24 March 2021

Competition Law and Digital Ecosystems: Learning to Walk Before We Run

Frederic Jenny
ESSEC Business School
Chairman, OECD Competition Committee

Summary
Competition authorities are under severe political pressure to intervene quickly against the
digital behemoth for a variety of reasons. Various expert reports have suggested that
traditional antitrust or competition law enforcement and merger control are inadequate or
insufficient to deal with competition issues in the digital sector. In this paper we review a
number of issues for which further economic thinking is required before competition
authorities can make informed and relevant decisions about competition on digital markets.
We review the difference between competition between firms and competition between
digital multi-products ecosystems, data and privacy issues, the relationship between
competition within ecosystems and competition between ecosystems, the implications of
consumer biases on competition in the digital sector, the competition issues raised by
gatekeepers, and the trade-offs involved in interoperability and data portability. We conclude
with a research agenda to help competition authorities avoid the risks of inadvertently giving
in to the political pressure of economic populism or ideology or issuing misguided decisions
which may be ineffective or, even worse, restrict competition or innovation in the digital
sector
************************

Competition authorities are under severe political pressure to intervene quickly against the digital
behemoth for a variety of reasons. The extraordinary financial success of a few companies, some of
which have benefited from the Covid-19 crisis, combined with the idea that they have been highly
successful at optimizing their tax status and avoiding paying taxes, have cemented a political consensus
that they should be scrutinized and that their power should be curtailed. The inability of antitrust
authorities to control the merger wave initiated by the GAFAM in the US and the fact that in Europe,
a limited number of cases brought by the EC Commission against Microsoft, Google or Amazon have
not had a tangible effect on their behavior has meant that a number of critics argue that direct
regulation of the GAFAM would provide a better way to control them than ineffective competition law
enforcement. The rapidity of the changes in the digital sector and the idea that once entrenched
market power may be exceedingly difficult to curb, have meant that competition authorities have felt
that they have to react fast, possibly even before they fully grasp the intricacies of the digital sector.

Electronic copy available at: https://ssrn.com/abstract=3776274

As the digital sector is the newest frontier in competition law enforcement, a number of recent reports,
some commissioned by competition authorities, have analyzed the specificities of the digital sector
and made suggestions on how to deal with competition issues in the digital sector.

For example, in March 2019 the report “Competition policy for the digital era” commissioned by
Margrete Vestager, EC Competition Commissioner, was made public. The report argued that the
specific characteristics of platforms, digital ecosystems, and the data economy required that
established concepts, doctrines and methodologies, as well as competition enforcement more
generally, be adapted and refined. In particular, the report suggested that the time frame and
standard of proof of the consumer welfare approach should be rethought to avoid under-enforcement.
It also suggested that competition enforcement should protect competition for the market as well as
competition in the market, that in some cases data access – and possibly data interoperability – may
need to be imposed through regulation. It added that the acquisition of targets with a low turnover
but a large and/or fast growing user base and a high future market potential needed to be controlled
and that injecting some “horizontal” elements into the analysis of “conglomerate” mergers may be
necessary to control the creation of dominant “ ecosystems” through mergers and acquisitions.

In May 2019, the Furman report (Report of the Digital Competition Expert Panel) was published in the
UK. This report suggested that digital markets are subject to ‘tipping’, in which a winner will take most
of the market, and that if concentration in digital markets can have benefits, it can also give rise to
substantial costs, