms
exercise control over whole platform ecosystems that are essentially impossible to contest by
existing or new market operators, irrespective of how innovative and efficient they may be.62
As a consequence contestability of these digital markets is limited even beyond the limitations
weak contestability of platform markets ). In addition, many
businesses are increasingly dependent on these gatekeeper platforms, which in many cases
leads to gross imbalances in bargaining power and, consequently, identified unfair practices
resulting in unfair business conditions for business users63 ( unfair business conditions for
business users ).
39 .
emerging and existing market failures
undermine effective competition and contestability in digital markets beyond the weak
contestability of and unfair business practices on platform markets. In the first place, business
models and market structures can change rapidly in markets with these features. The market
features set out in in Section 2.1 and the resulting market failures can also occur in markets
without gatekeeper platforms. In the second place, even within markets with gatekeeper
platforms, market dynamics can lead to the emergence of new business practices whose
effects will need to be examined in a thorough market investigation before deciding whether,
on balance, they are harmful or beneficial to competition on these markets. In the third place,
some market failures related to entrenched market positions might need to be remedied by
more comprehensive means than only addressing business practices.
2.2.1. Problems stemming from conduct by large gatekeeper platforms
40 . Many of the digital markets are characterised by a single player, and a few large platforms
that have become gatekeepers for many digital and non-digital products and services. These
gatekeeper platforms represent a key segment of the digital economy and have an important
role in providing third parties with online access to a large number of EEA consumers. For
example, more than 4 million businesses trade significantly via platforms,64 for whom these
platforms serve as important intermediaries to reach consumers.
EN
9
EN
41 . These gatekeeper platforms raise important concerns. First, their gatekeeper position and
economic power cannot be efficiently contested by actual or potential competitors due to the
importance of the market features presented above Second, due to their economic power and
imbalances in their commercial relationship, gatekeepers can undermine fair trading
conditions for their business users.65
2.2.1.1 .
Weak contestability of platform markets
42 . As mentioned above, there is an unprecedented concentration of economic power in
digital markets: seven of the large platforms account for 69% of the total EUR 6 trillion
platform economy, through vertical and horizontal integration. Large online platforms
intermediating between businesses and consumers are growing at an exponential pace. They
have several hundreds of millions of users (both businesses and citizens/consumers).66 Total
net revenues of some of these platforms (of billions of euros) double and triple over a few
years.
43 . Large gatekeeper platforms benefit significantly from the main features of digital markets
(described in Section 2.1). In this context, new entrants that may want to enter or expand in
platform markets characterised by a gatekeeper platform may find it extremely difficult to
overcome some of the inherent market features without operating a sufficiently large user
base.67 For instance, a new entrant must convince sufficient number of users (due to the
importance of network externalities) to coordinate their migration to a new service, taking e.g.
part of the social network along, or other associated data assets such as purchase or preference
histories, or ratings. This lack of contestability is extensively echoed in the academic
literature,68 pointing out that control over data by specific platforms (or a lack of venture
capital funding for businesses aiming to compete with incumbent digital platforms), are
significant barriers to entry.69
44 . These large gatekeeper platforms therefore have an entrenched market position, which is
hard to contest, and which they further expand through ecosystems. The largest platform
companies are active across many different markets, creating extended data-driven
ecosystems around their core activities, often cross-subsidising one service with data or
revenues from another. In this regard, a large numbers of respondents identified online
intermediation services, search engines, operating systems for smart devices, consumer
reviews, network and/or data infrastructure/cloud services, digital identity services, online
advertising intermediation services, payment services, fulfilment services and data
management platforms as activities that can strengthen the role of gatekeeper platforms when
any or all of these are integrated within a single corporate structure.70
45 . The (structurally) entrenched position of gatekeepers has shown to be lasting and essentially
unchallenged by competing platforms, thus leading to weak inter-platform competition. It is
sometimes argued that incumbent s services are often free and that competition is just one
click away or that it is vigorous in some segments. This is a too narrow and selective view of
the overall dynamics of the digital platform economy and market features described above
(see in particular Section 2.1).
EN
10
EN
2.2.1.2 .
Unfair business conditions for business users
46 . The relative bottleneck or economic power of gatekeeper platforms means that entrants or
and gatekeeper platforms
represent an important trading partner for them to be able to reach large numbers of
consumers in the EEA.
47 . As a result, there are stark imbalances of bargaining power between gatekeeper platforms and
business users, which allow large gatekeeper platforms to engage in unfair practices vis-a-vis
business users resulting in unfair business practices, as evidenced by competition enforcement
practice, the OPC71, including targeted submissions by stakeholders, and various studies (for
more details on the different forms of unfair practices, see Section 2.5.2).
48 . The identified unfair practices do not represent one-off problem. They are systemic and
recurrent features of the conduct by gatekeeper platforms, which are very often facilitated by
the features of the markets in which these gatekeeper platforms operate.72
2.2.2. Market failures that undermine competition and contestability in digital
markets
49 . Going beyond the identified problems of gatekeeper platforms, the market features explained
in Section 2.1 also contribute to more general, structural competition problems, in particular
in digital markets.
cases in various industries points to the emergence and existence of market failures that
cannot be tackled or that cannot be addressed in the most effective manner under the existing
EU competition rules. This is supported by numerous international studies (see Section 1).
50 . The term market failure indicates a situation in which a market does not allow consumers to
benefit from the results of effective competition in terms of low prices, better quality, as well
as more choice and innovation, while firms are able to earn supra-normal profits which are
not competed away over time.73 The following graph illustrates the stock price development
for five major big tech companies from 2014-2018.74 To the extent that stock prices reflect
market expectations of future profitability, this graph can be interpreted as measuring (future)
profits of the respective companies. When comparing these figures to the S&P 500 index75,
which grew by about 60-70% over the same period, this graph shows how the five companies
have
consistently outperformed the market average.
EN
11
EN
Figure 1 - Stock price development for 5 big tech companies, 2014-2018
51 . To get a sense of the absolute levels of profitability (rather than their development over time),
the following graph shows the revenue per employee of selected tech companies in 2019.
Given that labour costs are the relatively most important cost factor for tech companies, these
figures can be interpreted as a rough indicator for profitability levels. It is interesting to note
that the ranking of companies in this figure does not fully correspond to the ranking in Figure
1. While Netflix ranks highest in both stock price growth and revenue per employee,
evolution.
Figure 2 - Revenue (in USD) per employee of selected tech companies, 2019
EN
12
EN
52 . While markets typically feature self-correcting mechanisms, there can be obstacles that
prevent these mechanisms from operating, leading to non-transitory losses of economic
value.76 For instance, abnormally high profits in a market should not be sustainable in the long
run because they would attract new entry into this market. As the new competitors start
offering the same or very similar products as the incumbent(s), they will steal market share,
and hence profits, from them, until the abnormally high profits will gradually be competed
away. However, this self-correcting mechanism may be impaired when there are, for instance,
barriers to entry that make it very difficult or even impossible for potential competitors to
enter the market and challenge the incumbents. The first five market features described in
Section 2.1 all have in common that they may represent such a barrier to entry, because they
do not allow entrants to be cost effective (because of scale and scope economies), to replicate
ntegration), or
to induce consumers to switch away from the incumbent(s) (because of network effects,
switching costs, or asymmetric information). Such barriers of entry therefore allow
incumbents to sustain market power, which in turn leads to longer-term societal losses in
terms of higher prices and less product variety for consumers, and less dynamic innovation.77
53 . It is important to stress that the features of a market include both structural and behavioural
ones and that demand-side considerations, in particular the behaviour of customers, play an
equally important role in this regard. Therefore, in many cases, there is a combination of those
elements leading to or constituting a market failure.78
54 . Depending on whether harm is about to affect or has already affected the market, market
failures can be grouped into two main categories: (i) scenarios where there is an emerging
market failure (e.g. driven by anti-competitive monopolisation strategies or gatekeeper
scenarios; see Section 2.6.1) and (ii) existing market failures (e.g. driven by repeated
parallel leveraging strategies or tacit collusion; see Section 2.6.2).79
2.3. Why can the existing EU toolbox not deal with emerging and existing market
failures and with unfair business practices by large gatekeeper platforms?
2.3.1. Why can existing EU competition law not tackle the problems?
55 . Broadly speaking, the EU competition policy toolbox includes rules on antitrust, merger
control, state aid, and public undertakings and services. Generally, a distinction is made
between competition rules allowing for an intervention ex post or ex ante. Antitrust
enforcement under Articles 101 and 102 TFEU (and the accompanying implementing
regulations) belongs to the first category, as it aims at detecting anti-competitive behaviour by
companies that has the actual or likely effect of causing distortions of competition. They
therefore qualify as quasi-criminal proceedings. Sector inquiries are investigations that the
Commission carries out when it suspects possible breaches of the competition rules in specific
sectors of the economy. There is no possibility to impose remedies following a sector inquiry.
Merger control and state aid rules aim at preventing anti-competitive outcomes by assessing
ex ante whether a merger between undertakings or state aid would negatively affect
competition. Merger control and state aid are purely administrative proceedings.
EN
13
EN
56 . In light of the above, intervention under the existing EU competition rules can only occur if
one of the following preconditions are fulfilled: 1) a company is dominant and abuses its
dominant position, 2) there is an anticompetitive agreement, 3) there is a merger/acquisition
with EU dimension80 or 4) a Member State grants an aid. This means that the existing EU
competition rules cannot conceptually deal with existing market failures in the absence of
these preconditions (e.g. anti-competitive monopolisation by a non-dominant company with
market power below the dominance threshold).
57 . Given that the market features of digital markets are not caused by the companies operating
on a particular market, even though their distortive effects can sometimes be exacerbated
through their conduct, emerging market failures cannot be tackled on the basis of the
existing competition rules. The same applies to existing market failures, which refer to
markets that are not functioning well and delivering competitive outcomes, notably due to
their structure.
58 . In addition, in some instances, the existing EU competition rules may be able to prevent or
address a market failure, but not in the most effective manner (e.g. repeated parallel
leveraging into related markets by a dominant company). This is because the existing
competition rules do not allow for a holistic assessment of and a more principle-based
approach with regard to remedies for all markets concerned. For example, individual
instances of leveraging of market power by a dominant company into related markets can be
dealt with under Article 102 TFEU. However, if such conduct were to be carried out in a
repeated manner, the Commission would need to run several parallel investigations, without
necessarily being able to tackle the core problem effectively (see Section 2.6.2).
59 . In antitrust cases the Commission cannot aim at going beyond the infringement at stake to
address an underlying structural problem. Remedies must be limited to what is necessary, not
more than that, to achieve its goal. Thus, the remedies in antitrust cases must go as far as
necessary to bring the infringement to an end and should be the least restrictive possible. If
there are several appropriate measures that can address the infringement, the least onerous one
must be imposed.
60 . In light of the above, there is an enforcement gap which prevents effective and timely
intervention based on the existing EU competition rules against emerging and existing market
failures.
61 . In addition, the existing EU competition rules do not necessarily capture all unfair business
practices by large digital gatekeepers. This is because these practices do not necessarily have
an anticompetitive object or effect under Article 101 TFEU, or may not be captured by Article
102 TFEU, if the platform is not found to be dominant, or if there is no effect on
competition.81
2.3.2. Why is other EU legislation not sufficent to tackle the problems?
62 . Regulation (EU) 2019/1150 on fairness and transparency for business users of online
intermediation services
P2B Regulation
EN
14
EN
EU-level legislation specifically targeted at commercial issues engaged in by online
to more than 10,000 platforms in Europe and reflects the fact that a certain dependency of
professionals,
the fairness, transparency and redress rights and obligations that the P2B Regulation provides
are necessarily high-level and principles-based. Since this legal framework establishes a
address issues deriving from the concentration of economic power and unfair business
practices of a limited number of very large, gatekeeper platforms, nor does it address
emerging and existing market failures in digital markets more broadly.
63 . EU data protection legislation82 specifies the fundamental right to the protection of personal
data. It therefore covers business-to-citizen and government-to-citizen interactions, rather than
commercial and competition related issues. Article 20 of the General Data protection
regulation
provides a limited right to data portability83, though it is broadly
considered that there are still many implementation challenges and that this right is at present
insufficient to significantly lower entry barriers and to facilitate the contestability of
markets84.
64 . EU consumer law does address a range of potentially harmful practices, at EU level notably
through the Unfair Commercial Practices Directive (UCPD)85 and the Unfair Contract Terms
Directive (UCTD)86. While these directives define a number of relevant concepts, such as
'professional diligence' and 'good faith', their scope is explicitly limited to business-to-
consumer transactions. Conversely, the Misleading and Comparative Advertising Directive
(MCAD)87 covers certain B2B relations. However, the provisions set forth in the MCAD are
limited to a narrow subset of advertising practices, which are not specific to online platforms
or digital markets, and do not deal with the unfair business practices carried out by large
gatekeeper platforms.
2.4. What is the size of the problem?
65 . Reports have estimated that the digital economy accounts for between 4.5% to 15.5% of
global GDP in 2019, depending on the definition.88 The amount of data created by year in the
digital economy is growing at an exponential rate. In 2020, it is estimated to reach 47
zettabytes compared to 12 zettabytes in 2015. Forecasts point to 2 142 zettabytes in 2035.89 A
report from the CMA describes the growing importance of pricing algorithms, not only
adopted by large retailers such as Amazon, but also smaller online retailers.90
66 . The important role of large platforms for businesses to reach markets and consumers is
constantly strengthened due to the e-commerce growth in Europe. The B2C e-commerce
turnover was growing at an average pace of 13% between 2014 and 2019 with turnover
forecasted to hit EUR 621 billion in 2019 and is set to be worth EUR 717 billion in 2020.91
On average, 16.2% of retail trade in 2020 in Europe takes place online, almost double in
comparison to 2018.92 The share of online shoppers in Europe making cross-border online
purchase has also increased significantly over the past decade, nearing 50% in 2019. Cross-
EN
15
EN
border B2C e-commerce sales in Europe are projected to grow at a double-digit rate at least
through 2022.93
lockdowns due to the COVID 19 outbreak in 2020;94 consumers have shifted their habits
towards more search engines, social media and online entertaining media.95
67 . As explained in more detail in Section 2.1, the characteristics of digital markets often favor
the emergence of strong concentration, which tends to be accompanied by rising mark-ups
and weaker competition. The trend of increasing industry concentration has been
documented for both digital and non-digital industries alike. For instance, in 2014 the mean
96 had 4 percentage point higher sales concentration
than in 2000.97
68 . As regards trends in mark-ups, empirical studies suggest that company mark-ups have
increased by 4% to 6% for the period 2001-14, on average across country,98 and that the result
is mainly driven by the top of the mark-up distribution in the digital sector.99 For the top 10%
of the firms in the sample, the growth in mark-ups over the period 2001-14 amounted to 20%,
while the remaining firms in the sample exhibit a flat trend, i.e. mark-ups stayed roughly the
same.100 To the extent that this observed trend of increasing market power of this top 10% of
firms is a sign of insufficient competitive constraints faced by these firms, increasing
competition in these markets could contribute to slowing down the growth trend in mark-ups,
decrease prices and increase choice, quality and innovation. For example, a recent study
shows that more concentrated industries also feature a more negative relation between
markups and investment and innovation.101
69 . As to the platform economy, according to the Online Platform Economy Observatory, traffic
share is the most revealing indicator of the economic significance of online platforms.102 The
50 top online platforms, representing an average of over 60% of traffic share across the
Member States, achieved worldwide revenues of almost USD 340 billion (EUR 276 billion)
in 2018, and employed almost 600 000 people.
70 . The degree to which businesses of all kinds have integrated into and may depend on the
platform economy is illustrated by the fact that in some cases more than 50% of goods sold on
a market place can be from third-party sellers. In relation to software, there are over 26.4
million software developers in the world who depend entirely on large platforms providing
the infrastructure and setting the rules for the distribution of their apps.
71 . The problems identified above are affecting a large and expanding number of merchants and
small businesses across Europe, which form a significant part of the EU economy. The
number of enterprises depending on online platforms varies depending on the sector, but can
be estimated to reach at least 1 million merchants in the EU (underestimate), combining
sectors such as online retail, hotels and restaurant businesses, app developers, etc. The
increasing importance of digital channels and the dependence, especially of small businesses,
on such channels, means that the implications of unfair business practices by gatekeeper
platforms could be far-reaching.
EN
16
EN
72 . The reduced contestability of digital markets in which gatekeeper platforms operate moreover
seems to result in suboptimal innovation levels, with notably implications for societal
welfare..103 Relevant data, although of a general nature, supports the view that many markets
are becoming more concentrated and display less competition. Profit margins are widening,
with a few firms reaping a significant share. Innovation levels are also sub-optimal 104
2.5. Drivers of problems stemming from conduct by large gatekeeper platforms
2.5.1. Identified factors and practices limiting platform market contestability
73 . Gatekeeper platforms benefit from data dependency, strong (in)direct network effects, variety
and volume of data, information asymmetry, lack of multi-homing and consumer lock-in,
which reinforce entry barriers. In addition, other factors, including unfair practices by
gatekeeper platforms contribute to the weakened contestability of platform markets where a
gatekeeper platform is present.
74 . Furthermore, considering the market features that gatekeeper platforms benefit from, business
users in a data driven economy face important barriers to entry, such as limited or no access to
vast amount of data as well as lack of any or meaningful interoperability to access such data
that may be collected by gatekeeper platforms.
75 . For example, the Impact Assessment study and input to the OPC point to practices that
prevent both consumers and business users to switch. In the digital sphere data, being able to
port both historical and real-time data are an important precondition for both multi-homing
and switching.105 Business users and consumers alike repeatedly raise the issue of not being
able to use any other platform or service because the incumbent refuses to provide an
enhanced and continuous real-time ability to port personal and non-personal data in
interoperable format. These practices affect contestability, and
possibilities to move to or rely on alternative platforms or services.
76 . As mentioned above, behavioral biases reinforce switching costs and keep users locked106
into the gatekeeper platform. Platform companies routinely design their services to optimise
such as A/B testing, or finely targeted personalisation of their service offering. Some usage
patterns design choices that
users into certain decisions.107 In addition to biases
towards default settings/options and high information costs, researchers recently identified
seven categories of behavioral patterns.108
77 . In addition to behavioral biases, certain practices by large gatekeeper platforms contribute to
consumer lock-in, including through strengthening these biases through such unfair
practices. For instance, gatekeeper platforms are often requiring a user to sign up/register with
an email service of the
platform when using another of its products. The US
House of Representatives Judiciary Committee also describes lock-in strategies including free
tier offerings for cloud services.109
EN
17
EN
78 . Gatekeepers often also limit their business users ability and incentives to switch. One
example of such behavior is imposing on business users anti-
, by which
they
on
the platform, even though these alternative offers may be cheaper or otherwise potentially
attractive alternatives.110 For instance, a big app store, which represents a very important
trading partner for many business users, may not allow its business users to advertise
alternative subscription options to consumers. Also, so-called wide Most Favoured Nation
Clauses (MFNs) i.e. clauses that oblige third party business users to apply the same retail
prices or offer the same conditions on other platforms essentially freeze competition among
platforms and do not allow competing platforms to have lower priced/better offers.
79 . Finally, gatekeeper platforms may limit the access to or the interoperability of certain of their
platform services/functionalities with the services offered by business users, reserving those
functionalities to their own services.
2.5.2. Identified unfair practices vis-a-vis business users
80 . Online platforms intermediate an increasing number of transactions and are increasingly the
main vehicle for businesses for market access. A few gatekeeper platforms have become the
entry points to numerous markets - they link millions of enterprises with billions of
consumers. Businesses are increasingly dependent on limited number of these gatekeeper
platforms, which together with the strong data-driven network effects results in an imbalance
in bargaining power between business users and gatekeeper platforms. They have thus
become an important trading partner for these businesses to reach millions of consumers in
the EEA and serve as market-gateway, on which SMEs depend to be visible online and to
reach consumers both on national and cross-border markets.
81 .
revenue via onl
-
commerce. According to the Observatory estimates, around half of enterprises derived more
than 25% of their revenues from online platforms. For almost 10% of companies, online
platform sales exceed 75% of all revenues; while according to Statista estimates, in 2017,
18% of company revenues across the EU-28 came from e-commerce, with the highest
proportion in Ireland (33%) and Czech Republic (31%).
82 .
nce on platforms is the use of platforms to publish
online advertising. Of SMEs in the EU that sell online, more than 8 in 10 rely on search
engines to promote their products and services as a mean of marketing their products or
services. In 2018, an average of 26.2% of enterprises across the EU paid to advertise online.
In northern European countries, such as Sweden and Denmark, this figure was over 44%.
83 .
models have enabled them to create ecosystems for which they set the rules by which other
economic players should act. If set in an unfair manner, these rules can be detrimental to
EN
18
EN
84 . The enforcement experience and input to the OPC shows that unfair practices can have
different forms and shape. They can be grouped into broader categories the most prominent
examples are unfair business practices related to self-preferencing, data and access conditions.
85 . The broad category of self-
or self-
refers to practices in which a
usually vertically integrated platform applies dissimilar conditions to business users in
equivalent situations in this case favouring its own services directly or indirectly in
situations when it has a dual role of providing core platform services to business users and at
the same time competing with them when providing ancillary services.
86 . Self-preferencing occurs in many situations in the online and offline world (e.g. in
supermarkets with own brands). Such behaviour may not be considered generally anti-
competitive under EU competition rules or as unfair in all business relationships. However,
certain forms of self-preferencing may amount to an unfair business practice. An important
concern here is fair balancing of interests, in this case those of the gatekeeper platforms
versus that of their business users.111 In particular, the special position of gatekeeper platforms
who are playing a dual role and may engage in favouring their own services may lead to the
exclusion of alternatives by business users who are largely dependent on these gatekeeper
platforms to reach consumers, reducing choice for them, potentially undermining the quality
of service and increasing prices.112 For example, a big app store, which markets a number of
its own popular apps and at the same time maintains the same marketplace, access to which is
very important for its competitors to reach consumers, for competitors is self-preferencing
through applying more favourable policies for its own products and selective drafting rules
favouring its own products.
87 . Feedback to the OPC shows that business users consider self-preferencing to be a very
common practice deployed by large, vertically integrated platforms. Responses by business
users suggest that search and ranking algorithms often give preference
services, but also that a platform often has an incentive to bias its recommendations toward
the content provider charging a lower royalty.
88 . Certain forms of data related practices by a gatekeeper platform can also be considered
unfair.
89 . An example of a data related practice that could be considered unfair, and has been raised by
many stakeholders in the context of the OPC, is the situation where a gatekeeper platform
restricts business users from accessing and using the data that they provide, receive from their
customers
related practices which could also be considered as harmful relate to unfair restriction to inter-
operability and/or interconnection.
90 . Feedback to the OPC shows that business users are confronted with the imposition by large
platforms of proprietary services and an authentication through the platform even when third
party services/products are used to create a direct link with customers to the detriment of
third-party providers. Respondents suggest that gatekeeper platforms exclude business users
EN
19
EN
from access to user data and attempt to remove the direct link between the client and third
party suppliers (so-called disintermediation).
91 . Other examples include (i) gatekeeper platforms that use certain data that they received from
business users for a particular use, for instance advertising services, for other, unrelated
purposes, or (ii) a gatekeeper operating a marketplace benefits from its dual role and ability to
evaluate product, sales and customer data generated from the sales of goods provided by third
party merchant business users on its marketplace.
92 . Finally, with regard to access conditions, gatekeeper platforms, thanks to their position, can
often impose unfair terms of access to business users, for instance in relation to price for the
services they offer.
2.5.3. Fragmented regulation and oversight
93 . Various different national rules in the EU are emerging in partial response to the problems
identified, including significant initiatives in Belgium, France, Germany, and the Netherlands.
In addition, there is no necessarily coordination among different national authorities setting
laws vis-à-vis platforms, leading to potentially heterogeneous responses across the EU.
Fragmentation already exists with regard to platform-specific regulation, as for example in the
cases of transparency obligations and MFN clauses.113
94 . Fragmentation of the EU legal landscape creates a degree of legal uncertainty and
heterogeneous compliance costs that are generally easier to absorb for incumbents, but
generate significant burden for potential entrants and smaller business users, thus limiting the
ease of scaling-up across the internal market. Disparate laws may also endanger the benefits
ultimately resulting in unequal impacts on EU consumers. Fragmented regulation and
oversight put at risk the effective functioning of the Digital Single Market.
2.6. Problem drivers for market failures in digital markets
95 . As outlined in Section 2.2.2, there are two categories of market failures that undermine
competition and contestability: emerging market failures (see Section 2.6.1) and existing
market failures (see Section 2.6.2). There are different scenarios that can give rise to emerging
and existing market failures which cannot be tackled or addressed in the most effective
manner under the existing EU competition rules. Examples illustrating the most important
cases of emerging or existing market failures are set out in Section 2.6.3.
2.6.1. Emerging market failures
96 . Emerging market failures arise where certain market characteristics as those set out in see
Section 2.1, in isolation or combined with conduct of the companies operating on the market
concerned, fundamentally change the competitive process, leading to sudden and radical
decreases in competition preventing markets from self-correcting. Such emerging market
failures cannot be tackled on the basis of the existing EU competition rules, because the
market features underlying the emergence of such market failures are not caused by the
EN
20
EN
companies operating on the market concerned, even though their distortive effects can be
exacerbated by the
conduct.
2.6.2. Existing market failures
97 . Existing market failures refer to markets that, notably due to their structure, are not
functioning well and delivering competitive outcomes in terms of low prices, better quality, as
well as more choice and innovation. As explained in paragraph 50 above, while markets
typically feature self-correcting mechanisms, there can be obstacles that prevent these
mechanisms from operating, leading to non-transitory losses of economic value.114
2.6.3. Examples of emerging and existing market failures
98 . This section discusses examples for the scenarios that give rise to emerging and existing
market failures. As illustrated in Figure 3 below, there are some scenarios, namely leveraging
strategies (either standard or repeated parallel), and gatekeeper (or bottleneck ) scenarios,
which are relevant for both emerging and existing market failures. Other scenarios only affect
one or the other type of market failure. For instance, anti-competitive monopolisation only
concerns emerging market failures, while tacit collusion (possibly facilitated by pricing
algorithms) and common shareholding are examples of existing market failures. Each
scenario will be discussed in more detail below.
99 . The scenarios identified below do not represent an exhaustive list of such situations, but are
meant to illustrate the most important cases of emerging or existing market failures, which
cannot be tackled under the existing EU competition rules or which cannot be addressed in the
most effective manner.
Figure 3: Examples of emerging and existing market failures
EN
21
EN
2.6.3.1 .
Anti-competitive monopolisation as example of emerging market
failures
100 . Anti-competitive monopolisation refers to scenarios where one market player may rapidly
acquire market shares due to its capacity to put competitors at a disadvantage in the market
unfairly, for instance, by imposing unfair business practices or by limiting access to key
inputs, such as data.115 These markets will not yet have generated a dominant firm, but show
clear signs of increasing market power in the hands of one firm. As the industry matures and
becomes more concentrated, certain business practices - which are unobjectionable when used
in the infant stages of the industry (such as exclusive contracts, fidelity rebates, tying) - may
allow a firm to drive its competitors out of the industry or relegate them to unprofitable
market niches. In combination with other factors (such as scale economies or network effects)
they may lead to anti-competitive monopolisation.116
101 . One implication of the presence of network effects is that even markets where initially
multiple competitors are active become prone to tipping : once a firm has obtained a certain
advantage over rivals in terms of market share, its position may become unassailable and the
market may gravitate towards a situation of dominance or (quasi)-monopoly. An example
would be the market for comparison shopping websites, where the Commission found that, as
Google was able to take
traffic away from competitors, allowing Google Shopping to grow from a very marginal
position to a very strong dominant position.117 While the Google Shopping case shows that
Art. 102 TFEU allows for competition policy to intervene and impose corrective measures in
such markets after tipping has occurred, such an ex-post intervention may not always be able
to fully restore the market conditions that existed before the market tipped, in particular in fast
moving markets where competitors may quickly leave markets and may not want to reenter
again. Both the NCAs with relevant experience118, as well as the respondents to the OPC119,
generally consider that tipping is common or to some extent common in digital markets.
According to these respondents, the important or very important competition concerns that
arise in tipping markets are the following: 1) there will not be sufficient competition on the
market in the long run; 2) customers will not have enough choice; 3) customers will face
insufficient innovation; 4) efficient or innovative market players will disappear and 5)
customers may face higher prices.
102 . Both the NCAs with relevant experience120, as well as the respondents to the OPC121, agree
that anti-competitive monopolisation is a market scenario qualifying as important or very
important emerging market failure. In terms of sectors or markets where this scenario might
arise, respondents mentioned notably digital sectors/markets (including app stores, online
advertising, search services, social networks and online marketplaces), transport, automotive,
pharmaceuticals and telecoms. Most respondents who expressed a view considered that this
scenario is common or at least somewhat common in digital markets, including because of the
particular features of such markets.122 Those respondents who signaled relevant knowledge or
experience with tipping markets identified the following as fields where such tipping has
occurred: social networks, search services, e-commerce platforms, online advertising, online
EN
22
EN
messaging, app stores, operating systems, regulated markets like telecoms, energy and rail, as
well as online food delivery and accommodation bookings.
103 . Respondents generally considered that important or very important market features of a
tipping market are the following: 1) direct network effects; 2) indirect network effects; 3)
users predominantly single-home (i.e. they use typically one platform only) and 4) economies
of scale. Respondents generally considered that tipping is common or to some extent common
in digital markets.
2.6.3.2 .
Leveraging strategies as example of emerging and existing market
failures
104 . Leveraging strategies refer to practices whereby a firm with market power exerts such power
with a view to extending such power to a related market. Such strategies may give rise to both
emerging and existing market failures. Where a (possibly dominant) company with market
power in a core market applies such leveraging on multiple related markets simultaneously,
for instance by relying on large amounts of data, the term repeated parallel leveraging
strategies applies. The latter are more likely to give rise to existing market failures, rather
than emerging market failures.
105 . Repeated parallel leveraging strategies are a market scenario that could fall under Article 102
TFEU, if the leveraging is carried out by a market player with a dominant position. However,
if that player does not have a dominant position, Article 102 TFEU does not apply. Provided
there is a finding of dominance, the Commission could address single instances of this
conduct on the basis of the existing EU competition rules. However, if the market player
concerned engages in multiple leveraging strategies into different related markets, this would
require running several competition cases for each individual market onto which the
leveraging takes place. This may lead to diverging outcomes depending on the circumstances
of each case, without addressing the underlying problem at the heart of the leveraging
In such circumstances, a [market]
investigation into these practices in the digital industry may allow to achieve a uniform
approach that avoids the vagaries of the case-law, and might hence be superior to the attempt
of setting a policy through the precedential values of article 102 cases. 123
106 . Both the NCAs with relevant experience124, as well as the respondents to the OPC125, broadly
[a] (not necessarily dominant) company with market power in a core market
extend[ing] that market power to related markets
important or very important structural competition problem. In terms of sectors or markets
where these situations manifest themselves, respondents mentioned digital sectors/markets
(including app stores, operating systems, online advertising, search services, social networks,
messaging apps, e-commerce and online marketplaces), transport, automotive, and telecoms.
As examples and manifestations of the leveraging of market power, respondents pointed to
-prefer
126
-
subsidisation.
EN
23
EN
2.6.3.3 .
Gatekeeper/bottleneck scenarios as example of emerging and
existing market failures
107 . As pointed out in the OPC127 by many stakeholders, as well as recently documented by the US
House of Representatives Majority Staff report (see Section 1), so-
control access to a number of customers (and/or to a given input /service such as data) that
at least in the medium term cannot be reached (and/or replicated) otherwise. Gatekeeper
scenarios refer to situations where customers typically predominantly use one service
provider/platform for a specific purpose (single home). These gatekeepers may engage in
strategies that limit the ability of their business users to grow and multi-home, thereby
weakening competition on the whole market.
108 . There was consensus both among the NCAs with relevant experience128, as well as among the
respondents to the OPC129
Gatekeeper Scenarios
structural competition problem, pointing in particular to e-commerce, search services, social
networks, online advertising, app stores, operating systems, online messaging,
accommodation bookings, entertainment services (e.g. video distribution and broadcasting),
financial markets, energy and telecommunication markets, or transport, as sectors where such
gatekeeper scenarios have manifested themselves.
2.6.3.4 .
Tacit collusion as example of existing market failures
109 . Tacit collusion refers to a situation where competitors in an industry manage to behave in a
coordinated manner (i.e. jointly raise prices, or partition the market among each other) based
on a m
their actions. Given that in a tacit collusion scenario undertakings do not reach an agreement
nor engage in a concerted practice, this type of scenario falls outside Article 101 TFEU and
cannot be tackled by the Commission under the existing EU competition rules.
110 . NCAs130 and stakeholders responding to the OPC131
[h]ighly
concentrated markets where only one or few players are present, which allows to align their
market behaviour
competition problem. They consider that important or very important features of an
oligopolistic market with a high or substantial risk of tacit collusion are the following: 1) high
concentration levels; 2) competitors can monitor each other's behaviour; 3) high barriers to
enter and 4) the homogeneity of products. Those respondents also generally agreed that
Article 101 TFEU is not suitable for tackling tacit collusion.132
111 . More recently, there have been concerns expressed in particular in academia133 that new
technologies, in particular pricing algorithms, may facilitate collusive behaviour among
firms without the need to coordinate their actions explicitly.134 Pricing algorithms are
automated tools that allow very frequent changes to prices and other terms taking into account
all or most competing offers on the market. It appears at least possible that such algorithms, if
deployed by several competitors in the same industry, will learn to set supra-competitive
prices without ever communicating with each other.135 NCAs with relevant experience submit
that pricing algorithms are widely used for the distribution of mass-market products (such as
EN
24
EN
electronic goods, or household appliances) and services (such as plane tickets and
accommodation). They are used especially to set flexible prices for products or services in
several industries including tourism, transport, hospitality and entertainment, and can be an
important or very important source or part of the reasons for a structural competition
problem.136
112 . For instance, in its E-commerce sector inquiry, the Commission established that more than
half of the respondent retailers tracked the online prices of competitors and more than three
quarters of these retailers subsequently adjusted their own prices to those of their
competitors.137 Such automatic price adjustments are not per se problematic. However,
transparency in the
industry that would be difficult to achieve if this data had to be hand-collected. Price
transparency has been recognised as a factor facilitating collusion in past competition cases138,
and in at least one case, namely that of Computerised Reservation Systems, necessitated
regulatory measures to prevent the emergence of tacit collusion.139 The potential of algorithms
to facilitate collusion by creating more transparency could be amplified where such
algorithms are either programmed to, or learn by themselves to pursue collusive strategies, i.e.
to set prices above the competitive level, which are sustained by a complex punishment code
by which firms trying to undercut the collusive price to steal market share from their rivals
trigger a price war in the market which makes the original price reduction unprofitable.
113 .
ompetition law covers instances where algorithms amplify explicit
collusion, but could be more difficult to apply in relation to pure forms of tacit collusion,
which is generally not covered by antitrust rules. Given the concern that algorithms make
tacit collusion more frequent, there is an ongoing debate about the need to rethink some
fundamental antitrust concepts. 140 A similar concern has been expressed by an expert panel
Economic Advisory Group on Competition Policy
Commission on the UK experience with Market Investigations and possible lessons to draw
for the EU.141
114 . Among the stakeholders responding to the OPC, only a part had come across pricing
algorithms before, but those who did consider that they are an important or very important
source or part of the reasons for a structural competition problem.142 NCAs do not have
extensive experience with such market situations. Some of them therefore conducted studies
in order to explore different theories of harm when using pricing algorithms. They explain
that pricing algorithms are widely used for the distribution of mass-market products and
services especially to set flexible prices for products or services in several industries including
tourism, transport, hospitality and entertainment and generally consider that using pricing
algorithms can lead to the alignment of prices/less competition between market players.
2.6.3.5 .
Common shareholding as example of existing market failures
115 . Some markets are characterised by a high degree of common shareholdings (when one person
or institution owns shares of two or more firms) or cross-shareholdings (when two firms own
143 This may lead to ambiguous incentives for the owners of these shares,
EN
25
EN
and the managers who run the respective companies on behalf of their shareholders. In fact,
(or all) of its rivals, then this may soften competition between these firms, leading to higher
prices and lower quality for consumers. Common shareholder may induce such a softening of
competition for instance by making managers of companies in the same industry compete less
aggressively against each other by designing their compensation schemes in a way that
discourages such competition.
116 . While common shareholding was not explicitly mentioned in the questionnaire submitted to
the NCAs and the OPC, one NCA144 and several respondents to the OPC145 suggested
common shareholding as an additional market feature that can be a source or part of the
reasons for a market failure, notably as it may facilitate tacit collusion.
117 . Due to its structural nature, common shareholding itself is, like tacit collusion, not covered by
Article 101 and Article 102 TFEU. It cannot therefore be addressed under the existing EU
competition rules absent additional elements amounting to an anti-competitive agreement
pursuant to Article 101 TFEU or a breach of a collective dominant position pursuant to
Article 102 TFEU.
2.7. How will the problem evolve?
118 . As explained in Section 2.1.6, concentration and mark-ups in most digital markets have been
increasing over the last years, and there is no indication that this trend will be inverted during
the next years. In some cases, markets have already stabilised at a high concentration level
structure and do not show any evidence of possible increase in competitiveness in the future.
Data is also becoming more and more important, exacerbating the market failures associated
to the control of data.
119 . The COVID-19 crises has dramatically increased the importance of e-commerce and trading
.146 This has only accelerated the dependency of
users and businesses on the services provided by the larger gatekeeper platforms as
evidenced indirectly in the increase in stock market valuation of some of the largest platform
companies.
120 . Absent any EU intervention, the economic drivers are likely to increase, exacerbating the
observed problems. As an illustration, further development and use of voice assistants can
-activated services may
create concerns in relation to search for online products/services/information. The provision
of a single answer to a search request limits the possibility to access alternative results, thus
reducing choice and limiting competition.147
121 . Innovation would remain concentrated within a small number of gatekeepers, ultimately
ation- and data-friendly services148 provided by
a larger number of platforms than gatekeepers.
EN
26
EN
122 . The regulatory gap will be filled at national and non-EU level, increasing further the
coordination costs, fragmenting the internal market, and leading to greater legal uncertainty.
2.8. Problem tree
Figure : Problem tree
Market features
Problem drivers
Problems
Scale (or scope) economies, and
Identified factors and practices
high degree of vertical integration
limiting platform market
Weak contestability
contestability
of platform markets
Network effects
Fragmented regulation and
oversight
Unfair business
conditions for
Data dependency
Identified unfair practices vis-
a-vis business users
business users
Switching costs
Emerging market failures, e.g.:
anti-competitive
monopolisation
Market failures that
Assymetric information and
undermine
consumers' behavioural biases
competition and
contestability in
Existing market failures, e.g.
digital markets
tacit collusion common
High market concentration
ownership
3. WHY SHOULD THE EU ACT?
3.1. Legal basis
123 . Given the intrinsic cross-border nature of the services provided by gatekeeper platforms and
the risk of further regulatory fragmentation regarding functioning of the Single Market for
digital services, in particular in relation to gatekeeper platforms as well as functioning of
digital markets, Article 114 TFEU is the relevant legal basis for this initiative. One of the
objectives of the intervention is to identify harmonised set of rules relating to measures
tackling unfair behaviour by gatekeeper platforms, safeguarding contestability of digital
markets and ensure harmonised set of market investigation powers in case of existing or
emerging market failures.
124 . It is appropriate to combine Article 114 TFEU with Article 103 TFEU because certain unfair
behaviour by gatekeeper platforms and certain types of existing market failures could also be
addressed under Articles 101 or 102 TFEU but not in the most effective manner, as explained
in Section 2.3.1. Given that Article 114 TFEU is subject to stricter procedural requirements
than Article 103 TFEU, the former will have to be followed (see Article 294 TFEU).
125 . As set out above, the current regulatory approaches at Member States level are a patchwork of
existing or proposed regulatory solutions (see detailed description in Annex 5). This creates
EN
27
EN
legal uncertainty for companies operating in the internal market, whether at national or on a
pan-European basis. Moreover, the fact that emerging or existing market failures can be
addressed in some Member States but not in others and that the remedies imposed may differ
across Member States risks creating an appreciable distortion of competition in the internal
market and undermine fundamental freedoms protected by the Treaty. This distortion could be
prevented in the most effective manner by empowering the Commission and the Member
existing market failures in a harmonised system of parallel competences, similar to the
application of Articles 101 and 102 TFEU.149
3.2. Subsidiarity: necessity and added value of EU action
126 . The objectives of the intervention cannot be achieved by Member States acting alone, as the
problems are of a cross-border nature, and not limited to single Member States or to a subset
of Member States. The digital markets at stake (including those featuring gatekeeper
platforms) are often of a cross-border nature, as is evidenced by the volume of cross-border
trade, and the still untapped potential for future growth, as illustrated by the pattern and
volume of cross-border trade intermediated by digital platforms. Almost 24% of total online
trade in Europe is cross-border. It is estimated that by 2025 online marketplaces will represent
65% of cross-border online sales in Europe.150
127 . Even where these digital markets may be geographically defined as national in scope, the
problems at stake nevertheless remain of a cross-EU nature for three main reasons. First, the
goods and services offered by the market players concerned are typically of a cross-border
nature. Second, digital players typically operate across several Member States, if not on an
EU/EEA-wide basis, which is particularly the case for markets such as online advertising,
social media, online retail, cloud services, e-commerce or online search. This is not to say that
languages - however, the overall business strategy will normally be EU/EEA-wide.
128 . Accordingly, emerging and existing market failures in digital markets and unfair business
practices by large digital gatekeepers both have Union relevance, as they can arise across
borders and affect several Member States, thus not being limited to a specific national market
of a Member State.151
129 . As regards unfair business practices, in the absence of an EU measure, there is a high risk that
with national approaches, business users or application developers seeking to serve the
internal market will need to understand a range of diverse rule-sets and pursue actions in
multiple countries across the EU, which is likely to fragment the Single Market for digital
services, create barriers to expansion and compliance costs, especially for start-ups and SMEs.
A lack of harmonised rules in this space risks complicating the regulatory landscape faced by
platforms operating on a pan-European or indeed global basis. An intervention at the EU level
is therefore more efficient, insofar as it introduces a common set of rules across Member
States to address in a consistent manner the same unfair business practices carried out by large
digital gatekeepers across the Union.
EN
28
EN
130 . Similarly, intervention by individual Member States or NCAs would be ineffective in tackling
emerging and existing market failures across the Union. Each Member State can only address
structural competition problems in its own territory,152, imposing is own remedies, whereas
emerging and existing market failures may affect the territory of several Member States
because of the wider geographic scope of the relevant market concerned or the cross-border
business activities of the market players concerned. Addressing emerging and existing market
failures with a cross-border dimension at national level could also lead to inconsistencies in
the remedies imposed, with the ensuing risk of fragmenting the Digital Single Market.
131 . Therefore, by addressing emerging and existing market failures in digital markets and unfair
business practices by digital gatekeeper platforms at the EU level, the functioning of the
internal market will be improved through clear and harmonised rules that give all stakeholders
legal clarity and through an EU-wide intervention framework allowing to address emerging
and existing market failures in a timely and effective manner.
4. OBJECTIVES: WHAT SHOULD BE ACHIEVED?
4.1. General objectives
132 . The general objective of this initiative is to ensure the proper functioning of the internal
market by promoting effective competition in digital markets, in particular a fair and
contestable online platform environment. This objective feeds into the strategic course set out
in the Communication Shaping Europe's digital future
1.
4.2. Overview of specific objectives
133 . The specific objectives are outlined in the figure below, and explained in the following
sections.
Figure 5: Specific objectives
Ensure the proper functioning of the internal market
by promoting effective competition in digital markets, in particular a fair and
contestable online platform environment
Ensure
Ensure
Enhance
contestable and
contestable and
Address
coherence and
competitive
competitive
identified unfair
Promote
legal certainty in
markets by
markets by
conducts by
contestability of
the online
preventing the
addressing
large gatekeeper
platform
platform
emergence of
existing market
platforms
markets
environment to
market failures
failures on a
preserve the
on a case-by-
case-by-case
internal market
case basis in
basis in digital
digital markets
markets
EN
29
EN
4.2.1. Objectives related to problems stemming from the gatekeeper position of
large platforms
4.2.1.1 .
Address identified unfair conducts by large gatekeeper platforms
134 . The market dynamics surrounding gatekeeper platforms provide them with a position in
which their behaviour vis-à-vis business users remains unchecked, resulting in unfair
conducts. Therefore, the specific policy objective is to lay out a clearly-defined set of rules
addressing identified unfair behavior by digital gatekeeper platforms. This objective is
measurable in principle through careful monitoring and information-gathering provisions. It is
both achievable and realistic, as industry pressure and Member State action, as well as some
voluntary measures have already led to some pockets of change. A timeframe and indicators
for reaching this objective should be devised following a dedicated monitoring approach
based on the work of the Expert Group for the Observatory on the Online Platform Economy
and its supporting study.
4.2.1.2 .
Promote contestability of platform markets
135 . Weak inter-platform competition is one of the problems identified in Section 2.2.1.1. This
specific policy objective is therefore to promote platform markets contestability and benefits
associated with it, i.e. innovation, increased consumer choice.
136 . While gatekeeper platforms themselves significantly contribute to innovation (as witnessed by
their large R&D budgets, for example), the innovation objective here refers to innovative
entry in areas that challenge the incumbent to innovate even further in view of market entry
that brings innovation.153 Closely related to innovation, the consumer aspect of this objective
aims for more privacy-friendly and innovative services.
4.2.1.3 .
Enhance coherence and legal certainty in the online platform
environment to preserve the internal market
137 . There is an emerging fragmentation of the regulatory landscape and oversight in the Union, as
Member States address platform related problems currently not (effectively) covered by
existing regulation. This is suboptimal in light of the cross border nature of the platform
economy and the systemic importance of gatekeeper platforms for the internal market.
Divergent fragmentation would create legal uncertainty and higher regulatory burdens for
participants in the platform economy. Therefore a specific policy objective is to improve
coherent and effective oversight and enforcement of measures against platforms.
4.2.2. Objectives to prevent and address market failures in digital markets
4.2.2.1 .
Ensure contestable and competitive markets by preventing the
emergence of market failures in digital markets on a case-by-case basis
138 . As explained in Section 2.1 and Section 2.6.1, digital markets may display certain
characteristics which, in isolation or in combination with conduct of the companies operating
on the market concerned, fundamentally change the competitive process, leading to sudden
EN
30
EN
and radical decreases in competition preventing markets from self-correcting. Such emerging
market failures cannot be tackled on the basis of the existing EU competition rules. Therefore,
the specific policy objective is to allow identifying and addressing such emerging market
failures in digital markets on a case-by-case basis in a timely and effective manner to ensure
that these markets remain contestable and competitive. This will contribute to digital markets
delivering low prices, better quality, as well as more choice and innovation to the benefit of
European consumers.
4.2.2.2 .
Ensure contestable and competitive markets by addressing existing
market failures in digital markets on a case-by-case basis
139 . As explained in Section 2.6.2, certain markets may not be functioning well and delivering
competitive outcomes due to their structure. Such existing market failures cannot be tackled
or addressed in the most effective manner on the basis of the existing EU competition rules.
Therefore, the specific policy objective is to allow identifying and addressing such existing
market failures in digital markets on a case-by-case basis in a timely and effective manner to
ensure that these markets remain contestable and competitive. This will contribute to digital
markets delivering low prices, better quality, as well as more choice and innovation to the
benefit of European consumers.
4.3. How do the objectives link to the problems identified?
140 . The figure below shows how different objectives are linked with the problems and the
underlying problem drivers identified in Section 2. It also shows that all five specific
objectives contribute to achieving the general objective of ensuring the proper functioning of
the internal market by promoting effective competition in digital markets and in particular a
fair and contestable online platform environment.
EN
31
EN
Figure 6: Problem drivers, problems and objectives
Problem
Specific
General
drivers
Problems
objectives
objective
Identified factors and
Promote contestability, leading in
practices limiting platform
Weak
particular to innovation and
market contestability
contestability
consumer choice
of platform
markets
Enhance coherence and legal
Ensure the
Fragmented regulation and
certainty in the online platform
proper
oversight
environment in the internal
functioning
of the
Unfair
market
internal
business
market by
Identified unfair practices
conditions for
Address identified unfair
promoting
vis-a-vis business users
business users
conducts by large gatekeeper
platforms
effective
competition
in digital
Ensure contestable and
markets, in
Emerging market failures
competitive markets by
particular a
(e.g. anti-competitive
Market
preventing the emergence of
fair and
monopolisation)
failures that
market failures in digital markets
undermine
contestable
on a case-by-case basis
competition
online
and
platform
Ensure contestable and
environment
Existing market failures contestability
competitive markets by
(e.g. tacit collusion,
in digital
addressing existing market
common ownership)
markets
failures in digital markets on a
case-by-case basis
5. WHAT ARE THE AVAILABLE POLICY OPTIONS ?
5.1. High-level options
5.1.1. High-level option 1: No change to the current regulatory framwork
141 . Under this policy option the current regulatory framework would not change.
142 . This means that the Commission continues vigorous competition enforcement in digital
markets under the existing EU competition law framework, and the existing EU rules
(outlined in Section 2.3) continue to apply. Competition enforcement by the Commission
includes making full use of the existing tools within this framework, such as sector inquiries,
interim measures and adopting more prescriptive or restorative remedies compared to past
cases, when appropriate. In this scenario, the ongoing reviews of existing legislation
(horizontal and vertical agreements), as well as the evaluation of the Market Definition Notice
will continue as planned.154 However, as explained in Section 2.3.1, the existing EU
competition rules cannot tackle or address the problem drivers for market failures in digital
markets in the most effective manner.
143 . The other applicable EU rules include the P2B regulation, the GDPR and EU consumer law.
As described in Section 2.3.2, the P2B regulation does not address issues deriving from the
concentration of economic power and unfair business practices of a limited number of very
large gatekeeper platforms, nor does it address emerging and existing market failures in
digital markets more broadly. The GDPR covers business-to-citizen and government-to-
EN
32
EN
citizen interactions, rather than commercial and competition related issues. The provisions of
EU consumer law are not specific to online platforms or digital markets, and do not deal with
the unfair business practices of gatekeeper platforms nor with emerging and existing market
failures in digital markets.
144 . It follows that this option cannot tackle or address in the most effective manner the problem
divers identified and thus cannot achieve the related specific objectives.
5.1.2. High-level option 2: Separate legislative instruments
145 . Under this policy option, the Commission would establish two separate legislative
instruments to address the problems outlined in Section 2.2, which are routed in the inherent
features of digital markets, as described in Section 2.1, namely (i) the weak contestability and
identified unfair business practices of gatekeeper platforms, and (ii) emerging and existing
market failures in digital markets.
146 . Under this policy option, one legislative instrument would focus on large gatekeeper
platforms. It would impose on them a clearly defined set of prohibitions and obligations with
two aims: 1) to enhance the contestability of the markets on which these large gatekeepers
operate (through provisions such as mandating data-portability); and 2) to ban identified
unfair business practices (such as self-preferential display, ranking and rating) of these
gatekeepers.
147 . A separate legislative instrument would enact a competition-based market investigation
regime for digital markets that complements the existing competition law toolbox by
preventing and addressing market failures through suitable and proportionate market-specific
remedies with a view to ensuring contestable and competitive markets. This would allow the
Commission to identify and remedy market failures that cannot be tackled or addressed in the
most effective manner under the existing EU competition rules.
148 . While this policy option could tackle the two sets of problems identified, it would not do so in
the most effective manner. In fact, one of the repeated requests of stakeholders during the two
parallel public consultations was to ensure a coherent EU-level intervention where ex ante
rules for gatekeeper platforms and competition-based market investigation powers would
complement each other.155
149 . This option could not fully guarantee this coherence between the two legal instruments and
could not effectively ensure the synergies and complementarity between them. In particular,
this option would not allow establishing an effective feedback mechanism between the two
instruments which would ensure establishing a coherent enforcement practice. Enacting two
instruments could even lead to inconsistencies in how to achieve the most effective outcome
to ensure the proper functioning of the internal market.
5.1.3. High-level option 3: A single legislative instrument
150 . Under this policy option, one single legal act would combine the measures addressing the
issues identified in Sections 2.2.1.1 and 2.2.1.2 with regard to gatekeeper platforms (as
EN
33
EN
described in Section 5.2) and the establishment of a competition-based market investigation
regime (as described in Section 5.3) aimed at preventing and addressing market failures in
digital markets that are not or not fully addressed by the aforementioned measures.
151 . This option would overcome the inefficiencies identified in the high-level option 2 and would
ensure a holistic approach with two complementary and mutually reinforcing pillars.
152 . Pillar I of the single legislative act would consist of two parts. The first part would contain
prohibitions and obligations aimed at protecting dependent business users whose services are
intermediated by large gatekeeper platforms against already identified and clearly defined
unfair practices and mitigate commercial imbalances. These prohibitions and obligations
would provide an immediate and cross-market solution to tackle the identified unfair
practices.
153 . The second part would contain prohibitions and obligations aimed at enhancing the
contestability of markets characterised by large gatekeeper platforms through measures
focusing on promoting consumer switching, as well as mitigating network effects in order to
promote choice and innovation. These general obligations are targeting certain already
identifiable key elements that affect the contestability of markets on which large gatekeepers
operate and will reduce the risk of emerging market failures, while mitigating the existing
ones. The measures promoting contestability would reduce the risk of market failures and
mitigate the existing ones in markets characterised by large gatekeeper platforms.
154 . Pillar II of the single legislative act would establish a competition-based market investigation
regime with appropriate remedy powers for digital markets, which, in combination with
continued vigorous enforcement under the existing EU competition rules, will allow
addressing emerging and existing market failures in the dynamically changing digital
environment beyond the already identified issues in markets characterised by large gatekeeper
platforms. It will also ensure keeping the gatekeeper rules future-proof through an appropriate
feedback mechanism by which the competition-based market investigation regime would
contribute to enhancing Pillar I. Inversely, experience in the implementation of the specific
pre-identified obligations for gatekeepers could point to areas meriting a competition-based
market investigation.
EN
34
EN
Figure 7: Complementarity between Pillars
155 . In conclusion, high-level option 3 ensures the most efficient and effective way to achieve the
objective.
5.2. Policy options under Pillar I
5.2.1. Link to related initiatives
156 . In parallel to this initiative, the Commission is also proposing rules on platform responsibility
under the DSA, when it comes to the issue of illegal third-party content, goods, or services
intermediated by platforms. Those issues relate to the supply of digital services across borders
in the internal market and are not directly related to notions of bottleneck power or unfair
business practices, nor to contestability objectives under the present initiative. While the
preferred option under that initiative
5.2.2. Description of the retained policy options
5.2.2.1 .
Option I.A: Ex ante rules on self-executing
blacklist
whitelist
by gatekeeper platforms
157 . This option would provide a new ex ante regulatory framework, which would apply to the
core platform services of gatekeeper platforms. Based on the characteristics of these
services as well as of the concentrated nature of the markets in which they intervene, the core
platform services in scope would be: (i) online intermediation services (incl. esp. market
places, app stores and social networks and their advertising services), (ii) online search
engines (and their advertising services), (iii) operating systems and (iv) cloud services.156
EN
35
EN
158 . The new framework would not replace, but rather complement the horizontally applicable
provisions of the P2B Regulation, which would continue to apply to all online intermediation
services and online search engines, not merely those with a gatekeeper role.
159 . To enable an accurate identification of the gatekeeper platforms and provide legal certainty
for market operators, the new framework would lay down criteria based on which a large
online platform provider and its services would be designated as a gatekeeper platform by the
competent EU regulatory body. Such criteria would reflect the problem drivers. First, the
market position and bargaining power held by certain core platform services would be
captured by size and dependency criteria. Second, the weak inter-platform competition that
results from such core platform services having become entrenched would be captured by a
durability criterion. These criteria would translate into mainly quantitative and some
qualitative parameters to be applied under two thresholds.
160 . A primary threshold would set a high bar of exclusively quantitative parameters above
which companies and their services would automatically be designated gatekeeper platforms.
The use of quantitative parameters only will provide legal certainty and expedite the
designation process. These parameters could include EEA turnover (at the group-level of the
companies controlling the platform), number of monthly active EU users (separately for each
of the relevant platform services in scope), number of EU business users using the platform
service, as well as additional parameters reflecting the durability criterion (e.g. average time
spent by consumers and EEA turnover derived from digital advertising, stability and growth
of the global and EU user bases). A platform meeting the criteria under the first threshold
would automatically be considered as gatekeeper and the designation by the regulatory body
as gatekeeper would be declaratory, given that the parameters are quantitative.
161 . A secondary threshold would allow the competent regulatory body to designate additional,
large online platforms and their services that fall short of the very high bar set by the primary
threshold but that nonetheless operate large core platform services that exhibit significant
dependencies, which are moreover entrenched. This designation would rely on some
qualitative parameters (e.g. data advantage; vertical integration; level of innovative entry) in
addition to quantitative ones. Importantly, online platforms falling below the secondary
threshold would be automatically excluded from the scope of the new regulatory framework.
This automatic exclusion will further add to providing legal certainty to market operators.
This system would be complemented by a requirement for large online platforms to notify a
limited set of data corresponding to the scoping parameters.
162 . Once designated as gatekeeper platforms, platforms (and their services in scope) would have
to comply with clearly defined self-executing prohibitions
blacklist ) and positive
obligations whitelist ) relating to identified unfair practices as further explained below.
These rules would apply irrespective of the technology used to provide the services (e.g.,
desktop, mobile or voice assistant technology). These prohibitions and obligations would
apply to unfair business practices identified through law enforcement experience, feedback to
the OPC, input from the Impact Assessment study and targeted submissions by the
stakeholders.
EN
36
EN
163 . In relation to such self-executing prohibitions and obligations, the task of the competent
regulatory body would be to ensure compliance with the rules without any substantive
assessment of the practice in question.
164 . These self-executing prohibitions and obligations under the new framework would address the
factors and practices concerning market contestability (Section 2.51) and would address
identified unfair practices towards business users (Section 2.5.2).
165 . As regards the objective of facilitating contestability, possible rules could include prohibitions
on gatekeeper platforms from (i) restricting business users to offer the same goods and
services to consumers under different conditions through other platforms than through the
gatekeeper own platform or services; (ii) restricting business users from promoting on the
rom concluding contracts with
services anti-steering 157 (iii) restricting the installation of third party applications or
application stores through channels
service side-loading
access, sign up or register to the platform; or (v) automatically signing users in to more than
one of
-in option for each of these individual
services.158 The aforementioned practices restrict the ability of their business users to use
orm, to offer
products or services to consumers or. In addition, some of these practices are unfair as they
require users to accept services they do not really want or prevent them from refusing
conditions of access to services that are not necessary for the provision of these services.
166 . As regards the objective of tackling identified unfair practices, a list of such self-executing
prohibitions and obligations addressed to gatekeeper platforms would include different types
of unfair practices, such as unfair data related practices, unfair self-preferencing practices and
Section 2.5.2).
167 . As regards unfair data related practices, the new rules would for example ban gatekeeper
platforms from (i) using non-publicly available, aggregated or non-aggregated data generated,
inferred from or collected through the commercial activities of business users of the
gatekeeper platform for
own consumer-facing commercial activities; or (ii)
using data received from business users through the provision of advertising services for any
other purpose than these advertising service. Another rule could be (iii) an obligation on a
gatekeeper platform to provide users access to their own data that is generated by using the
platform and allow them to transfer that data. These practices are considered particularly
unfair, since they allow gatekeeper platforms to unfairly benefit, vis-à-vis dependent business
users, from data they obtain due to their dual role, and to undermine the level playing field
and ability of business users to operate on the markets concerned.
168 . As regards unfair self-preferencing practices, the new rules would for example ban gatekeeper
platforms from applying preferential display/ranking in online search engines or online
EN
37
EN
intermediation services. These practices by gatekeeper platforms are considered particularly
unfair in an imbalanced commercial relationship, since they unfairly favour gatekeeper
engage with their consumers in view of more targeted and commercially more attractive
offers.
169 . The objective of these self-executing prohibitions and obligations is to address unfair
practices and enhance contestability, by ensuring a more level playing field between
gatekeeper platforms and protecting the weaker business partner in commercial relationships
with a gatekeeper platform.
170 . For all self-executing prohibitions and obligations covered by the new framework, gatekeeper
platforms would be allowed to provide limited justifications, on general, non-economic
considerations (e.g. public morality, public policy or public security). Such justifications
would not cover, or seek to replicate, economic justifications for a specific type of conduct as
known and applied under EU competition rules.
171 . Finally, this option would involve implementation, supervision and enforcement at the EU
level by the Commission as the competent regulatory body. Given the pan-European reach of
the targeted companies, a decentralised enforcement model does not seem to be a conceivable
option. However, to integrate the national expertise in the platform economy, the initiative
would envisage that the Commission consults a network of
before taking
decisions (e.g. on non-compliance; fines).
172 . To facilitate implementation, supervision and enforcement by the Commission as the
competent regulatory body, the new rules will also envisage that gatekeeper platforms would
provide to the Commission any information that is necessary to ensure compliance with the
rules and enable it to continuously monitor market developments. Furthermore, the rules
for the implementation of the
new rules and obligations.
5.2.2.2 .
Option I.B: Ex ante rules on prohibitions and obligations
concerning clearly defined unfair practices by gatekeeper platforms
greylist
173 . This option would build on the previous option I.A of laying down clearly defined self-
executing prohibitions and obligations applicable to designated gatekeeper platforms.
174 . Beyond the aforementioned self-executing prohibitions and obligations, this option would in
addition define a closed list of obligations and prohibitions for a further set of identified unfair
practices, which would however require an intervention by the competent regulatory body in
their concrete application ( greylist
s and obligations under
the new framework would equally address the factors and practices that limit market
contestability (Section 2.5.1) and would address identified unfair practices towards business
users (Section 2.5.2). As regards the former, a greylist would include obligations concerning
EN
38
EN
conditions of data access, portability and interoperability. As regards the latter, the grey list
would aim at ensuring fair and non-discriminatory
and platforms.
175 . In relation to the objective of facilitating contestability, the new rules would include data-
related obligations on gatekeeper platforms to (i) take appropriate and reasonable measures to
enable the business users and consumers of their services to use any other service by means of
providing continuous portability of personal and non-personal data and appropriate forms of
interoperability159 or (ii) interoperate with or allow access to third-party ancillary services at
the request of its business users or third party service providers where the gatekeeper provides
ancillary services to its online intermediation services (e.g. data analysis services, payment).
These obligations are considered particularly important to enable users to use equally valid
such practices, users would be significantly limited in their ability to multi-home and/or use
alternative platforms or services. This could consequently lead to the unfair exploitation of
users and undermine the contestability of the platform markets concerned.
176 . As regards the objective of tackling unfair practices by gatekeeper platforms, the new rules
would include obligations to ensure conditions of fair and non-discriminatory access to
160 Such obligations would include a prohibition
of, for example, unfair or discriminatory pricing, or other conditions, for the access to
ir since they would allow gatekeeper
platforms to extract excessive value from their commercial relationship with dependent users,
which would not be possible in a more balanced commercial relationship.
177 . The required intervention by the competent regulator for applying the grey list would take two
different forms:
178 . First, some of the clearly defined obligations would still require an implementation-step,
whereby the regulatory body would explain how one or more of the obligations apply to the
specific gatekeeper platform. For example, it would be for the competent regulatory body to
define appropriate and reasonable technical and contractual measures that the gatekeeper
platform would need to take to enable the business users and consumers of their platform or
services to use any other platform or service (i.e. to multi-home). Another example is an
obligation on gatekeeper platforms to provide non-discriminatory interconnection and
interoperability between their service and other related services. In this case, it would again be
for the competent regulatory body to define appropriate and reasonable technical and
contractual measures to ensure that interoperation.
179 . Second, some of the clearly defined obligations contain an explicit requirement to act
fairly/not to act unfairly and therefore will require a fairness assessment in their application to
the specific case. For example, it would be for the competent regulatory authority to assess
ricing behaviour could
be considered unfair and determine how such unfairness should be addressed. In another
EN
39
EN
example, it would be for the competent regulatory body to assess and establish in a concrete
case whether gatekeeper platforms give access to the platform on fair terms. The fairness
assessment pursues different objectives from competition law and as such will not be carried
out pursuant to the EU competition rules. If such unfairness would be confirmed in the
specific case, it would be for the competent regulatory body to determine measures for
addressing it.
180 . Similarly to option I.A, also under this option it would be the Commission who would be
responsible for the implementation, supervision and enforcement of the new framework.
Equally, to integrate national expertise in the platform economy, this option would also
envisage that the Commission consults a network of
before taking decisions (e.g.
non-compliance; tailoring of obligations; fines).
5.2.3. Policy options discarded at an earlier stage.
181 . As indicated in the IIA, an option of amending the P2B Regulation was considered. Further
horizontal rules could be established for all 10 000 online intermediation services and search
engines that are currently falling within the scope of the P2B Regulation. This could cover
prescriptive rules on different specific practices that are currently addressed by transparency
obligations and beyond.
182 . However, such an option would target all platforms. It is therefore in mismatch to the
problems and its drivers, which relate to the subset of very large gatekeepers. Imposing
stringent measures horizontally would be disproportionate and have a negative impact on
innovation and competition in the online platform space. Changing the scope of P2B
Regulation so that only large gatekeepers are subject to regulation is not a conceivable way
forward since it would eliminate the beneficial impact of its fairness and transparency rules
addressed to non-gatekeeper platforms.
183 . Furthermore, at the IIA stage, an option was considered which would empower a regulatory
body to collect information from large online platforms acting as gatekeepers, supported by
enforcement powers in case of refusal to supply this information. The purpose would be to
better inform the implementation of the existing legal framework by gaining, for example,
further insights into gatekeepers
users and consumers, scope of data gathering, treatment of their own downstream operations
compared with those of third parties and indicators of the outcomes resulting from these
practices. However, this option would only improve the understanding of the issues at stake. It
may be expected to have some reputational lever, but that would be insufficient to address the
problems or affect their drivers, which are already well identified and circumscribed.
5.3. Policy options under Pillar II
184 . In order to prevent emerging and address existing market failures in digital markets (described
in Section 2.6), the Commission has considered two main policy options to frame an
appropriate competition-based market investigation regime. The two market investigation
regimes considered have a fundamentally different underlying rationale as described below.
EN
40
EN
5.3.1. Option II.A: A conduct-focused dominance-based market investigation
regime
185 . This option would allow the Commission to address competition concerns arising from
unilateral conduct by dominant companies without any prior finding of an infringement
pursuant to Article 102 TFEU.161 This regime would require the Commission to establish
dominance of a company in a relevant market before assessing whether there is an emerging
or existing market failure on a market characterised by the presence of a dominant company.
Following a thorough market investigation, and subject to a clear legal test to identify an
emerging or existing market failure, it would allow the Commission to impose appropriate
and proportionate remedies addressing the conduct of the dominant company without any
finding of an infringement of the existing EU competition rules. It follows that it would not
involve any fines or rights to launch damage claims. The Commission could also recommend
legislative action to improve the functioning of the market concerned.
5.3.2. Option II.B: A market structure-based market investigation regime
186. This option would allow the Commission to identify and remedy market failures that cannot
be tackled or addressed in the most effective manner under the existing EU competition rules.
Thus, unlike Option II.A, this regime would not focus on any specific company, but rather on
any market suffering from emerging or existing market failures. This option would therefore
consist in establishing a competition-based market investigation regime similar to already
existing and tested market investigation instruments of this kind in other jurisdictions such as
the United Kingdom, Iceland, Greece, South Africa and Mexico. This market investigation
regime would be based on a clear and predictable legal test allowing the Commission to
intervene with appropriate and proportionate remedies after having established on the basis of
a thorough market investigation that emerging or existing market failures prevent a given
market from functioning properly, irrespective of the conduct of the undertakings operating
on that market.
187 . Under this option, the Commission would be competent not only to remedy the conduct of
specific companies, but also to impose appropriate and proportionate market-wide remedies,
after having mapped all features of a market in order to establish whether there is an emerging
or existing market failure on the market or the markets concerned (where the relevant features
exist in more than one relevant market).
188 . As under the previous option, there would be no finding of an infringement, no fines and no
damage claims. The Commission could also recommend legislative action to improve the
functioning of the market concerned.
5.3.3. Sub-options II.C and II.D Scope of the market investigation regime
189 . Irrespective of whether the market investigation regime is conduct-focused and dominance-
based or rather market-structure based, the question of its scope can result in two further sub-
options.
EN
41
EN
a) Sub-option II.C: Market investigation regime with a digital scope
190 . Under this sub-option, the market investigation regime would allow preventing emerging and
addressing existing market failures in digital markets. Digital markets would be identified on
the basis of relevant criteria, such as the fact that products or services are offered over the
internet, including products connected to the internet.
b) Sub-option II.D: Market investigation regime with a horizontal scope
191 . Under this sub-option, similarly to the existing EU competition rules, the market investigation
regime would apply beyond digital markets, and allow preventing emerging and addressing
existing market failures in all markets of the economy. This option would therefore result in a
market investigation regime with horizontal scope.
6. WHAT ARE THE IMPACTS OF THE POLICY OPTIONS?
6.1. What are the impacts of the policy options under Pillar I?
192 . This section presents the main impacts of the policy options as compared to the baseline with
the aim of identifying proportionate measures.
193 . The categories of stakeholders which would be affected, directly or indirectly, by the retained
policy options are: platforms (in scope and out of scope), business users depending on
platforms (e.g. hotels), business users competing with platforms (e.g. another platform),
competitors (e.g. innovative entrants), consumers, regulatory authorities. Impacts for these
stakeholder categories have been assessed as completely as possible in the following sub-
sections covering the internal market, growth, innovation, competition, platforms, business
users, SMEs, consumers and regulatory authorities.
194 . To set the background, among respondents who replied to the relevant question in the OPC,
91% agree that there is a need to consider dedicated regulatory rules to address negative
societal and economic effects of gatekeeper power of gatekeeper platforms.162 This view if
supported by many targeted submissions by different groups of stakeholders, such as SME
platforms and their associations, telecom operators and their associations as well by national
regulatory authorities in different sectors (e.g. electronic communication services).
195 . For the impacts developed in this section see also Annex 3, which specifies in detail who will
be affected by the preferred option and how.
6.1.1. Internal market
196 . Preventing fragmentation of the internal market is one of the most important policy objectives
enshrined in the Treaties of the EU and preserving the cross-border nature of the platform
economy contributes to this objective. A 2016 European Commission Communication on the
opportunities and challenges of online platforms for the Digital Single Market stresses the
pivotal role of online platforms in the European single market. Services and products such as
search engines, price comparison websites, online marketplaces and creative content outlets
EN
42
EN
offer strong links to the rest of the economy. One million EU companies in Europe use online
platforms to trade goods and services, and more than 50% of these are SMEs. The report
estimates a turnover of about EUR 602 billion for the European B2C e-commerce sector,
growing at almost 14% per year.163
197 . Christensen et al (2018) estimate, using the RHOMOLO model164, that implementing the third
pillar of the Investment Plan for Europe, including efficiency gains from the Digital Single
Market, would contribute to a 1.5% increase in GDP per year until 2030 and create between
1.0 and 1.4 million jobs.165 In particular, the impact of a more efficient Digital Single Market
ranges from 0.44 to 0.82% changes in GDP and between 307 and 561 additional FTEs.
198 . Option I.A would already allow some aligning of platform-related rules across the EU
through horizontal measures. Option I.A leaves scope however, for a quite a number of unfair
business practices to remain unaddressed or left to treatment at national level, which raises a
risk of fragmentation. Option I.B would capture a complete scope of identified unfair
practices by gatekeeper platforms that would be treated in a harmonised way across the EU.
Due to enforcement at EU level foreseen under both options, also the implementation and
enforcement would be coherent across all Member States, thus fostering the Single Market for
digital services.
6.1.2. Growth and trade
199 . The platform economy contributes heavily to the European economy as revealed by its size
and is expected to continue to grow steadily. Traffic share is one of the most important
proxies of the sector. The top 50 online platforms represent 60% of the traffic share in Europe
reaching revenues for about EUR 276 billion in 2018 and employing almost 600 000 people.
200 . The European market of online platforms makes a significant contribution to GDP and the
European economy as a whole. Revenues of the sector in Germany for instance reached EUR
33 billion in 2015, compared to EUR 320 billion worldwide. Similarly, the top 13 vloggers166
of one of the largest platforms earned together about USD 54 million which is assumed to end
up as consumption or investment in the national GDP.167 This shows the relative size of the
sector in the economy168.
201 . As explained in the problem definition (Section 2
position is difficult to contest
due to the features of digital markets as well as due to unfair business practices by gatekeeper
platforms. In terms of the total potential scale of the macroeconomic impacts associated with
policy options to reduce entry barriers, as mentioned in Section 6.2.1, one can observe that the
B2C e-commerce turnover was growing at an average pace of 13% between 2014 and 2019
with turnover forecasted to hit EUR 621 billion in 2019 and is set to be worth EUR 717
billion in 2020. This sector is expected to increase in value by around 14% per year.169
202 .
chilling effect on sales.
Business users argue that such unfair practices (e.g. pretended privacy considerations,
limitation to data access, etc.) would lead to up to 15% loss in their sales. Both options and B
in particular would reinforce trust in the platform business environment since it foresees an
EN
43
EN
adaptable framework, based both on a clear set of obligations/prohibitions and a flexible
greylist which might require an assessment of the applicability of the conducts to the specific
case.
203 . Importantly, competitive entrants contribute to growth in the digital sphere; prohibitions
and/or obligations under Option I.B in particular e.g. data access or interoperability - can
allow entrants to grow and compete effectively. Given that consumer trust in gatekeepers is
essential, no prohibition and/or obligation under Options A or B should lower standards of
security or privacy. Option I.A rules (clearly defined self-executing obligations and
prohibitions) for gatekeepers would improve business environment but would not address in
depth issues related to some clearly defined unfair business practices which might require
competent regulatory body to tailor implementation of a clearly defined obligation or
establish by means of a fairness assessment how a given prohibition and/or obligation should
be tailored to a specific gatekeeper platform. Option I.B would allow for continuous
monitoring and gradual/timely redress of unfair business practices and weak market
contestability, thus contributing to a more competitive platform ecosystem.
204 . Building on a clear set of self-executing prohibitions and obligations and tackling clearly
defined unfair practices through prohibitions and/or obligations that would require tailoring of
these measures to a given gatekeeper platform, Option I.B would provide an adequate solution
to the problems identified in the platform economy and would free its growth potential.
205 . As already explained in the recent P2B impact assessment, trade intermediated through online
platforms is expected to follow an upward trend as most consumers opt for platforms when
purchasing goods and services online170. In addition, the COVID-19 crisis accelerated the
shift to online retail at an unprecedented pace pointing to the importance of the online
platform economy. Usage of digital devices has increased significantly during the COVID
pandemic, which is likely to increase the relative importance of online platforms compared
with the off-line world. Specifically, following the lockdown, one global survey found that
consumers spent more time on social media and mobile applications (cited by 47% and 36%
respectively).171
206 . Cross-border ecommerce in Europe was worth EUR 143 billion in 2019 (without travel). And
59% of this market, EUR 84 billion, is generated by online marketplaces. Marketplaces with
European capital represent 11% of the market, an increase of 17% compared to one year
before.172
6.1.3. SMEs
207 . The above figures show the growth potential for EU SMEs to reach consumers through
platforms across the globe. SMEs in the context of the initiative under assessment could be
both platforms and business users. Options A and B would have a positive impact on business
users and SME platforms. Business users would benefit from a more balanced relationship
with gatekeepers while SME platforms will have enhanced opportunities to scale up and
compete with these gatekeepers. Options A and B foreseeing a comprehensive form of
regulatory oversight would allow both categories of SMEs to benefit from a more innovative
EN
44
EN
and competitive business environment incentivising them to seize the digital single market
opportunities and grow.
6.1.4. Competition and Innovation
208 . There are strong links between patterns of competition and innovation. Preventing patents or
pre-emptive activities, for instance, is one way to gain monopoly power and to increase
barriers to entry. If this pattern is dominant, the pace of innovation in the long-run slows
down.173 Although the sector of online platform invests heavily in innovation, smaller
companies that depend on gatekeepers are discouraged to innovate so as not to compete with
the gatekeeper. Hence, innovation that contributes to such dependency is incentivised.174
209 . The evidence shows the concentration of R&D investment among few dominant firms, and
with a sustained trend. The trends in the investment in R&D depicted in our Impact
Assessment study suggest a cluster of high volumes of investment among big five companies;
and a widening gap across time between large and small companies. The Impact Assessment
study shows that financial resources that could be invested in R&D are diverted to mergers
and acquisitions, which results in higher market concentration instead of improvements in the
quality and quantity of products and services for consumers. T the pattern of innovation
dedicated to competing 'for the market' has a detrimental effect on consumer choice and
surplus.175
210 . Moreover, market concentration results in accumulation of cash-flow that is available for
R&D investment and innovation or mergers & acquisitions. The Impact Assessment study
illustrates the concentration of liquidity among the top five companies, each of them ranging
between 10% up to 30%, while the remaining 17 companies are on average below 1%. i Five
companies accumulate 90% of total free cash-flow (USD 90 612 million) that could be
distributed among all 22 companies if competition were lower. This suggest that smaller
companies may face some financial constraints, failing to attract venture capital to finance
R&D projects, while large firms have enough own funds to embark on innovation.
211 . Option I.A is expected to have a positive impact on innovation since it would create a fairer
but given
innovation capacity.
212 . Option I.B
, thus also creating a healthier business
environment for other platforms contributing to restoring and/or installing competitive
dynamics in the platform economy. Alternative platforms are currently facing a number of
challenges e.g. for developing compelling offers (lack of data and consumers (due to strong
network effects)), for accessing venture capital for competing services, portability, risk of
leverage, etc. Also, business users (e.g. e-commerce merchants, service providers and
application developers) face issues such as dependency, unfair contractual relations, unequal
distribution of revenues/profits, exclusion. In light of this, the expectations for Option I.B are
EN
45
EN
to spur overall technological innovation in the platform markets (concentrated so far within a
limited number of gatekeepers) to other market players, thus creating more competition and
innovation to the ultimate benefit of consumers. Option I.B is in this respect estimated to yield
direct benefits of many billions of Euros annually, in addition to improved innovation levels
and entrepreneurship, which are complex to quantify in precise terms but likely equally if not
more important in size and impact.
6.1.5. Competitiveness
213 . As stressed in the EU Better Regulation Guidelines, the three components of competitiveness
are cost/price, innovation and international competitiveness.176 The competitiveness of the
online platform market is enhanced if the sector can produce good quality or original
products/services, is undistorted and competitive, and is capable of innovation. This is pivotal
when considering introducing remedies (policy options) that do not affect the competitiveness
of the online platforms and incentivise competitive behaviours.
214 . Carayannis et al (2014) shows that innovation and productivity are important drivers for
competitiveness. It was shown above that dynamic markets with the right incentives to
innovate foster total factor productivity and growth.177 A more efficient Digital Single Market
with the right incentives to innovate should contribute to a more competitive EU digital
economy. Yet, a 2019 BEUC report stresses that regulatory policy is pivotal to ensure
regulatory costs and administrative burden are kept to a minimum while minimising data
harvesting and handling, ensuring rights to privacy, encouraging fair behaviour between the
most powerful platforms and ensuring consumer protection.178
215 . Option I.A would address a range of clearly identified harmful practices. Option I.B sets a
clear regulatory framework imposing requirements on gatekeepers in relation to further
clearly identified harmful practices and is able to assess specific unfair conducts and further
target remedies (if needed) to the unfair conduct identified. Both therefore contributing to the
overall competitiveness of enterprises depending on platforms
6.1.6. Platforms
216 . First and foremost, it should be stressed that there is a wide ranging consensus across various
operators with different sizes and business models in the tech community that there is a need
for rules addressing the
179
217 . Second, the targeted scope of both options imposing rules only on the largest gatekeeper
platforms strongly contributes to the proportionality of any potentially resulting compliance
costs.
218 . Compliance costs under both options would largely substitute for the already high costs large
platforms incur for complying with divergent regulatory measures gradually put in place in
different Member States. It could be objectively argued that compliance costs under all
options considered would be reasonable. Such costs would imply some additional legal
compliance officers to check company policies against the new rules; some employees to
EN
46
EN
interface with the regulator and respond to requests for information. Given that compliance
costs are expected to be limited, platforms under scope would have little incentive to pass on
costs to consumers or to small business users (e.g. by limiting their access to a platform).
219 . However, indirect (other than non-compliance) costs may be higher, as proposed measures are
expected to have impact on gatekeeper platforms business models and potentially reduce their
supra-normal profits. The impact of such changes on gatekeeper platforms is difficult to
quantify. While some loss of revenue for gatekeeper platforms is expected, there are no
indications that this would result in significantly higher fees and/or reduced quality for
and
platforms need, due to the market features discussed in Section 2.1, an important number of
consumers in order to be able to (i) attract businesses (and vice versa) thus allowing online
matching of offer and demand, and (ii) benefit from the virtual growth cycle characterising
the platform economy.
220 . The Options would moreover not be geared towards eliminating legitimate monetisation
opportunities for gatekeeper platforms. They would aim to eliminate unfair behavior towards
business users, thus rather enhancing trust in the platform business model. A set of measures
that contribute to a more dynamic online platform ecosystem and more contestable market
will particularly benefit SME online platforms who would face lower barriers when entering
the market. It can therefore be expected that an increased contestability of the markets will,
even with some changes to their business model due to the regulatory intervention, continue
to incentivise gatekeeper platforms to bring innovative products to the market and compete
for consumers and business users.
221 . Moreover, the Impact Assessment study for this initiative estimates that gatekeeper platforms´
financial resources that could be invested in R&D are currently diverted to mergers and
acquisitions (M&A), which results in higher market concentration instead of improvements in
the quality and quantity of products and services for consumers. This pattern of innovation
dedicated to competing 'for the market' has a detrimental effect on consumer choice and
surplus. In addition, the positive impact on innovation stemming from higher market
contestability is not limited only to diversion of money from M&A to R&D. Other expected
indirect effects include an increase in entrepreneurship and creation of new products and
solutions meeting consumers' needs rather than focused on exploiting a gatekeeping position.
This may have a multiplicative effect increasing the size of the European single market, and
hence, GDP and online cross-border trade. The Options are estimated to allow to recover to a
large extent this opportunity cost. The present options would thus have a clearly positive
effect on overall welfare.
6.1.7. Business users
222 . Both third party business users (non-competing with the gatekeeper) and business users
competing with the gatekeeper (e.g. a retailer selling on an e-commerce platform) are
expected to significantly benefit from both options, as these are geared at preventing unfair
practices that go to
EN
47
EN
while Option I.B would address a larger number of such practices. Given that the rules only
aim to prevent unfair and harmful conduct, they should not hamper entry even by gatekeepers
based on fair means of competition.
6.1.8. Employment
223 . The impact of a more contestable online platform market on employment directly in the sector
is ambiguous. On the one hand, if the ecosystem becomes more dynamic and grows further in
size, in theory it could absorb additional FTEs. However, if companies decide to use the
additional profits on R&D or labour-saving production processes, then the effect is
negative.180 Hence, the direct effect on employment depends on how many new firms enter
the market and challenge the gatekeepers, their ability to create new jobs, and the
corresponding behaviour across the dominant platforms.
224 . However, there is an indirect positive effect on other sectors of the economy, considering the
links between the digital economy and growth. Above all, even under the assumption that no
additional jobs would be created, given the millions of people employed in the sector and the
millions of SMEs depending on online platforms to reach their consumers, taking adequate
measures to ensure the proper functioning of the platform economy would safeguard these
millions of jobs.
6.1.9. Consumers
225 . This impact assessment supports the wide consensu
in terms of choice, convenience and price of online products and services. However, due to
the distorted intra-platform competition and the limited inter-platform competition, consumers
have a reduced choice in terms of number of competitive platforms, informed choice
decisions181, and data/privacy-friendly services.182
226 . In addition, increased market concentration is detrimental for the consumer surplus as it
results mainly in lower choice and higher prices/costs.
227 . Although data to estimate the loss in consumer surplus is limited, there is some illustrative
evidence. For example, if commission fee in large app store were halved from 30% to 15%,
the average prices of apps in this app store could fall, which would increase consumer surplus
up to EUR 490 million in the EU per year based on Statista data.183
228 . Moreover, if it is assumed that current expenditures on advertising per users are excessive and
driven by high market concentration, such amounts could be a proxy for the consumer
detriment. For example, in 2019 EUR 546 was spent in the UK on advertising per user184.
Given the total number of users in the platform economy is about 3.3 billion185, the consumer
surplus would reach EUR 1,803 billion, assuming the number of consumers does not change
with the level of competition in the market.186
229 . The choices for consumers are limited by lock-in effects and lack of innovative alternatives
that are restricted by the gatekeeper platform(s) unfair business practices. In the longer run,
they risk experiencing lower quality and/or less innovative services and/or higher prices. The
EN
48
EN
initiative assessed under the present analysis aims at addressing these concerns with a view to
230 . Options I.A and I.B - Both options would indirectly contribute to safeguard value added for
consumers and would contribute to ensuring
interests187. This would be achieved by contributing
platform (intra-platform competition) and among platforms (inter-platform competition), (b)
stronger contestability of the markets where gatekeepers are present and (c) better functioning
of the internal market through enhanced regulatory oversight at EU level.
231 . The question has arisen whether additional regulatory costs for gatekeeper platforms would
not translate into higher prices for consumers. As explained above under 6.1.6., consumers are
at the core of platfo
Hence, gatekeepers would not risk changing their successful business model and losing
consumers by setting prices for currently free services. Where g
-
priced,
services would be perceived by consumers differently as compared to increases in already
existing (i.e. not zero) prices. Any attempt from gatekeepers to make users pay would imply
the risk for them of reducing the attractiveness of their services and of encouraging users to
switch to other platforms continuing to offer their services free of charge.
6.1.10. Regulatory Authorities
232 . Both Options imply enforcement costs to be essentially incurred by the EU Commission. It
may also imply costs for national regulatory authorities if they were to be consulted on the
Commission enforcement decisions, but these costs would be limited due to the purely
consultative, non-operational nature.
233 . Option I.B would imply additional resources and costs for the Commission for acquiring
information/data, monitoring, analysis and decision-taking under the greylist. They are
estimated in the order of one Commission unit. However it is considered that these costs will
largely outbalance the benefits of reducing the impact of practices which severely undermine
incontestable positions.
6.1.11. International trade
234 . The vast majority of gatekeeper platforms are active globally and have their corporate
headquarters outside the EU. Both Options can therefore have trade implications. However,
such implications would be, if anything, political in nature, as both Options are consistent
235 . The Options are designed in such a way as to target any gatekeeper platform in an objective
and non-discriminatory manner, therefore potentially also platforms headquartered in the EU,
as long as they fulfil the objective criteria (see Section 5.2.2.1) for being designated a
gatekeeper platform, as defined in the law. Importantly, the global gatekeeper platforms have
EN
49
EN
an important EU presence, including legal representation, large numbers of personnel and
physical infrastructure including cloud, fulfilment or retail outlets. The objective scoping
criteria applicable to both options target this EU presence and do not take into account the
location of the corporate headquarters of the company in question. In doing so, the options
will be future-
-
discrimination under the General Agreement on Tariffs and Trade. The regulator will also be
required to keep the designation of gatekeeper platforms up to date by identifying new players
or by removing existing ones, including any that may have their corporate headquarters inside
the EU.
236 . It is also important to add that in the US, where many gatekeeper platforms have their
headquarters, an intense debate is on-going about the need to regulate gatekeeper platforms188,
including antitrust hearings of Amazon, Apple, Facebook, and Google in the US Congress
House of Representatives.189
the Judiciary issued a Majority Staff Report in which a broad range of significant remedies are
proposed, following a detailed assessment of the effects of number of unfair and
anticompetitive practices by these platforms, in order to restore competition in digital
markets.190 These remedies notably include structural separation, line of business restrictions
as well as non-discrimination rules for dominant platforms including on access and pricing.
237 . Similar calls for ex ante regulation of platforms are also voiced in other non-EU countries,
such as Japan or Australia.191 These developments point to the global consensus that has been
building on the need to complement competition policy with additional ex ante measures,
limiting the risk of negative impacts of the new EU regime on international trade. What is
more, the present initiative would establish a proportionate regulatory framework that should
promote a fair and contestable online platform environment in the EU, one in which new
platforms can emerge and scale-up, to the benefit of users around the globe, not just in the
EU.
6.2. What are the impacts of the policy opitions under Pillar II?
238 . This section summarises the main impacts of Options II.A and II.B described in Sections
5.3.1 and 5.3.2 as compared to the baseline scenario. Given that the impacts described below
are valid for both a conduct-focused dominance-based market investigation regime (Option
II.A) and a market structure-based the market investigation regime (Option II.B), the
following sections refer simply to the impact of the market investigation regime. However,
the magnitude of the impact of each option differs as explained in Section 6.2.10.
239 . The assessment of the impact of these options is to a large extent qualitative as a
quantification of the effects is only partially feasible. In fact, the assessment of the options
depends on the number of cases investigated and the markets concerned. Given that these
elements are very difficult to predict any quantification of the impacts would at any rate be
subject to a large error margin. Nevertheless, whenever feasible some quantification is
provided.
EN
50
EN
6.2.1. Internal market and competition
240 . The adoption of a market investigation regime would complement the existing competition
rules, thus leading to a more comprehensive competition-based enforcement toolkit that is
able to address emerging and existing market failures in digital markets in a timely and
effective manner. This would result in more open and competitive markets, where companies
compete more fairly on their merits, and enable them to generate wealth and create jobs.
241 . Several studies illustrate the importance of an efficient enforcement of competition law for
ensuring competition on markets.192 Hylton and Deng (2007) examined how different antitrust
systems affect the level of competition in individual countries and found that increasing the
range of instruments available to competition authorities has a significant impact on the
intensity of competition in the country's economy.193 One of the key objectives of the market
investigation regime is to remove the features that give rise to barriers to entry and expansion
for businesses. Boosting effective competition enforcement would mean that the internal
market would be reinforced and be fairer for businesses and consumers.
242 . A study requested by the IMCO committee of the European Parliament estimated the size of
the EU electronic communications and services sector in 2019 at EUR 86.1 billion, data and
AI at EUR 51.6 billion and e-commerce sector at EUR 14.6 billion.194 According to this
study, interventions aiming at increasing the contestability of the digital sector, resulting in
lower prices and greater consumer choice, productivity gains and innovation, would have a
significant positive and growing contribution to achieve all of the potential benefits of a
Digital Single Market.
6.2.2. Growth and productivity
243 . The adoption of a market investigation regime would lead to a more comprehensive
competition-based enforcement toolkit. More effective competition enforcement would
ultimately increase productivity, which is a key driver for economic growth.
244 . Several empirical studies confirm that the enforcement of competition law leads to more
competition on markets, which in turn results in higher productivity in affected industries,
which translates into economic growth195, and that industries where there is a high level of
competition experience statistically significant faster productivity and economic growth196.
Other studies also confirm the positive effects of competition on the productive efficiency of
companies due to (i) between-firms effect, by which better companies succeed while the
worst ones fail and leave the market, and (ii) a within-firm effect by which companies in
competitive environments are better managed.197
245 . It is difficult to estimate the impact on economic growth and productivity of the adoption of
the market investigation regime since the proposed changes are of a nature that is not easily
quantifiable. This is because a more comprehensive competition-based enforcement toolkit is
likely to give rise to general benefits to the economy as a whole rather than to specific and
quantifiable savings or benefits. In addition, the economic literature trying to measure those
benefits is scarce.
EN
51
EN
246 . Despite these obstacles, Buccirossi et al. (2011, 2013) developed a methodology to measure
the impact that competition policy enforcement has on the economy.198 These authors
estimated the impact of the level of competition enforcement, measured by the Competition
Policy Indicators ( CPIs )199, on growth in Total Factor Productivity ( TFP )200. TFP growth
has had an important impact on GDP in the EU. Comprehensive competition policy
enforcement is expected to influence TFP growth because it helps keeping markets open,
thereby ensuring that new, innovative and more productive firms are not foreclosed from
markets, and at the same time putting pressure on incumbents to either improve or lose market
share. Given that TFP growth in the EU as a whole has been below 1% for the last ten years,
even if one assumes that the introduction of additional powers to the Commission results in a
relatively small increase in the effectiveness of competition policy enforcement, that impact
on TFP would be signifcant, translating into a significant boost to productivity. For instance, a
10% increase in CPI would lead to an increase of more than 40% in TFP growth.201
247 . Dutz and Hayri (1999) found a strong correlation between long-run GDP growth, and
effective enforcement of competition rules, on the basis of a cross-section of 52 countries.202
Moreover, the positive effects on productivity growth are not only felt in the sectors where
such increased competition takes place, but they also spread to downstream markets and
throughout the economy.203 A JRC study from 2017 find that a 1% reduction in the fixed cost
of entry in the overall market for product and services would result in an increase in GDP of
0.1%. Such effects are expected to increase further in the long-run fostered by the positive
effect on TFP and employment.204
248 . The impact of the market investigation regime on economic growth and productivity is
expected to be particularly relevant when addressing emerging and existing market failures in
digital markets where, as explained in Section 2, market features can lead or contribute to
preventing healthy competition between market players. This is even more the case today
given the increasing weight of the digital economy (see Sections 2.4 and 2.7).
6.2.3. Innovation
249 . The adoption of a market investigation regime would lead to a more comprehensive
competition-based enforcement toolkit allowing businesses to compete more fairly on their
merits. This incentivises them to innovate and offer a better range of higher quality products
and services that meet consumers' expectations. Firms facing more competition from rivals
innovate more than monopolies. Greater competition also drives efficiency in processes,
technology and service.
250 . According to Federico et al. (2019), a significant amount of innovation is driven by disruptive
firms. By making its offer to customers attractive in a new way, a disruptive firm can destroy
a great deal of incumbent profit while creating a large amount of consumer surplus.
Competition enforcement precisely seeks to protect the competitive process by which
disruptive firms challenge the status quo.205 Several empirical studies confirm that an increase
in competition leads to a significant increase in research and development investment by
neck-and-neck firms.206 Conversely, the view according to which market concentration or
EN
52
EN
large firm size is associated with a higher level of innovation is not supported by empirical
evidence.207 Shapiro (2012) highlights the considerable empirical evidence that greater
competition spurs innovation.208
251 . As explained in an UNCTAD report (2020),209 because digital markets show a particularly
high speed of evolution, the ability to innovate, including through the adoption and innovative
use of technologies, is likely to become a particularly important capability for firms and
economies in the digital era.210 Adressing the market features impeding a more competitive
environment in the digital markets is thus of particular relevance.
6.2.4. International trade
252 . The adoption of a market investigation regime would lead to a more comprehensive
competition-based enforcement toolkit and make EU markets more open, competitive and
attractive to investors. Greater competition also enhances the ability of businesses to compete,
both on their home markets and internationally. Borrell and Tolosa (2008) assessed the
combined effect of competition and other policies, particularly open trade policies, concluding
that competition law and policies aimed at opening trade reinforce each other and should be
considered as complementary.211
253 . The promotion of a higher competitiveness of digital markets is of particular importance in
increasing trade and investment flows. According to an UNCTAD report,212 digitalisation
contributes significantly to increasing the scale, scope and speed of trade. ICT products are
already a significant part of the global trade (in 2017 they are estimated to have reached USD
530 billion, representing 10% of total global trade in services).
254 . Similarly to Pillar I, the market investigation regime, and in particular Option II.B which
focus on market structure and not on particular businesses, would be designed in such a way
that it would target any company active in a given digital market in an objective and non-
discriminatory manner. In doing so, the market investigation regime would be future-proof
the non-discrimination
principle under the General Agreement on Tariffs and Trade.
6.2.5. Employment and inequality
255 . The adoption of a market investigation regime would lead to a more comprehensive
competition-based enforcement toolkit and would lead to greater competition on EU markets.
An overview by the OECD of the main literature covering the links between competition and
employment confirms that competition stimulates employment growth in the long term.213
The aggregate effect mainly results from a positive impact on TFP growth, which increases
labour demand, and through aggregate demand, given that more competition lowers prices
and therefore tends to increase real wages. This generates a virtuous circle of output and
demand growth in the long run.214 Greater competition by stimulating economic growth, has
also a beneficial effect on income equality.
EN
53
EN
256 . In the short run, the response to increased competition can lead to an increase in
unemployment, e.g. through process innovation that replaces labour intensive machinery with
new machines to increase productivity at the cost of labour. However, econometric
simulations of the effect of increased competition leading to redundancies in an industry
demonstrate a return to a steady growth path with rising employment after 2 to 3 years.215
Benetatou et al (2020) found that CPI has a positive and statistically significant effect on
labour productivity growth.216
257 . In relation to income, Ennis et al. (2017) attempted to calculate a rough measure of the short-
run money transfer from poor to rich due to market power for 12 OECD economies and
showed that, on average, for each dollar of monopoly profits, a total of USD 0.37 is
transferred from the 90% poorest to the 10% richest.217 Ennis et al (2017) suggest that, in an
average country, market power would increase by 17% the wealth of the best-off 10%.218 The
results were similar to the results found by Ennis and Kim (2017), suggesting that 10% to
24% of the wealth of the best-off decile may be attributed to market power.219
258 . One important aspect to consider as regards the impact of the market investigation regime on
employment is the growing importance of digital markets in job creation. According to an
UNCTAD report,220 digital transformation has strongly contributed to job creation across the
G20. Between 2006 and 2016, total employment in the G20 grew by 13%, a net gain of
almost 127 million jobs with highly digital-intensive sectors contributing with 43% of these
net job gains. Jobs in the ICT sector comprised 11.8% of total employment of the G20
countries, in 2017.
259 . The Covid-19 crisis called for the adoption of new labour regulations favouring teleworking
regimes. Digital services are of extreme importance as tools enabling teleworking regimes.
Therefore, making these services more accessible is even more important today for a
functional labour market.
6.2.6. Consumers
260 . The adoption of a market investigation regime would lead to a more comprehensive
competition-based enforcement toolkit and thus protect EU consumers from business
practices that keep the prices of goods and services artificially high. This in turn would ensure
that consumers have access to better quality, wider choice and innovative goods and services
at affordable prices. Numerous studies confirm the benefits of competitive markets for
consumers.221
261 . The CMA estimates the average direct financial benefit to UK consumers of its whole activity
(which includes antitrust, merger control, consumer protection enforcement, as well as market
studies and market investigations) and associated costs.222 For the period 2017 to 2019, the
total estimated annual consumer benefit of its activities was GBP 1109 million against an
average cost of GBP 76 million, yielding a ratio of direct benefits to costs of 14.6.223 A
significant share of these consumer benefits relate to the
(i.e.
GBP 820 million).
EN
54
EN
262 . During the last 10 years of enforcement by the UK competition authority, the annual
consumer benefits resulting from its market investigations ranged between GDP 345 million
and GDP 887 million, corresponding to 0.02% to 0.05% of UK GDP. If one sizes this to the
EU GDP level and converts it to current prices, annual EU consumer benefits would have the
potential to range between EUR 2.7 and 6.3 billion (see Annex 3). However, it is not
expected, at least during the first years of application, that the application of the market
investigation regime would result in similar annual consumer benefits, given that the
Commission would still have to build the expertise in applying this competition instrument. It
is thus more relevant to use as a reference the average consumer benefit per market
investigation, which sized to the EU GDP could be as high as EUR 2 - 2.5 billion.
263 . One important aspect to consider as regards the potential consumer benefits of the market
investigation regime is the growing importance of digital markets for consumers. According
to the Digital Economy and Society Index (DESI) 2020 , internet use has continued to
increase year-on-year with 85% of Europeans surfing the internet at least once per week.224
Using the internet for listening to music, playing games or watching videos is still the most
common activity (81% of individuals). Reading news online is the second most popular
activity (72% of individuals), followed by e-commerce (71%), bank online (66%) and social
networks (65%). According to Eurostat figures, more than 6 out of 10 consumers from the
EU28 made online purchases in 2019, the highest proportion made purchases three to five
times in a period of three months and bought goods or services for a total of between EUR
100 to EUR 499.225 Improving competition enforcement in digital markets is thus particularly
relevant for the protection of European consumers.
6.2.7. Regulatory Authorities
264 . The adoption of a market investigation regime would lead to a more comprehensive
competition-based enforcement toolkit, which would make it better 'value for money'.
would enable a
more effective competition enforcement, leading to more competition on markets.226
265 . The burden that would result for the Commission from the use of the market investigation
regime in digital markets is low compared to the scale of the benefits in terms of a more
effective competition enforcement and overall benefits for the economy, as set out notably in
Section 6.2.2. The introduction of a market investigation regime would not bring about
significant structural changes implying high costs, given that it would be implemented in the
context of the existing competition law enforcement structures.
266 . The main cost involved for the Commission would consist in the staff required for running the
market investigations. Depending on the number of markets concerned and the complexity of
the issues under investigation, under Option II.B, per investigation a total of approximately 10
to 15 full-
should be between 15 and 20 FTEs. These estimates are based on the staff requirements for
more complex Commission antitrust investigations, which generally require between 10 and
EN
55
EN
15 FTEs. A market investigation regime, which focuses not only on the conduct of a dominant
company but on market structure, would likely require a higher number of FTEs.
267 . For the purpose of remedy design and monitoring, additional resources would be required. Per
investigation, a minimum of 2 to 3 FTEs would be required during a period of approximately
5 years, depending on the complexity of the issues identified and the remedies required. Other
additional costs of monitoring could be incurred, namely for external experts, but any estimate
of those costs would be highly speculative without regard to the markets and market failures
at stake. In this regard, it should be noted that a market investigation regime aimed at
addressing emerging and existing market failures would generate expertise that would also
create positive spill-over effects for remedy design in antitrust and merger cases. This would
result in an increase in the ability of the Commission to enforce EU competition law more
effectively in those fields, leading to more competition on the markets concerned. The
gathered expertise could also be deployed to initiate or inform ongoing legislative proposals.
6.2.8. Businesses
268 . The adoption of the market investigation regime would lead to a more comprehensive
competition-based enforcement toolkit, which would result in more open and competitive
markets where companies compete on their merits. As explained in the previous sections, this
would allow businesses to benefit from higher productivity growth, and more incentives to
innovate and to offer a broader range of high-quality products and services. This would also
create the conditions to make European's markets more attractive to investors.
269 . Businesses subject to a market investigation would face administrative costs when
undertaking administrative activities needed to comply with obligations to provide
information. They may also face compliance costs if remedies are imposed at the end of a
market investigation.
270 . There is scarce evidence available about the amount of administrative costs incurred in the
context of competition enforcement. Baker (2003) indicates the average cost with an antitrust
case to be around USD 2.5 million (covering filing fees, lawyers and economic
consultants).227 This would be equivalent to around EUR 2.5 million at current prices.228
Neven (2006) discusses the relative importance of economic and legal fees gathered from the
records of the Airtours case (1999).229 This case was litigated in court after the Commission
prohibited the acquisition by Airtours of First Choice. Airtours succeeded in its appeal in the
Court of First Instance (CFI) and the Commission was ordered to pay the cost that Airtours
had incurred in the procedure. The fees claimed by Airtours added up to more than EUR 2.2
million, corresponding in the current value to EUR 2.7 million.230
271 . There are no reasons to believe that the administrative costs associated to an investigation
under the market investigation regime would be higher than the ones estimated by the above
studies in relation to antitrust investigations.231 Expected administrative costs to businesses
would be at most of a magnitude of EUR 2.5 to 3 million per investigation. In case of smaller
players the costs would be significantly lower as the information required would be
proportional to the size of their businesses.
EN
56
EN
272 . As concerns compliance costs, the Commission for each of the market investigations would
assess the proportionality of the possible remedies. These costs are however not possible to
quantify upfront as the requirements associated to any future intervention are now very
speculative. It stands to reason
these costs [would] be significantly lower than those that
may arise to disadvantaged firms and, ultimately, consumers if, in the absence of early
intervention, the factors that favor concentration, partnered with consumer and/or firm
conducts that merit investigation, indeed cause this concentration 232 A guiding example
normally collect information from parties about the potential cost of
implementing and complying with its remedies. In evaluating such information, the CC will
bear in mind that it has less information than the parties have about how such potential costs
have been estimated and that there might be incentives for parties to overstate the cost of
those remedies that they do not support. The CC is likely to place most weight on estimates of
implementation and compliance costs where parties have provided a clear explanation of how
the estimate was reached, together with supporting evidence as to the assumptions used to
derive those estimates. 233
6.2.9. SMEs
273 . In the particular case of SMEs, the adoption of a market investigation regime would lead to
the creation of a more level playing field allowing SMEs to compete more fairly and grow
throughout the internal market. In fact, by targetting the features of a market that give rise or
contribute to market failures, an intervention under the market investigation regime would
decrease barriers to entry and expansion for smaller innovative players.
274 . In addition, access to digital markets allow SMEs to increase their productivity and reduce
their costs. According to a study from OECD countries, in 2015 only 20% of SMEs engaged
in sales through e-commerce, against 40% of large firms. This digital gap slows productivity
growth and widens inequalities. More competitive digital markets resulting in more affordable
services would allow SMEs an easier access to the digital technologies. Ultimately, given that
SMEs are the bulk of many national economies, a massive adoption of digital technologies by
them will generate a shift of aggregate productivity and welfare.234
6.2.10. Summary of impacts
275 . The two main options described in Sections 5.3.1 and 5.3.2 would lead to a more effective
and coherent competition enforcement regime and thus to an increase in competition in the
markets where the Commission decides to intervene. The main differences between the two
main options in relation to the impact on the above-described parameters reside in the markets
on which the Commission would be able to intervene. In particular, as explained in Section
7.2 below, the Commission would not be able to address under Option II.A most of the
market failure scenarios identified, such as anticompetitive monopolisation, tacit collusion
and common shareholdings. This implies that the impact on the parameters discussed in this
section is expected to be significantly higher under Option II.B.
EN
57
EN
276 . In relation to administrative and compliance costs for businesses, Option II.A would impose
lower costs as it would imply an enforcement in less markets and remedies would only be
enforced against dominant companies. Costs for the Commission would also be lower under
Option II.A, but Option II.B would compensate this by allowing for more effective
competition enforcement.
277 . The magnitude of the impact would not differ significantly between Option II.C and Option
II.D. As explained for each parameter in Sections 6.2.1 to 6.2.8, the digital economy is
becoming more and more important as a driver of productivity and growth, trade and
investment, innovation, employment and income distribution. It is also becoming increasingly
relevant for consumers and SMEs. Moreover, although the emerging and existing market
failures that the market investigation regime aims to address occur in all sectors, digital
markets are particularly prone to the emergence of quasi-monopolistic market structures, as
explained in Section 2.6. Therefore, also a market investigation regime with a horizontal
scope is likely to focus on emerging and existing market failures in those markets.
7. HOW DO THE OPTIONS COMPARE?
278 . This section assesses the effectiveness, efficiency, coherence and proportionality of the
different policy options as compared to the baseline scenario. It does so first within each pillar
discussed in Section 5 and then across the preferred options within each pillar.
7.1. How do the options under Pillar I compare?
279 . This section assesses the effectiveness, efficiency and coherence of the different policy
options as compared to the baseline scenario as explained above.
7.1.1. Effectiveness
280 . Option I.A would tackle unfair practices on the basis of a well-defined self-executing
prohibitions and obligations addressed to large gatekeepers with significant bottleneck power.
This option would be partially effective in as much as such unfair practices can be clearly
identified for the companies in scope and do not need specification in their implementation.
Self-executing rules would be immediately applicable and thus have direct effects.
281 . However, the effectiveness of this blacklist approach would also be limited by three factors: a
series of unfair practices
) would not be captured, even though they too would
constitute equally harmful unfair practices; the list established would essentially be backward
and be less future proof, as it would be based on past behaviours, while regular
updating of the blacklist would be legally cumbersome.
282 . This option would partially contribute to innovative entry, because a series of egregious
blacklist practices that harm such entry would be prohibited, such as specific harmful forms of
self-preferencing, or data related unfair practices. Consumer choice could also increase
directly, notably requiring consumer portability provisions for gatekeepers, making it easier
for companies to switch. Indirectly, consumer benefits would also derive from lower prices
EN
58
EN
for gatekeepers, although rules would need to be designed to avoid adverse effects on security
and privacy, for instance.
283 . Option I.B would be effective in curtaining a wider range of unfair practices than in the self-
executing blacklist under Option I.A, since
would be added, which would lay
down clearly defined and closed list of unfair practices that however require limited
intervention by the competent regulatory body. This option would therefore capture wider
range of unfair practices, be more flexible and therefore more future-proof than the one under
the previous option. It would still provide a high degree of legal certainty through a clear
signaling effect to the market. The need for intervention of the regulatory body does not
significantly affect effectiveness, as it would be limited contextualisation in the clearly
circumscribed frame of a specific provision.
284 . The impact of Option I.B on innovation and consumer choices would be very similar than
under the previous option. Option I.B also meets the objective of effective and coherent EU-
wide oversight through the establishment of a single regulator at the EU level, in cooperation
with a network of national authorities. This regulatory set-up would have information
gathering powers that would help provide regulators with early information on potentially
unfair practices.
7.1.2. Efficiency
285 . The efficiency comparison is based on a cost-benefit estimate235 developed below, which
compares the administrative costs associated with the different options and comparing the
costs with the direct and indirect benefits in each case.
286 . Option I.A: In addition to regulatory enforcement costs, the costs under this option should
consider regulatory burden for gatekeepers while the main benefit would be a healthier and
fairer business environment in terms of competition and innovation. As explained under
Section 5.2.2.1, this option would require resources to enforce this measure as well as national
regulators to respond to eventual consultations from the Commission.
287 . Option I.B: As explained under Section 5.2.2.2, this option would equally require resources
to enforce this measure as well as national regulators to respond to eventual consultations
from the Commission. Given that the platform economy is related to a number of policy areas,
enforcement costs would include coordination costs of national authorities and EU regulators
(e.g. in case of data related practices that may require assistance by the data protection
authorities when tailoring appropriate obligations). In terms of profits, while a quantitative
estimate of profits under different options proves impossible to draw, it is an objective
qualitative assessment to consider that the impact of putting in place an effective and
proportionate regulation addressing dysfunctions in the platform economy would lead to a
different profit distribution. Current excessive gatekeeper profits that could be extracted based
on their grossly imbalanced bargaining power would be distributed to business users and
consumers; hence, the more appropriate (effective but also proportionate) the regulatory
measures, the more optimal the re-distribution of profit.
EN
59
EN
288 . Based on benchmarks of similar practices within the Commission, networks and national
authorities, the administrative costs for the Commission are estimated by an external study at
EUR 8 million, while the compliance costs for market operators are expected to be between
EUR 11.25 million and 12.5 million. The benefits of option I.A and I.B could include price
reductions and associated increases in consumer welfare, as well as greater innovation
potential amongst smaller businesses. Assuming that price reductions are passed onto
consumers, the consumer surplus is estimated to increase by EUR 18-44 billion. In total, when
the full potential of the platform economy is unlocked, it is expected that between EUR 43.7
billion and EUR 174.5 billion in opportunity cost over 10 years in comparison to the baseline
scenario would be recovered.236
7.1.3. Coherence
289 .
with (a) the
7.1.3.1 .
Coherence with the Digital Strategy
290 . Both options
ensuring a fair and competitive digital economy, one of the three main pillars of the policy
orientation and objectives announced in the Communication Shaping Europe's digital future .
Both would a coherent, effective and proportionate ex ante toolbox to address problems in the
digital economy that currently cannot be tackled.
7.1.3.2 .
Coherence with the DSA-package
291 . Both options are coherent with and complementary to the proposal for the update of the e-
Commerce Directive under the DSA-package. While the ECD review is a horizontal initiative
focusing on the issues such as free provision of information society services in the Single
Market, liability of online intermediaries for third party content or safety of users online, the
present options are concerned with economic imbalances, unfair business practices by
gatekeeper platforms and their negative consequences. To the extent that the other initiative
contemplates an asymmetric approach which may impose stronger due diligence obligations
on large online platforms, consistency will be ensured in defining the relevant criteria, while
taking into account the different objectives of the initiatives.
7.1.4. Coherence with other instruments
292 . Both options align with other EU instruments, including with the Charter of Fundamental
Rights, the General Data Protection Regulation, Consumer Law acquis and upcoming
legislative proposals like the Data Act. The definitions to be used under both options are
coherent with definitions used in EU existing legislation, in particular the P2B Regulation.
With their scope targeted to gatekeeper platforms, they complement well the horizontal
obligations for all platforms under that Regulation. Both complement competition law by
addressing practices that either fall outside the existing EU competition rules, or cannot be
effectively addressed by these rules. Both options would minimise the harmful structural
EN
60
EN
effects of these behaviours ex ante
same
behaviours ex post. Option I.B would - while recognising the differences align with the
experiences from the targeted and tailor-made ex ante regulation of specific sectors, including
the rules applicable to electronic communication services.
Proportionality and Subsidiarity
293 . Option I.A is proportionate since it allows achieving the objectives in a targeted manner
imposing limited burden both on gatekeepers and national authorities. It would however leave
uncovered some unfair practices that are equally harmful as the ones identified under the
blacklist but may require a more tailored approach. It may hence underperform in terms of
to tackle those. Option I.A would thus leave greater room for action at national level but at the
same time raise a risk of a regulatory fragmentation.
294 . Option I.B allows achieving the objectives effectively since it sets a comprehensive ex ante
framework providing both for a list of self-executing obligations and prohibitions and a tailor-
made approach through the greylist. At the same time, compliance cost for gatekeepers is
reasonable thus allowing to safeguard the benefits they create for the internal market. Also,
Option I.B is proportionate since it would also address the wider possible range of practices at
EU-wide level while at the same time providing for the possibility of a tailored and
contextualised application of the rules. Option I.B foresees cooperation with NCAs and with
sectorial bodies.
7.1.6. Conclusion
Table 1: Pillar I policy options comparison
Effectiveness
Efficiency
Coherence
Proportionality
I.A Blacklist/whitelist
+
+
++
+
I.B Blacklist/whitelist and
++
++
++
++
grey list
295 . Option I.A
innovation, consumer choice and better oversight. This option is efficient to the extent that
irrespective of the compliance and enforcement costs it entails, it also allows creating rules for
gatekeepers incentivizing them to put in place fairer trading practices. This option is coherent
with other EU instruments.
296 . Option I.B appears as the most apt to address not only current, but also future issues arising
gradually in the quickly changing platform environment; it is hence fully effective in fulfilling
enforcement costs it allows achieving substantial benefits for businesses and consumers by
-normal profits and creating healthier business environment.
EN
61
EN
297 . Both options respect the subsidiarity principle. Proportionality of both options is ensured by
targeting gatekeepers and the selection of clearly harmful practices only. Option I.B which
achieves the objectives more effectively, and allows for proportionality considerations in the
contextualisation and implementation of the greylist practices, and is therefore considered
more proportionate than Option I.A.
298 . Option I.B therefore emerges as the preferred option.
7.2. How do the options under Pillar II compare?
7.2.1. Effectiveness
299 . Under Option II.A (i.e. the conduct and dominance-focused approach) the condition for
intervention under the instrument, similarly to the existing Article 102 TFEU, is establishing
dominance. In order to carry out such analysis, the competition authority in charge of
enforcing the instrument has to, first, investigate and define the relevant markets and, second,
establish who the players are on the market and identify their market position. Once
dominance is established, the competition authority would focus its analyses on the conduct
of the dominant company that may result in or contribute to emerging and existing market
failures. Following the investigation the authority may take different measures, in particular
imposing remedy on the target company.
300 . Under Option II.B (i.e. market-structure-focused approach) the authority would first conduct
a market investigation to map all relevant features of the market including (i) the structure of
the market concerned or any aspect thereof; (ii) the conduct of persons supplying or acquiring
goods or services who operate on that market, whether that conduct occurs in the same market
or not; (iii) and the conduct relating to the market concerned of customers of any person who
supplies or acquires goods or services. The features of a market that may be investigated
include both structural and behavioural ones, including demand and supply-side
considerations and consumer behaviour.
301 . The market structure-based approach includes an investigation of markets where dominant
companies are present to analyse whether there are market features leading or contributing to
emerging and existing market failures. In many cases, anti-competitive conduct by a single
firm may be associated with structural features of the market, for example barriers to entry or
regulation and government policies, or conduct by customers.
302 . Under the market-structure-based approach, the Commission would seek to achieve a
-functioning market in the future. The
market investigation regime would allow the Commission to take an appropriate course of
action that may imply a complex package of different measures, such as (i) make non-binding
recommendations to companies (e.g. proposing codes of conducts and best practices); (ii)
inform and make recommendations/proposals to sectorial regulators; (iii) inform and make
legislative recommendations; and (iv) impose appropriate and proportionate remedies on
companies to deal with identified and demonstrated scenarios of emerging and existing
market failures.
EN
62
EN
303 . In order to achieve Specific Objective (1), i.e. preventing market failures, Option II.A
would not be effective. In cases of an unlawful monopolisation, the dominance by nature
cannot be demonstrated as it has not yet materialised. Under the conduct-focused dominance-
based approach, parallel leveraging could also only be addressed when the company is
dominant. In turn, under Option II.B, the Commission would be able to assess all scenarios
of risk of market failures, by assessing whether the features of the market may lead to a risk of
market failure, irrespective of the fact that there is not yet a dominant undertaking.237
304 . Achieving the Specific Objective (2), i.e. addressing market failures, Option II.A would not
be as effective as Option II.B. This is because market failures are not always solely linked to
a single dominant company. For example, several market scenarios identified as market
failures, such as tacit collusion, the use of pricing algorithms, common shareholding
commonly involve a multiplicity of non-dominant companies and could only be addressed
under the structure-based approach. Similarly, some gatekeepers will often not yet reach the
dominance threshold but their conducts might very well risk to cause market failures.
Conversely, with the market-structure-focused approach the Commission could address all
market failures, as it would investigate the features of a market, to identify whether they lead
to any possible situation of lack of competition, not limited to those where there is a dominant
player.238
305 . Among respondents to the OPC with a relevant knowledge there is a strong view that the
market investigation regime should focus market failures, being applicable to all undertakings
in a market, including non-dominant companies.239 The general view among NCAs with
relevant experience is that the market investigation regime should depart from the traditional
dominance concept and thus be applicable to all undertakings in a market, including dominant
but also non-dominant companies.240 This is because a conduct-focused approach would fail
to cover many important emerging and existing market failures that are not caused by
dominant companies including gatekeeper scenarios where dominance cannot necessarily be
identif[y] a wide set of theories of harm which may also
include narrow oligopolies or markets that will likely move towards dominance if unchecked.
Thus, a dominance-based competition tool would not address several forms of consumer
harm that are due to competition problems. Therefore, [a conduct-focused, dominance-based]
would in our opinion be inferior to a market structure-based competition tool 241
306 . In relation to the sub-options on the scope, both Option II.C and Option II.D would be
equally effective in addressing Specific Objective (1), i.e. preventing market failures and
Specific Objective (2), i.e. addressing market failures, in digital markets.
307 . Option D would also be applicable to non-digital markets, which are not directly part of the
objectives of this initiative. In fact, respondents to the OPC indicated that emerging and
existing market failures occur in all sectors and markets and highlighted that no sector is
immune to (potential) market failures.242 At the same time, a high number of respondents who
indicated that market failures can occur in all sectors and markets mainly pointed to digital
examples. Respondents indicating that emerging or existing market failures mainly or solely
occur in digital sectors/markets argued that the characteristics of the digital sector (e.g.
EN
63
EN
economies of scale and scope, data accumulation and dependency, network effects, lock-in,
zero pricing) make digital markets particularly prone to the emergence of quasi-monopolistic
market structures. Similarly, according to half of the responding NCAs, emerging and existing
market failures may occur in all sectors/markets, whereas others argued that emerging and
existing market failures may occur in some specific sectors/markets, including but not limited
to digital sectors/markets.243 At the same time, NCAs suggested that digital markets are more
prominently affected by emerging and existing market failures than other markets.
308 . The European Consumer Organisation pointed out that competition law enforcement in digital
markets, though important, has not been effective enough in dealing with all problems in
these markets and consequently not able to remedy, let alone prevent, harm to consumers in a
timely manner.244 Indeed, there is an extensive economic literature and numerous reports as
explained in Section 2 describing the growth of digital markets and their particular
characteristics that makes them prone to market failures, and where resources would be better
focused at least at the initial stage of the launch of any new investigation regime. This
emerging and shared view speaks for introducing the market investigation regime first in
digital markets ensuring however via a review clause or other means that once the evaluation
process described in Section 9 allows, the Commission may propose extending the scope of
the investigation regime beyond digital markets.
7.2.2. Efficency
309 . Option II.A would have some, but limited value added in terms of efficiency as compared to
the current competition law instrument under Article 102 TFEU. The main advantage of a
dominance-based approach would be that the intervention would be based on a well-
established competition law concept, i.e. dominance. At the same time, it has the disadvantage
that the market investigation regime would not bring substantial value-added compared to the
investigations under Article 102 TFEU. It would turn on establishing the dominance, which,
apart from being subject to a burdensome and lengthy assessment, would not be able to tackle
fully the specific objectives identified. More specifically, the Commission under the
dominance-based approach would need to define the relevant markets in order to establish
dominance before it can focus on the actual conducts. The regime by its non-incriminating
nature could however potentially solicit a more cooperative attitude by the undertaking
targeted by the investigation.
310 . Option II.B would, however, ensure more efficiency. The Commission would have to
investigate thoroughly the relevant markets, without however having to establish dominance.
The focus would be on the market features as such and the identification and remediation of
the emerging or existing market failures. Option II.B would thus allow the Commission to
address market failure scenarios such as anticompetitive monopolisation, tacit collusion, the
use of pricing algorithms, common shareholdings, and tipping markets in the absence of
dominance. This implies that the impact of Option II.B on the parameters discussed in Section
6.2 is expected to be significantly higher and the potential to generate consumer benefits of
the order of EUR 2-2.5 billion per market investigation would also be higher (see Section
6.2.6). This would more than compensate the possible higher administrative costs to
EN
64
EN
businesses associated to Option II.B, as these are more than off-set by the gains of a market
investigation intervention focused on the market structure.
311 . In relation to the sub-options on the scope, both Option II.C and Option II.D, the later would
require more resources and could, in the short term, risk diverting the focus of the available
resources to market investigations on non-digital markets where the emerging and existing
market failures are less prone to emerge. In the long term, after acquiring the expertise of
conducting market investigations in digital markets and benefiting from the design and
monitoring of the remedies adopted after those investigation, resources could also be applied
to non-digital market investigations without any losses of efficiency. In that case,
Commission may propose extending the scope of the investigation regime beyond digital
markets.
7.2.3. Coherence
312 . In terms of coherence, Option II.A and Option II.B would complement the current EU
policies in the digital sector and would lead to a more comprehensive EU competition law
regime. A market investigation regime could potentially detect new uncharted practices that
might at a later stage could become subject of further regulatory measures provided they can
be well-defined and occur systematically. Focusing specifically on the positive and negative
obligations on large gatekeeper platforms, those rules can efficiently be complemented by a
market-structure based investigatory regime that would be able to capture situations not
already covered by the list of obligations.
313 . In relation to Option II.C and Option II.D both would be coherent with the current EU
policies. Option II.C would be coherent with the recent policy initiatives focusing on the
creation of borderless, fair, and contestable Single Market for digital services (see Section 1),
while Option II.D is coherent with the EU competition rules (such as mergers and antitrust,
see Section 2.3.1) which apply indifferently to all sectors.
314 . The introduction of the market investigation regime, under any of the options considered
would be subject to full respect of the fundamental rights to fair proceedings and good
administration as enshrined in the EU Charter of Fundamental Rights and the European
proceedings under the market investigation regime are administrative in nature and not
criminal or quasi-criminal, the fundamental rights of the Charter enjoyed in the case of
criminal proceedings would not apply.245 However, when acting under the market
by ensuring that undertakings involved enjoy effective fair process rights such as the right to
be heard, the right to a reasoned decision and access to judicial review, including the
possibility to challenge enforcement measures. These rights apply in case of administrative
proceedings.246 This design to preserve fundamental rights is also consistent with if not
superior to the safeguards applicable similar investigation regimes elsewhere in the
world.247
EN
65
EN
7.2.4. Proportionality
315 . Option II.A is proportionate since it allows achieving the objective in a targeted manner
imposing limited burden on undertakings in digital markets. Option II.A does not contain any
additional prescriptive general obligations for companies. It would however require the
cooperation of those companies that are subject to the investigation and/or the remedies.
Overall given the dominance requirement to apply the tool, the circle of companies potentially
subject to the remedies would be narrower.
316 . Option II.B allows achieving the objectives whilst remaining proportionate. Similarly to
Option II.A, it does not contain any additional prescriptive general obligations for companies
and require the cooperation of those companies that are subject to the investigation and/or the
remedies.
317 . Option II.C is proportionate to the objective, as it limits the scope of the market investigation
regime to digital markets where as Section 2 shows the market failures are more prevalent.
318 . Option II.D, corresponds to the evidence that shows that market failures are not limited to
digital markets. At the same time, extending the scope of the initiative beyond digital markets
would go beyond the objectives.
7.2.5. Conclusion
Table 2: Pillar II: policy options comparison
Effectiveness
Efficiency
Coherence Proportionality
Preventing Addressing
Preventing Addressing
market
market
market
market
failures
failures
failures
failures
Option II.A
0
+
0
+
++
++
Option II.B
++
++
++
++
++
++
Sub-option
++
++
++
++
+
++
II.C
Sub-option
++
++
+
+
+
+
II.D
319 . Option II.A is not effective and efficient in preventing emerging market failures and would
not be effective is addressing all existing market failures, as many of the examples of market
failures described in Section 2.6.3 would not be dealt with this option. This option is coherent
with other EU instruments and proportionate since it allows achieving the objective in a
targeted manner imposing limited burden on undertakings in digital markets.
320 . Option II.B is highly effective and efficient in both preventing emerging market failures and
addressing existing market failures. This option is coherent with other EU instruments and
proportionate since it allows achieving the objective in a targeted manner imposing limited
burden on undertakings in digital markets.
EN
66
EN
321 . Option II.C is highly effective in both preventing emerging market failures and addressing
existing market failures. This option would be highly efficiency by focusing on the markets
which are more prone to market failures. This option is coherent with other EU instruments.
This option is also proportionate since it allows achieving the objective in a targeted manner
imposing limited burden on undertakings in digital markets.
322 . Option II.D is highly effective in both preventing emerging market failures and addressing
existing market failures. This option would be risk the loss of efficiency by diverting
resources to less prominent problems in non-digital markets. This option is coherent with
other EU instruments. This option is only partially proportionate as it goes beyond the
objectives.
323 . Based on the above, the most effective, efficient and coherent policy option is a market-
structure focused investigatory regime (Option II.B) that in its initial phase focuses on digital
markets where emerging and existing market failures are the most prominently present (sub-
Option II.C). The market investigation regime, at least in its initial phase shall focus on the
markets where emerging and existing market failures are the most pronounced.
8. PREFERRED OPTION
324 . Based on the preceding analysis in Sections 6 and 7, the preferred policy option is a single
legal instrument consisting of combination of Option I.B under Pillar I to comprehensively
address the issues identified in relation to large gatekeeper platforms, and Option II.BC, under
Pillar II, i.e. a competition-based market investigation regime to address emerging and
existing market failures in digital markets that the existing EU competition rules cannot
tackle, or cannot address in the most effective manner.
8.1. Description of Pillar I
325 . The core substantive elements of this option are explained in Section 5.2. It provides for a
new ex ante regulatory framework applicable to designated gatekeeper platforms, based on
two building blocks: (1) self-executing prohibitions and obligations ( blacklist
) and
(2) a closed list of unfair practices, in form of prohibitions or obligations that would require
greylist ). Such regulatory
intervention would be twofold: First, some obligations would require concretisation as to its
implementation by the gatekeeper platform concerned. Second, some unfair practices would
require contextualisation and fairness assessment as to their applicability in the case at hand.
326 . The option would provide for a procedural framework ensuring effective implementation and
enforcement while respecting due diligence and rights of defence. The Commission would
enjoy appropriate and effective implementation, enforcement as well as sanctioning powers in
case of non-compliance. The designation decision addressed to gatekeepers would be subject
to judicial review and would foresee reassessment at the request of the affected operator in
case of material changes concerning the designation criteria. Case specific intervention under
the greylist could be subject to prior market testing and allow gatekeeper platform to propose
measures on how to comply with their obligations in the greylist. Finally, the Commission
EN
67
EN
could adopt sanctions in case of a non-compliance by the gatekeeper platform(s) with the
blacklist and greylist as further concretised by the Commission.
8.2. Description of Pillar II
327 . The market investigation regime, inspired by similar and tested existing competition
instruments, in particular by that of the UK, is described in detail in Annex 5.1.3. In a
nutshell, the main procedural milestones of the proceedings, such as the opening decision, the
preliminary report and the final decision imposing appropriate and protonate remedies will be
set out in the main legislative act and while detailed rules will be set out in an implementing
regulation. During the investigation, the Commission will investigate the market(s) at stake to
identify emerging and existing market failures under appropriate and effective investigative
powers.
328 . Under the Market Investigation Regime, the Commission would not find infringements by
individual undertakings nor impose fines, therefore the proceedings would be of an
administrative rather than quasi-criminal nature. To ensure compliance with the EU Charter of
fundamental rights in particular the proceeding will ensure appropriate procedural safeguards,
checks and balances, transparency, stakeholder engagement and judicial review.
8.3. Functioning of the feedback loop between Pillars I and II
329 . As explained in Section 5.1.3 the combination of these two sections in one instrument would
lead to substantial and mutual synergies between the two pillars. The practical modalities to
formalise the feedback mechanism between the two pillars will be developed in either the
legislative proposal or the guidance documents related thereto.
330 . The implementation, monitoring and enforcement practice of the regulatory body (who, for
the greylist practices, would have to confirm the applicability of certain obligations to specific
gatekeeper platforms and/or to establish how one or more of the obligations apply in a
specific case) will result in information gathering that could point towards the need for a
market investigation in certain areas (for instance, where numerous complaints flag new
issues which emerge or existent issues which persist despite the application of the ex-ante
measures).
331 . In turn, the market investigation regime will serve to keep the gatekeeper rules future-proof,
where its applications results in the identification of systemic issues in the markets where
large gatekeepers subject to the gatekeeper rules operate. In particular, unfair practices by the
large gatekeepers which are not currently identified in the blacklist or greylist could be
detected through the market investigation regime. In that scenario, the conclusions of a market
investigation could, together with other evidence sources, indicate a need for revising the
practices included in the blacklist and greylist.
EN
68
EN
9. HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED?
9.1. How will the impact of Pillar I be monitored and evaluated?
332 . Given the dynamic nature of online platforms, monitoring and evaluation of impacts needs to
constitute an important part of the proposal. It also responds to explicit demands by
stakeholders, including Member States (e.g. France), for a dedicated monitoring function, and
reflects the self-standing monitoring option considered in the IIA. The monitoring therefore
will be divided into two parts: (1) continuous monitoring which will report on the latest
developments in the market every second year potentially involving the EU Observatory of
the Online Platform Economy and (2) operational objectives and specific indicators to
measure them.
333 . Regular and continuous monitoring will focus cover the following main aspects:
a) Monitoring on scope related issues (e.g. indicators for the designation of gatekeepers,
range of designated gatekeepers and its evolution, use of the margin of appreciation in the
designation)
b) Monitoring on unfair practices (compliance, enforcement patterns, evolution)
c) Monitoring as a trigger for launch of a market investigation
d) Interplay between the two pillars of the instrument
334 . The following specific indicators will used concerning the specific objectives.
Table 3: Measuring indicators
Specific objective
Operational objectives
Measuring indicators
Number of compliance
Address identified unfair
Preventing unfair data related practices and
interventions by the Commission
conducts by large
ensuring the compliance with obligations
per gatekeeper platform/per year
gatekeeper platforms
Preventing identified unfair self-
Number of sanction decisions per
preferencing practices
gatekeeper platform/per year
Promote contestability,
Share of users multi-homing with
Preventing unfair practices concerning
leading in particular to
different platforms or services
increased innovation,
platform
choice for consumers
Share of users switching between
different platforms and services
Enhance legal certainty in
Limit the diverging national regulatory
Number of regulatory interventions
the online platform
interventions
at the national level
environment to preserve
the internal market
Ensure coherent interpretation of
Number of clarification requests
prohibitions and obligations
per year
335 . The monitoring will also take due account of the conceptual work of the Expert Group of the
Online Platform economy under its work stream on Measurement and Economic Indicators.250
EN
69
EN
9.2. How will the impact of Pillar II be monitored and evaluated?
336 . This section describes the monitoring and evaluation that could be applied to assess the
impact of the market investgiation regime, in particular concerning the objectives of
preventing and adressing market failures resulting in inefficient market outcomes in terms of
higher prices, less choice, lower quality and lower innovation to the detriment of European
consumers.
337 . By way of background, it should be noticed that the benefits of the market investigation
regime similarly to the traditional instruments of competition policy would materialise
through specific investigations initiated under the regime. These investigations will affect
precisely defined markets (an analytical concept widely used in EU competition law), which
NACE codes. At this point in time, it is impossible to foresee which markets will be
investigated, what combination of market features and conduct may give rise to emerging and
existing market failures, and which remedies may be found suitable and proportionate to solve
these emerging and existing market failures.
338 . Therefore, the monitoring of the actual impact of the market investigation regime will occur at
two levels:
a) on a general and more qualitative level, documenting the application of the regime;
and
b) on a specific and quantitative level, by performing ex-post studies on selected
investigations run under the regime.
339 . Regarding the first level of monitoring, the market investigation regime will be monitored
from the start of its implementation on an annual basis following the methodology of the
annual Report on Competition Policy.251 This report lists the ongoing merger, antitrust and
state aid investigations, sector inquiries and decisions adopted during the respective year and
provides a brief description of the relevance of the main decisions. In the future, this report
will also list and describe the investigations opened under the market investigation regime and
the final decisions taken, including an overview of the importance of those decisions for the
internal market. Another element that ca
resources deployed for those investigations, in particular as regards the number of officials
who were involved in each investigation.
340 . This will allow an evaluation of the frequency with which the market investigation regime is
used in its application.
341 . As regards the second level of monitoring, the Commission will evaluate a selection of
decisions taken under the market investigation regime. The ex post evaluation of a given
decision under the market investigation regime would be carried out after 2 to 3 years from
the date of the respective remedy implementation. This would also allow sufficient time for
the remedies to produce effects in the markets under consideration. Similar to the case of the
EN
70
EN
application of Articles 101 and 102 TFEU and the Merger Regulation, the Commission
cannot regularly monitor the impact of all its interventions by collecting data on a continuous
basis. This is because of the administrative burden that such a process would impose on
companies and also the lack of resourses available to the Commission. Nevertheless, a
selection of the most important decisions under the market investigation regime (i.e. those
targeting larger markets and those likely to generate higher benefits) will be subject to the
process of ex post evaluation.
342 . The core part of the ex post evaluation of a given decision under the market investigation
regime will be the identification of the effects of the decision on the market and their
comparison with the situation in the counterfactual scenario. The results of the ex post
evaluation reports will also include a discussion on the similarities and/or discrepancies in
comparison to the relevant (peer-reviewed) literature and will provide useful conclusions for
any future remedy selection and implementation. Relevant examples of this type of ex post
evaluations are those produced by the CMA regarding their market investigation decisions.
From the 19 market investigations conducted so far, the UK Competition Authority has
evaluated three decisions, each time on the basis of different elements of evidence and
methodologies.252
343 . The monitoring methodology of the results achieved under the market investigation regime
will depend on the specific decision. Contrary to other legislative proposals, this initiative
does not aim at implementing pre-determined measures in a given market/sector that the
Commission could consequently start monitoring shortly after implementation. Instead, this
initiative will allow the Commission to intervene in markets whenever a emerging and
existing market failure is detected. The outcome of such investigation cannot be foreseen
upfront. A market investigation may result in no intervention if the Commission concludes
based on the results of the market investigation that an intervention is not justified.
Alternatively, it can result in an intervention that will be tailor-made to the strcutural
competition problems identifed and demonstrated in the course of the investigation. This
implies that the monitoring indicators and arrangements will be determined on a case-by-case
basis, which means that, at this stage, only general considerations can be made about the
possible methodologies and data sources to be used (see Annex 5.1.4 for a presentation of
these general considerations).
344 . The ex post evaluation of an intervention under the market investigation regime is likely to
omit elements that can result in both an improvement of competition or additional costs for
companies. In the former case, this leads to an under-estimation of the benefits and in the
latter case, this leads to an over-estimation of the net benefits. For instance, a decision under
the market investigation regime is likely to generate deterrent effects.253 As expained in a
report by the CMA (2017)254, a common thread in all the literature about deterrence effects of
competition enforcement is that there are significant practical difficulties involved in
estimating deterrence. Another example is the cost incurred by businesses as a result of the
implementation of remedies. These are very difficult to estimate, in particular because it is not
possible to impose on companies the burden of providing information about the costs and
benefits of each intervention.255
EN
71
EN
1 https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12418-Digital-Services-Act-package-
ex-ante-regulatory-instrument-of-very-large-online-platforms-acting-as-gatekeepers
2 https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12416-New-competition-tool
3 Inception Impact Assessment of the New Competition Tool, at page 3; and Inception Impact Assessment for
the Digital Services Act package: Ex ante regulatory instrument for large online platforms with significant
network effects acting as gate-
4 J. Crémer, Y.-A. de Montjoye & H. Schweitzer (2018), Digital policy for the digital era, available at
https://ec.europa.eu/competition/information/digitisation_2018/report_en.html.
5 https://ec.europa.eu/digital-single-market/en/eu-observatory-online-platform-economy.
6 Investigation of Competition in Digital Markets, US House of Representatives Majority Staff Report of
October 2020.
7 https://www.accc.gov.au/system/files/Digital%20platforms%20inquiry%20-%20final%20report.pdf.
8 https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/
785547/unlocking_digital_competition_furman_review_web.pdf
9 https://research.chicagobooth.edu/-/media/research/stigler/pdfs/market-structure-
report.pdf?la=en&hash=E08C7C9AA7367F2D612DE24F814074BA43CAED8C.
10 Mission
letter
to
Executive
Vice-President
Vestager,
10
September
2019,
https://ec.europa.eu/commission/sites/beta-political/files/mission-letter-margrethe-vestager_2019_en.pdf.
11
12 Ibid, at page 5.
13 Pursuant to Article 225 of the Treaty on the Functioning of the European Union;
https://www.europarl.europa.eu/doceo/document/JURI-PR-650529_EN.pdf;
https://www.europarl.europa.eu/doceo/document/LIBE-AM-653762_EN.pdf.
14 https://www.europarl.europa.eu/committees/en/digital-services-act-improving-the-funct/product-
details/20200922CAN57451.
15
by reducing barriers to market entry and by regulating large
platforms, an internal market instrument imposing ex-ante regulatory remedies on these large platforms has
the potential to open up markets to new entrants, including SMEs and start-ups, thereby promoting consumer
choice and driving innovation beyond what can be achieved by competition law enforcement alone
the acquisition of significant market power by dominant platforms has
each exerting market dominance over their competitors and imposing their business practices on users
calls on the Commission to assess the possibility of defining fair contractual conditions to facilitate
data sharing with the aim of addressing imbalances in market power; suggests, to this end, to explore options
to facilitate the interoperability and portability of data
16 European Parliament resolution of 18 June 2020 on competition policy annual report 2019
(2019/2131(INI)), 18 June 2020, available at https://www.europarl.europa.eu/doceo/document/TA-9-2020-
0158_EN.pdf.
17
-
Council
Conclusions of 22 March 2019.
18
https://data.consilium.europa.eu/doc/document/ST-8711-2020-INIT/en/pdf.
19 Council of the European Union, Conclusions on a deepened Single Market for a strong recovery and a
competitive, sustainable Europe, 11 September 2020.
20 Ibid.
21 Council of the European Union, Special Meeting of the European Council (1 and 2 October 2020)
conclusions, 2 October 2020.
EN
72
EN
22 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm - Expert advice
for
the
Impact
Assessment
of
a
New
Competition
Tool,
Chapter
2,
https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0420575enn.pdf.
23 These features are present in digital markets as regards both business-to-
-to-
24 For instance, scale economies may translate in a quality advantage (rather than a pure cost advantage) for a
search engine: the more queries a search engine receives, the better the results it can provide since its
algorithm will be improved by learning.
25 J. Crémer, Y.-A. de Montjoye & H. Schweitzer (2018), Digital policy for the digital era, available at
https://ec.europa.eu/competition/information/digitisation_2018/report_en.html.
26 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool,
https://ec.europa.eu/competition/consultations/2020_new_comp_tool/summary_contributions_NCAs_respon
ses.pdf. Some NCAs indicated that some of the questions in the questionnaire did not apply to them, because
they did not have come across this particular feature or scenario in their recent case work. When reporting on
the views expressed by NCAs on particular issues, this IA only reflects the views of those NCAs that did in
with relevant experience
27 See
Summary
of
the
Stakeholder
Consultation
on
the
New
Competition
Tool,
https://ec.europa.eu/competition/consultations/2020_new_comp_tool/summary_stakeholder_consultation.pdf
and
https://ec.europa.eu/competition/consultations/2020_new_comp_tool/contributions_outside_eu_survey.zip.
on on the New Competition
28 See ICN Report on Digital Markets available at https://www.internationalcompetitionnetwork.org/wp-
content/uploads/2020/07/UCWG-Report-on-dominance-in-digital-markets.pdf.
29 Case AT.39740 Google Search (Shopping), Commission Decision of 27 June 2017, paragraph 272.
30 Case AT.40099 Google Android, Commission Decision of 18 July 2018, paragraph 462.
31 Case AT.40153 E-book MFNs and related matters (Amazon), Article 9 Decision of 4 May 2017, paragraph
65.
32 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
33 See Summary of the Stakeholder Consultation on the New Competition Tool.
34 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm - Expert advice
for the Impact Assessment of a New Competition Tool, Chapter 2.
35 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
36 See Summary of the Stakeholder Consultation on the New Competition Tool.
37 Case COMP/C-3/37.792 Microsoft, Commission Decision of 24 March 2004, paragraph 420.
38 Case AT.40099 Google Android, Commission Decision of 18 July 2018, paragraph 464.
39 J. Crémer, Y.-A. de Montjoye & H. Schweitzer (2018), Digital policy for the digital era, available at
https://ec.europa.eu/competition/information/digitisation_2018/report_en.html.
40 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
41 See Summary of the Stakeholder Consultation on the New Competition Tool.
42 Case M. M.8788 Apple/Shazam, Commission Decision of 6 September 2018, paragraphs 221 ff.
43 Case AT.39740 Google Search (Shopping), Commission Decision of 27 June 2017, paragraph 287.
44 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
45 See Summary of the Stakeholder Consultation on the New Competition Tool.
46 See Support study to the Observatory for the Online Platform Economy, report on the main obstacles and
opportunities for multihoming.
47 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
48 See Summary of the Stakeholder Consultation on the New Competition Tool.
EN
73
EN
49 This section quotes Evangelos Katsamakas & Heba Madany (2019) Effects of user cognitive biases on
platform competition, Journal of Decision Systems, 28:2, 138-161, DOI:10.1080/12460125.2019.1620566.
50 Case AT.40153 E-book MFNs and related matters (Amazon), Article 9 Decision of 4 May 2017, paragraph
65.
51 Case COMP/C-3/37.792 Microsoft, Commission Decision of 24 March 2004, paragraph 463.
52 Case AT.40099 Google Android, Commission Decision of 18 July 2018, paragraph 470.
53 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
54 See Summary of the Stakeholder Consultation on the New Competition Tool.
55 Case AT.40099 Google Android, Commission Decision of 18 July 2018, Recital 781.
56 Case AT.40099 Google Android, Commission Decision of 18 July 2018, Recital 783.
57 M. Bajgar, G. Berlingieri, S. Calligaris, C. Criscuolo and J. Timmisn: Industry Concentration in Europe and
North America, OECD Productivity Working Paper (2019); 2016 Economic Report of the President,
available
at:
https://obamawhitehouse.archives.gov/administration/eop/cea/economic-report-of-the-
President/2016; G. Grullon, Y. Larkin, R. Michaely: Are US Industries Becoming More Concentrated?,
Review of Finance, Volume 23, Issue 4, July 2019, Pages 697 743; Market Concentration - Note by Jason
Furman,
Hearing
on
Market
Concentration,
7
June
2018
(available
at:
https://one.oecd.org/document/DAF/COMP/WD(2018)67/en/pdf); G. Gutiérrez and Th. Philippon: Declining
Competition and Investment in the U.S., NBER Working Paper 23583, 2017; G. Gutiérrez and Th. Philippon:
How European Markets Became Free: A Study of Institutional Drift, NBER Working Paper 24700, 2018.
58 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
59 See Summary of the Stakeholder Consultation on the New Competition Tool.
60 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm - Expert advice
for the Impact Assessment of a New Competition Tool, Chapter 2, and G. S. Crawford, P. Rey, and M.
Schnitzer (2020), An Economic Evaluation of the EC
Section V.C.
61 Fletcher, Amelia, Market Investigations for Digital Platforms: Panacea or Complement? (August 6, 2020).
Available at SSRN: https://ssrn.com/abstract=3668289. Amelia Fletcher is a Professor of Competition Policy
at University of East Anglia (Norwich, UK) and Deputy Director at the Centre for Competition Policy. She is
also a Non-Executive Director of the CMA, and a member of the Enforcement Decision Panel at Ofgem. She
was recently a member of the HM Treasury-commissioned Digital Competition Expert Panel, which reported
in March 2019. She has been a Non-Executive Director of the Financial Conduct Authority (2013-20) and
Payment Systems Regulator (2014-20) and Chief Economist at the Office of Fair Trading (2001-2013).
62
certain platforms and their ecosystems
have become unavoidable to access a large variety of contents and services on the internet. Those structuring
platforms have become gatekeepers not only within their services, but for the internet at large
63 In this Impact Assessment the notion of unfair business practices refers to both terms and conditions as well
as the actual business practices of gatekeeper platforms.
64 https://platformobservatory.eu/state-of-play/economic-significance/.
65 In the OPC on Ex Ante Rules, the vast majority of respondents fully agree (71%) or agree to a certain extent
(20%) that there is a need to consider dedicated regulatory rules to address negative societal and economic
effects of gatekeeper power of large platforms. See Synopsis Report Open Public Consultation Ex Ante
Rules.
66 Observatory expert group Progress report on the Measurement of the Online Platform Economy.
67 In the OPC on Ex Ante Rules, respondents in general consider that unfair practices by gatekeeper platforms
have a concerning impact on competition, innovation and consumer choice. Competition is hampered when
gatekeeper platforms create barriers for new entrants to enter the market, thereby resulting in reduction of
investments and innovation and consumer choice stifling. Unfair practices are considered to be the means by
which digital platforms increase the cost of switching or multi-homing for users, thereby limiting market
contestability and preserving their market power. See Synopsis Report Open Public Consultation Ex Ante
Rules.
EN
74
EN
68 Biglaiser, Calvano and Crémer (2019); Afilipoaie, Donders and Ballon (2019).
69 In the OPC on Ex Ante Rules, all categories of stakeholders consider that startups and small companies are
more and more dependent on large platforms to be able to reach their customers. Their reliance on platforms
is considered to vary heavily according to their fields and/or business models. It is argued that start-ups often
end up in a dependent relationship with these mega-platforms from the very beginning. See Synopsis Report
Open Public Consultation Ex Ante Rules.
70 In the OPC on Ex Ante Rules, several hundreds of respondents identified each of these activities.
71 In the OPC on Ex Ante Rules, different stakeholders groups refer to number of unfair practices by gatekeeper
platforms. Some examples (by stakeholder group): (i) telecom operators: unfair and abusive commission
fees, lack of data sharing, unilateral imposition of unfair terms and conditions, imposition of proprietary
services, unfair self-preferencing; (ii) content creators or publishers: unfair self-preferencing, unfair data
pooling, lack of transparency on news feed, inability to offer services to consumers outside gatekeeper
platform or services, inability to obtain customers data or (iii) other business users: arbitrary de-ranking of
the service, excessive extraction of data or limitation on access to consumers data. See Synopsis Report Open
Public Consultation Ex Ante Rules.
72 In the OPC on Ex Ante Rules, 88% of the businesses and business users who replied (155 in total),
encountered issues concerning trading conditions on large platforms. In general, most of the issues presented
by the users are due to a perceived imbalance in bargaining power between platforms and business users,
which hampers competition, fosters uncertainty in relation to contractual terms and also results in lock-in of
consumers. Tensions also arise from the fact that a platform can (1) provide online intermediation or online
search services for business users, and, at the same time, (2) competes with some of its business users on
certain products or services. Another major issue raised has to do with access to data and risk of
disintermediation. See Synopsis Report Open Public Consultation Ex Ante Rules.
73 See for instance J. De Loecker, J. Eeckhout, G. Unger, The Rise of Market Power and the Macroeconomic
Implications, The Quarterly Journal of Economics, Volume 135, Issue 2, May 2020, Pages 561 644; J. De
Loecker, J. Eeckhout: Global Market Power, NBER Working Paper No. 24768, 2018; S. Barkai: Declining
Labor and Capital Shares, Journal of Finance, Volume75, Issue5, October 2020; S. Calligaris, C. Criscuolo,
and L. Marcolin (2018), Mark-ups in the Digital Era, OECD Science, Technology and Industry Working
Papers 2018/10, on trends in firm-level mark-ups across 26 countries for the period 2001-14. They find that
average mark-ups are
sectors over the 2001-14 period.
74 Stock prices of each company are normalised to 100 in 2014, i.e. they are expressed relative to their
respective value in 2014. This graph therefore allows to compare the development of stock prices across
different companies, but not their absolute level.
75 The S&P 500 is a stock market index that measures the stock performance of 500 large companies listed on
stock exchanges in the United States. It is one of the most commonly followed equity indices.
76 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm - Expert advice
for the Impact Assessment of a New Competition Tool, Chapter 2.
77 See M. Motta: Competition Policy -Theory and Practice, Cambridge University Press, 2004.
78 See R. Whish: The New Competition Tool: Legal comparative study of existing competition tools aimed at
addressing structural competition problems, with a particular focu
,
https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0420573enn.pdf.
79 The existence of emerging and existing market failures has been further confirmed by NCA stakeholders
responding to the OPC and further research. See Summary of the contributions of the NCAs to the impact
assessment of the new competition tool. These results of this public consultation will be discussed in more
detail in the following sections, notably Section 2.6.3.
80
aggregate turnover of the undertakings
concerned exceeds given thresholds; irrespective of whether or not the undertakings effecting the
concentration have their seat or their principal fields of activity in the EU, provided they have substantial
EN
75
EN
operations there. Council Regulation (EC) No 139/2004 of 20 January 2004 on the control of concentrations
between undertakings (the EC Merger Regulation), OJ L 24, 29.1.2004, p. 1 22, Article 1(2).
81 While certain forms of unfair business practices can be abusive under Article 102(a) TFEU, this not only
requires a dominant undertaking but generally also an effect on competition. If an undertaking imposes on its
trading partners or obtains from them terms and conditions that are unjustified, disproportionate or without
consideration but without affecting competition on the market, competition law generally does not apply (See
recital 9 of Regulation (EC) No 1/2003. Some national competition laws also prohibit the abuse of economic
dependence). Such behaviour resulting from imbalances in bargaining power that do not affect competition is
usually the domain of unfair trading laws.
82 Regulation (EU) 2016/679 on the protection of natural persons with regard to the processing of personal data
and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection
Regulation), OJ/L 119/1 (2016).
83 While some voluntary efforts for data portability by some platforms have been underway since 2017 in the
large online platforms, which means that actual or potential competitors do not (yet) benefit from this project.
84 See
for
example
on
challenges
of
the
implementation:
https://www.researchgate.net/publication/321198844_The_right_to_data_portability_in_the_GDPR_Towards
_user-centric_interoperability_of_digital_services.
85 https://eur-lex.europa.eu/legal-content/EN/TXT/?qid=1601555056590&uri=CELEX:32005L0029
86 https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=celex%3A31993L0013
87 https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32006L0114
88
https://unctad.org/en/Pages/DTL/STI_and_ICTs/ICT4D-Report.aspx.
89 Digital Economy Compass 2019, Statista, available at https://www.statista.com/study/52194/digital-
economy-compass/#professional.
90 CMA (2017), Pricing algorithms.
91 Ecommerce report 2019.
92 Statista.
93
European-Cross-Border-B2C-E-Commerce-Market-2020-Double-Digit-Growth-Expected-after-
2020 .
94 https://platformobservatory.eu/news/covid-19-and-online-platform-economy/.
95 https://platformobservatory.eu/news/covid-19-and-online-platform-economy/.
96 This notion
digital intensity
digital intensity
Marcolin, M Squicciarini (2018), A taxonomy of digital intensive sectors, OECD Science, Technology and
Industry Working Papers 2018/14.
97 Bajgar, M., Berlingieri, G., Calligaris, S., Criscuolo, C. and Timmis, J. (2019), Industry Concentration in
Europe and North America, OECD Productivity Working Papers, 2019-18, OECD Publishing, Paris.
98 The study used firm-level data sourced from the commercial dataset Orbis® by Bureau van Dijk (BVD). It
provides information on firms' localisation, annual balance sheet and income statements, although the
number of observations per country can vary significantly. It covers the period 2001-14 for 26 countries:
Australia, Austria, Belgium, Bulgaria, Denmark, Estonia, France, Finland, Hungary, Germany, Indonesia,
India, Ireland, Italy, Japan, Republic of Korea, Luxembourg, the Netherlands, Portugal, Romania, Slovenia,
Spain, Sweden, Turkey, the United Kingdom, United States. See also Federico J. Díez, Daniel Leigh &
Suchanan Tambunlertchai, Global Market Power and its Macroeconomic Implications, IMF Working Paper
WP/18/137 (June 2018), https://www.imf.org/en/Publications/WP/Issues/2018/06/15/Global-Market-Power-
and-its-Macroeconomic-Implications-45975.
EN
76
EN
99 See S. Calligaris, C. Criscuolo, and L. Marcolin (2018), Mark-ups in the Digital Era, OECD Science,
Technology and Industry Working Papers 2018/10, De Loecker, J. and J. Eeckhout (2017), The Rise of
Market Power and the Macroeconomic Implications, NBER Working Paper 23687.
100 See S. Calligaris, C. Criscuolo, and L. Marcolin (2018), Mark-ups in the Digital Era, OECD Science,
Technology
and
Industry
Working
Papers
2018/10,
available
at
https://www.oecd-
ilibrary.org/docserver/4efe2d25-en.pdf?expires=1601160653&id.
101 https://www.imf.org/en/Publications/WP/Issues/2018/06/15/Global-Market-Power-and-its-Macroeconomic-
Implications-45975 (see also footnote 74).
102 The economic significance of platforms can be assessed quantitatively, taking in several major areas of
interest: revenues/traffic share, trade flows. https://platformobservatory.eu/state-of-play/economic-
significance/
103 Digitalisation and its impact on innovation, R&I Paper Series, Working Paper 2020/07, Ariel Ezrachi,
Maurice E. Stucke, European Union 2020, p. 34.
104 Ibid., p. 22.
105
meaningful interoperability as important barriers to entry and called for measures that would address them. In
addition, telecom operators recognised the right of data portability in Article 20 GDPR, but referred to the
fact that its scope is limited to specific cases and subject to specific legal bases for processing. In particular,
this right does not foresee continued and far-reaching access possibilities to different categories of data but is
-in effects for individuals.
106 As explained earlier, lock-in refers to situations where users of platforms find it difficult to switch to another
service.
107
benefit an online service by coercing,
steering, or deceiving users into making unintended and potentially harmful decisions S Lewandowsky and
108 Based on a crawl of 11000 websites https://dl.acm.org/doi/10.1145/3359183.
109 https://judiciary.house.gov/calendar/eventsingle.aspx?EventID=3113.
110 Several stakeholders, such as media publishers or game developers, raised concerns about this specific issue
in the OPC on Ex Ante Rules as well as through different legal actions taken both in the Europe and the US
(see for example: https://www.theverge.com/2020/8/13/21367963/epic-fortnite-legal-complaint-apple-ios-
app-store-removal-injunctive-relief).
111 See I Graef (2018) and (2019).
112 Observatory expert group report on differentiated treatment.
113 Most favoured nation or MFN clauses refer to a potentially harmful practice by online platforms which
lower prices) through other means.
114 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm - Expert advice
for the Impact Assessment of a New Competition Tool, Chapter 2.
115 There may of course be instances where it is instead desirable to limit access to data, for instance out of a
concern for privacy. However, the tensions between contestability of a market on the one hand, and data
protection on the other, can be eased where, for instance, anonymised data are sufficient for a competitor to
compete effectively, or where consumers have given their express consent to port their data to a new service
provider.
116 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm Expert advice
for the Impact Assessment of a New Competition Tool, Chapter 3.
117 See Commission Decision Google Search (Shopping), Case AT.39740, public version of the decision
available at: https://ec.europa.eu/competition/antitrust/cases/dec_docs/39740/39740_14996_3.pdf.
118 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
119 See Summary of the Stakeholder Consultation on the New Competition Tool.
EN
77
EN
120 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
121 See Summary of the Stakeholder Consultation on the New Competition Tool.
122 See Summary of the Stakeholder Consultation on the New Competition Tool.
123 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm - Expert advice
for the Impact Assessment of a New Competition Tool, p. 39.
124 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
125 See Summary of the Stakeholder Consultation on the New Competition Tool.
126 Envelopment is a strategy by which a gatekeeper platform (the enveloper) operating in a multi-sided market
(the origin market) enters a second multi-sided market (the target market) by leveraging the data obtained
from its shared user relationships.
127 See Synopsis Report Open Public Consultation Ex Ante Rules.
128 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
129 See Summary of the Stakeholder Consultation on the New Competition Tool.
130 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
131 See Summary of the Stakeholder Consultation on the New Competition Tool.
132 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool. See
Summary of the Stakeholder Consultation on the New Competition Tool.
133 See for instance Ezrachi, A. & Stucke, M. E. (2016). Virtual Competition: The Promise and Perils of the
Algorithm-Driven Economy. Harvard University Press; Harrington, J. (2018). Developing Competition Law
for Collusion by Autonomous Artificial Agents. Journal of Competition Law & Economics, Vol 14(3), pp.
331 363; or Calvano, E., Calzolari, G., Denicolò, V. & Pastorello, S. (2019a). Algorithmic Pricing: What
Implications for Competition Policy? Review of Industrial Organization. Vol 55, pp. 155-171; J. Moore, E.
Pfister, & H. Piffaut (2020): Some Reflections on Algorithms, Tacit Collusion, and the Regulatory
Framework, CPI Antitrust Chronicle.
134 See also European Commission (2017). Algorithms and Collusion Note from the European Union. OECD
Roundtable on Algorithms and Collusion, 21-23 June 2017; Competition & Markets Authority (2018),
Pricing Algorithms: Economic working paper on the use of algorithms to facilitate collusion and
personalised pricing; Autorité de la Concurrence and Bunderkartellamt (2019). Algorithms and Competition;
or Autoridade da Concurrência (2019), Digital ecosystems, Big Data and Algorithms.
135 E. Calvano, G. Calzolari, V. Denicolò, S. Pastorello (2020): Artificial Intelligence, Algorithmic Pricing and
Collusion, American Economic Review (forthcoming) show in a lab environment that Q-learning algorithms
can give rise to supra-competitive prices. Assad, S., Clark, R., Ershov, D. and Xu, L., Algorithmic Pricing
and Competition: Empirical Evidence from the German Retail Gasoline Market (2020), CESifo Working
Paper No. 8521, find that profit margins increase by almost 30% in duopolistic gasoline markets where both
gas stations adopt the algorithmic pricing software.
136 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
137 See, Commission Staff Working Document accompanying the Final Report on the E-commerce Sector
Inquiry, document SWD(2017) 154 final of 10.5.2017, para. 149.
138
Case
AT.
39850
-
Container
Shipping,
see
press
release
available
at:
https://ec.europa.eu/commission/presscorner/detail/en/IP_16_2446
139 Regulation (EC) No 80/2009 of the European Parliament and of the Council of 14 January 2009 on a Code of
Conduct for computerised reservation systems (evaluated in January 2020)
140 OECD, Executive Summary of the Roundtable on Algorithms and Collusion, 26 September 2018.
141 G. S. Crawford, P. Rey, and M. Schnitzer (2020),
. Not yet published. See Annex 4.b.
142 See Summary of the Stakeholder Consultation on the New Competition Tool.
143 For a report on common shareholding in Europe, see S. Frazzani, K. Noti, M. P. Schinkel, J. Seldeslachts, A.
Banal Estañol, N. Boot, and C. Angelici: Barriers to Competition through Joint Ownership by Institutional
Investors, 2020; and Rosati, N., Bomprezzi, P., Ferraresi, M., Frigo, A., Nardo, M.: Common Shareholding in
EN
78
EN
Europe, EUR 30312, JRC, Ispra, 2020, ISBN 978-92-76-20876-1, doi:10.2760/734264, JRC121476
[CHECK IF ALREADY PUBLIC]; for an investigation of the U.S., see M. Backus, C. Conlon, M.
Sinkinson: Theory and Measurement of Common Ownership, AEA Papers and Proceedings, Vol. 110, May
2020, pp. 557-60.
144 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
145 See Summary of the Stakeholder Consultation on the New Competition Tool.
146 [TBA]
147 Competition in the voice-assistant markets will become more and more difficult as the algorithms
data. Incumbent platforms benefitting from large volume and variety of datasets will be able to provide more
refined search results through their own assistants. This is an important competitive advantage vis-à-vis
smaller and/or start-up platforms.
148 2019 BEUC report refers to the importance for regulatory policies to minimising data harvesting and
handling, and ensuring rights to privacy.
149 A number of NCAs expressed an interest in either adopting national market investigation instruments or in
being empowered to apply or to participate in a market investigation regime set up at EU level, see Summary
of the contributions of the NCAs to the impact assessment of the new competition tool. In that case, the
European Competition Network could serve as a mechanism for the necessary discussion and coordination in
the application and enforcement of the instrument.
150 ICF IA study, draft final report.
151
th
across the Union, in particular in digital markets of cross-border nature. See Summary of the Stakeholder
Consultation on the New Competition Tool; and Summary of the Questionnaire to the NCAs on the New
Competition Tool. While respondents indicated that emerging and existing market failures may occur in all
industry sectors, several respondents emphasised that they are particularly prominent in the digital sphere.
152 In addition, Member States may not have the means to adopt appropriate measures to tackle emerging and
existing market failures. Only some NCAs of Member States have instruments that enable them to tackle, to
a certain extent, emerging and existing market failures, such as Greece and Romania. It can be expected that
further Member States will adopt such national tools. Eight NCAs signalled that the competition rules
applicable in their respective Member States have been amended in order to deal with emerging and existing
market failures or that there are plans for doing so, namely Belgium, Bulgaria, Austria, Romania, Lithuania,
Iceland, Germany and Greece. See Summary of the Questionnaire to the NCAs on the New Competition
Tool.
153 JRC, Entry and Contestability, 2020
154
https://ec.europa.eu/competition/consultations/2018_vber/index_en.html (VBER),
https://ec.europa.eu/competition/consultations/2019_hbers/index_en.html (HBERs) and
https://ec.europa.eu/competition/consultations/2020_market_definition_notice/index_en.html (MDN).
155 In the OPC on the Ex Ante Rules, when asked to rate the suitability of the options considered in the parallel
consultations on the New Competition Tool and Ex Ante Rules, the largest group of correspondents rated a
combination of obligations and prohibitions for gatekeeper platforms with a tool to address structural risks
and lack of competition in (digital) markets as most effective (52%) or very effective (15%).
156 The draft Impact Assessment Support Study contains an analysis of business areas including mobile operating
systems, app stores, desktop operating systems, search, social media, advertising (incl. search, display &
video), e-commerce and cloud services.
157 In the feedback to the OPC and the ICF IA study on Ex Ante Rules, several respondents referred to an
arbitrary restrictions by gatekeeper platforms of their ability to freely communicate with their users on
EN
79
EN
different operating systems, in particular restrictions on their ability to inform customers about special offers
and promotions as well as
158 In the feedback to the OPC and ICF IA study on Ex Ante Rules different stakeholders groups raised concerns
about all these practices and their unfairness. As mentioned above, concerns about several of these practices
have also been raised in the context of initiation of the formal proceedings and also in the recent US Report
House of Representatives Majority Staff report.
159 During the OPC on Ex Ante Rules, several stakeholders raised a concern that limited data portability and data
access due to lack of interoperability (e.g., APIs, limits to sharing customer data) creates obstacle for
emerging competitors and also favours consumers lock-in.
160 For example, the replies to the OPC on Ex Ante Rules point to concerns expressed by several stakeholders
about the capacity for gatekeeper platforms to shape the entire ecosystem, acting as quasi private regulators.
This allows them to impose in a non-transparent manner and at their discretion their own technical standards,
trading conditions, and contractual terms. Stakeholders point to the fact that refusing them is not an option as,
due to the lack of competitive offerings, there is no alternative for reaching consumers.
161 As mentioned in footnote 149 above, several NCAs expressed an interest in either adopting national market
investigation instruments or in being empowered to apply or to participate in a market investigation regime
set up at EU level (see Summary of the contributions of the NCAs to the impact assessment of the new
competition tool). While the policy options outlined in the following consider the Commission as being in
charge of a market investigation, a cooperation mechanism could be introduced, based on the model of the
European Competition Network, to enable NCAs to participate and coordinate in the application and
enforcement of the instrument. This would be possible under all policy options and would enable a
coordinated enforcement of the market investigation regime, similar to the mechanism for the enforcement of
Articles 101 and 102 under the European Competition Network.
162 71% of respondents fully agree and 21% agree to a certain extent. See Synopsis Report Open Public
Consultation Ex Ante Rules.
163 https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52016DC0288&from=EN
164 For more information see: https://ec.europa.eu/jrc/en/rhomolo
165 Christensen, M., Conte, A., Di Pietro, F., Lecca, P., Mandras, G. and Salotti, S., 2018. The third pillar of the
Investment Plan for Europe: An impact assessment using the RHOMOLO model (No. 02/2018). JRC
Working
Papers
on
Territorial
Modelling
and
Analysis.
Available
at:
https://ec.europa.eu/jrc/sites/jrcsh/files/jrc113746.pdf
166 People making short films (recording thoughts, ideas or opinions on a subject) and posts them on internet.
167 Please note that the impact for platforms is understood as the monetary effect (spillover) on the economy,
while the impact for youtubers is measured by their earnings/revenues.
168 Similarly, but for the US economy, Barefoot et al (2019) estimated a 5.6%e growth in GDP per year between
2006 and 2016, accounting for 6.5% of current dollar GDP, 3.9% of employment, and 6.7% of employee
compensation. Barefoot, K., Curtis, D., William, A., Nicholson, R. and Omohundro, R., 2019. Research
Spotlight Measuring the Digital Economy. Washington, DC: US Bureau of Economic Affairs. Available at:
https://www.bea.gov/research/papers/2018/defining-and-measuring-digital-economy
169 European Ecommerce Report 2017 While the causal link between GDP growth and the economy of online
platforms is difficult to demonstrate, considering these figures, it is reasonable to expect a relatively
significant positive and growing contribution of the platform economy to the digital internal market and
economic growth.
170 P.130 Impact Assessment Annexes, SWD(2018) 138 final: 71% of consumers would have preferred
platforms for their purchases. This figure is an underestimate given the COVID epidemics but provides
already an idea of the important use of platforms by consumers.
171 Hootsuite Digital 2020 global statshot report https://thenextweb.com/growth-quarters/2020/04/24/report-
most-important-data-on-digital-audiences-during-coronavirus/
EN
80
EN
172 Cross-border Europe, annual analysis of the best global cross-border platforms operating in Europe, EU 28
173 Gilbert, R.J. and Newbery, D.M., 1982. Preemptive patenting and the persistence of monopoly. The
American Economic Review, pp.514-526.
174 P2B SWD
175 IA study, Draft Final Report
176 https://ec.europa.eu/info/files/better-regulation-toolbox-20_en
177 Carayannis, E. and Grigoroudis, E., 2014. Linking innovation, productivity, and competitiveness:
implications for policy and practice. The Journal of Technology Transfer, 39(2), pp.199-218.
https://ideas.repec.org/a/kap/jtecht/v39y2014i2p199-218.html
178 https://www.beuc.eu/publications/beuc-x-2019-054_competition_policy_in_digital_markets.pdf
179 OPC, direct submissions.
180 https://ec.europa.eu/digital-single-market/en/news/regulation-promoting-fairness-and-transparency-business-
users-online-intermediation-services
181 The Online Observatory report on differentiated treatment shows that they may be subject to search diversion,
i.e. platforms may have incentives to a biased order of products/services presentation, which would divert
consumers from products/services they initially intended to buy, pushing them to purchase more and/or more
expensive products/services.
182
privacy concerns. Facebook analytics is a well-known
example.
183 IA study, Draft Final Report
184 https://assets.publishing.service.gov.uk/media/5efc57ed3a6f4023d242ed56/Final_report_1_July_2020_.pdf
185 Statista, total number of users in 2019
186 IA study, Draft Final Report
187 Online platforms benefit from asymmetry of information (they dispose of large data sets compared to
machine learning techniques to facilitate targeting, discriminatory practices and behavioural manipulation.
BEUC considers that such practices can have an impact on demand and distribution of wealth the most
vulnerable consumers might end up paying higher prices than under a competitive price scenario (when
personalisation is combined with commerci
willingness to pay). They may also be used to target biases and reinforce existing or desired viewpoints with
188 Stigler Center Report, George J. Stigler Center for the Study of the Economy and the State The University of
Chicago Booth School of Business, July 2019.
189 https://judiciary.house.gov/calendar/eventsingle.aspx?EventID=3113
190 Investigation of Competition in Digital Markets, US House of Representatives Majority Staff Report of
October 2020.
191 ACCC Report, Digital Platforms Inquiry Final Report, June 2019; Japanese Fair Trade Commission Report
regarding trading practices on digital platforms, October 2019
192 Competition authorities and academics have published a large number of ex-post studies of the results of
enforcement actions, which were surveyed by the OECD in 2013: "Evaluation of competition enforcement
Competition Policy: The EU
post evaluations of specific policy interventions and an assessment of the broader impact of competition
policy. There are also studies measuring the effectiveness of competition enforcement across a large number
of cases. For example, Dutz, M. & Vagliasindi, M. (2000), "Competition policy implementation in transition
economies: An empirical assessment", European Economic Review 44 (4-6), 762-772, used data on a number
of transition economies to show that better implementation of competition law leads to better competition.
193 Hylton, K &
-341.
EN
81
EN
194
itizens and
https://www.bruegel.org/wp-
content/uploads/2019/02/IPOL_STU2019631044_EN.pdf.
195 See OECD, 2014, Fact-sheet on how competition policy affects macro-economic outcomes. Ahn S. (2002)
similarly concluded that
rom various kinds of
and long-run economic growth". Ahn, S. (2002). "Competition, Innovation and Productivity Growth: A
Review of Theory and Evidence", OECD Economics Working Paper No. 317
196 See Nickell, S. (1996) "Competition and Corporate Performance", Journal of Political Economy 104(4), 724-
746, which found that the most competitive firms experienced productivity growth rates 3.8-4.6% higher than
the least competitive. See also Disney, R., Haskel, J., & Heden, Y. (2003). "Restructuring and productivity
growth in UK manufacturing", The Economic Journal, 113(489), 666-694; Blundell, R., Griffith, R., & van
Reenen, J. (1999), "Market Share, Market Value and Innovation in a Panel of British Manufacturing Firms.
Review of Economic Studies", 66(3), 529-54; Januszewski, S. I., Köke, J. & Winter, J. K. (2002), "Product
market competition, corporate governance and firm performance: an empirical analysis for Germany",
Research in Economics, 56(3), 299-332.
197
-
Investment in Knowledge Capital, Growth and Innovation.
198 Buccirossi, P., Ciari, L., Duso, T., Spagnolo, G., & Vitale, C., (2013), "Competition Policy and Productivity
Growth: An Empirical Assessment", Review of Economics and Statistics, 95(4), 1324-1336; and Bucirossi,
P., Ciari, L., Duso, T., Spagnolo, G. & Vitale, C. (2011), "Measuring the deterrence properties of competition
policy: the competition policy indices", Journal of Competition Law & Economics, 7(1), 165-204.
199 CPI ranges between 0 and 1 and is intended to measure the quality of competition policy enforcement. The
CPIs covered seven features resulting in seven individual indicators that were used to calculate an aggregate
CPI incorporating all the information on the competition policy regime of each country.
200 TFP is a widely used measure of productivity in an economy which basically describes how efficient the
economy is in the use of all relevant inputs.
201 It is not possible to replicate the study of Buccirossi et al. (2011, 2013) in order to estimate the quantitative
impact that the market investigation regime would have on the CPI, and hence on TFP growth. This is
because the CPI was built on the basis of the current legislative tools available to the Commission as regards
its antitrust enforcement. However, it is fair to expect a significant impact. According to the calculations by
Buccirossi et al. (2011, 2013) the estimated coefficient of the CPI index is around 0.09. This means that an
increase of 0.1 in the CPI index leads to an increase in TFP growth of 0.009 percentage points. For example,
at the average values in the study of TFP (~1%) and CPI (0.4976), a 10% increase in CPI (e.g. an increase of
0.04976) would lead to an increase in the growth rate of TFP of 0.448 percentage points (i.e.. an increase of
more than 40% in TFP growth, from 1% to 1.448%).
202 See Dutz, M. & Hayri, A. (1999), "Does more intense competition lead to higher growth?", Policy Research
Working Paper No. 2320, World Bank, found a strong correlation between long-run GDP growth, and
effective enforcement of competition rules, on the basis of a cross-section of 52 countries.
203 Barone, G., & Cingano, F. (2008). "Service Regulation and Growth: Evidence from OECD Countries", Bank
of Italy Temi di Discussione (Working Paper) No, 675, demonstrate that productivity growth in
manufacturing is harmed by regulations that reduce competition in services (especially financial services and
energy provision). Studies carried out by Bourlès, R., Cette, G., Lopez, J., Mairesse, J., & Nicoletti, G.
(2013). "Do product market regulations in upstream sectors curb productivity growth? Panel data evidence
for OECD countries. Review of Economics and Statistics, 95(5), 1750-1768", and Forlani, E. (2010).
"Competition in the Service Sector and the Performances of Manufacturing Firms: Does Liberalization
Matter? CESifo Working Paper No. 2942", reaches similar results.
EN
82
EN
204 Benedetti-Fassil, C., Sanchez-Martinez, M. Christensen, P. & Robledo-Böttcher, N. (2017).
205
206 Aghion, P., Bloom, N., Blundell, R., Griffith, R., & Howitt, P. (2005). "Competition and Innovation: an
Inverted-U Relationship", The Quarterly Journal of Economics, 120(2), 701-728. On empirical work, see
Aghion, P., Bechtold, S., Cassar, L., & Herz, H. (2014), "The causal effects of competition on innovation:
Experimental evidence", Journal of Law, Economics, and Organization 34 (2), 162-195.
207 Ahn, S. (2002), "Competition, Innovation and Productivity Growth: A Review of Theory and Evidence",
OECD Economics Working Paper No. 317; Gilbert, R. (2007), "Competition and innovation", Competition
Policy Centre, UC Berkeley.
208
and Scott Stern (eds.), The Rate and Direction of Inventive Activity Revisited, pages 361-404.
209
210
211 Borrell, J. R., & Tolosa, M. (2008). "Endogenous antitrust: cross-country evidence on the impact of
competition-enhancing policies on productivity". Applied Economics Letters, 15(11), 827-831. The authors
found that the impact of antitrust enforcement on total factor productivity is positive and statistically
significant, implying that competition policy effectiveness raises productivity. The estimates suggest that
increasing the average antitrust effectiveness in one standard deviation would increase average total factor
productivity by 28%. The study assesses the combined effect of both competition and trade policy, as
examining competition law alone over-states its effect on productivity growth.
212
213 "Does competition kill or create jobs?", OECD Global Forum on Competition, DAF/COMP/GF(2015)9.
214 See also Dierx A., Heikkonen J., Ilzkovitz F., Pataracchia B., Ratto M., Thum-Thysen A. & Varga J. (2015),
"Distributional macroeconomic effects of EU competition policy A general equilibrium analysis", paper to
be published in a World Bank-OECD publication on Competition Policy, Shared Prosperity and Inclusive
Growth, who estimate that enforcement of the EU competition rules by the European Commission has a
sizeable impact on the creation of new jobs (they estimate around 650 000 after 10 years).
215 "Does competition kill or create jobs?", OECD Global Forum on Competition, DAF/COMP/GF(2015)9,
paragraph 78.
216
5 3076
217
International.
218
219 Ennis, S. & Kim, Y.,
Competition Policy for Shared Prosperity and Inclusive Growth, Washington, DC: World Bank Publishing.
220
221 See for instance Ahn, S. (2002), op. cit. See also for example, a study by the Directorate-General for
Competition of the European Commission on "The Economic Impact of enforcement of competition policies
in the functioning of EU energy markets" (2015), which found that the Commission's decision finding an
abuse of dominance by E.ON lead to a reduction in prices for both wholesalers and retailers to the benefit of
consumers. See also the Note by the UNCTAD Secretariat (2014) "The benefits of competition policy for
consumers".
222 For a comprehensive description of the UK market investigation instrument see R. Whish (2020), The New
Competition Tool: Legal comparative study of existing competition tools aimed at addressing structural
competition problems, with a particular focus on the
.
EN
83
EN
223 See the Impact Assessment 2018/2019 of the UK Competition and Markets Authority, which estimated the
-year period and the ratio of
these benefits to costs.
224 See https://ec.europa.eu/digital-single-market/en/use-internet.
225 See https://ec.europa.eu/eurostat/statistics-explained/pdfscache/46776.pdf.
226 See P. Larouche and A. de Streel (2020), Interplay between the New Competition Tool and Sector-Specific
Regulation in the EU,
https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0120577enn.pdf.
227
-50.
228 Based
on
USD
to
EUR
exchange
rate
published
by
ECB
(https://www.ecb.europa.eu/stats/policy_and_exchange_rates/euro_reference_exchange_rates/html/eurofxref-
graph-usd.en.html)
and
PPP
conversion
figures
published
by
Eurostat
(https://ec.europa.eu/eurostat/databrowser/view/tec00001/default/table?lang=en).
229
006, 741-791.
230 Using
PPP
conversion
figures
published
by
Eurostat
(https://ec.europa.eu/eurostat/databrowser/view/tec00001/default/table?lang=en).
231 PriceWaterhouseCoopers (2003) investigated whether time and costs of business devoted to multi-
jurisdictional merger reviews. The study finds that on average the costs per jurisdiction correspond to EUR
3.3 million in external merger review costs, 65% of these costs are legal fees, 19% are filing fees and 14%
are fees for other advisers. Updating these costs to current values, they would correspond to EUR 4.3 million.
These are however be less representative of the costs of market investigation.
232 G. S. Crawford, P. Rey, and M. Schnitzer (2020),
Section V.C. Not yet published. See Annex 4.b.
233 Competition Commission (2013), Guidelines for market investigations: Their role, procedures, assessment
and
remedies.
https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/284390/cc3
_revised.pdf
234 See https://www.oecd.org/going-digital/sme/resources/D4SME-Brochure.pdf.
235 A robust quantitative cost-benefit analysis proves difficult for the initiative under consideration. We therefore
236 IA study (draft final report)
237 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm - Expert advice
for the Impact Assessment of a New Competition Tool, section 2.3.
238 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm - Expert advice
for the Impact Assessment of a New Competition Tool, section 2.3.
239 See Summary of the Stakeholder Consultation on the New Competition Tool.
240 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
241 See M. Motta and Martin Peitz (2020): Intervention trigger and underlying theories of harm - Expert advice
for the Impact Assessment of a New Competition Tool. See also G. S. Crawford, P. Rey, and M. Schnitzer
(2020),
one of the benefits of the
New Competition Tool lies in its ability to address the conduct and practices of non-dominant firms; hence,
we see no benefit to limiting its applicability to dominant firms
242 See Summary of the Stakeholder Consultation on the New Competition Tool.
243 See Summary of the contributions of the NCAs to the impact assessment of the new competition tool.
244 See workshop with BEUC members on the ongoing Impact Assessment for a possible New Competition
Tool, 1 October 2020.
245 See H. Schweitzer (2020): The New Competition Tool: Its institutional set up and procedural design, Chapter
EN
84
EN
246 Ibid, at Chapter V
https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0420574enn.pdf
247 See R. Whish (2020), The New Competition Tool: Legal comparative study of existing competition tools
investigation tool, at Chapter 4
Chapter 7
248 See Richard Whish (2020), The New Competition Tool: Legal comparative study of existing competition tools
investigation tool.
249 See Summary of the Stakeholder Consultation on the New Competition Tool and H. Schweitzer (2020): The
New Competition Tool: Its institutional set up and procedural design.
250 https://platformobservatory.eu/app/uploads/2020/07/ProgressReport_Workstream_on_Measurement_and_
Economic_Indicators_2020.pdf
251 See https://ec.europa.eu/competition/publications/annual_report/index.html.
252 For instance, in the case of its BAA market investigation, the CMA made an evaluation on the basis of
qualitative elements, namely based on stakeholders interviews and site visits resulting in several case studies,
as well as quantitative elements, including a descriptive and econometric analysis by independent
consultants. One of the models run aimed at testing whether positive changes in overall airport transport
movements and passenger numbers at the divested airports were above that which would have been expected
in the counterfactual scenario. For this the external consultants compared trends in time for three divested
airports and a set of comparable UK airports using a fixed-effects panel regression and test for breaks in time
series. In the evaluation of the home credit market investigation, the UK competition authority interviewed
those involved in the design and implementation of the remedies and undertook some descriptive analysis,
examining monthly data from the LendersCompared website and market-level and firm-specific annual data.
253 Several studies suggest that the deterrent effect of antitrust enforcement may be substantial. See Baarsma B.,
Kemp R., van der Nol, R., & Seldeslachts J. (2012), "Let's not stick together: anticipation of cartel and
merger control in the Netherlands", De Economist, which estimated on the basis of a survey of lawyers and
other advisors that for every sanction decision taken by the Dutch competition authority there are almost five
other cases in which a prohibited act has been terminated or modified in response to advice on competition
law. A study commissioned by the former UK competition authority, the Office of Fair Trading (OFT):
London Economics (2011), "The impact of competition interventions on compliance and deterrence", OFT
Report no 1391, found, based on a survey of more than 800 companies and a small number of law firms, that
for each abuse of dominance case, 12 potential infringements are deterred.
254
255 In case companies would consider that the costs of intervention would outweigh the benefits, they could
voluntarily submit information supporting their view.
EN
85
EN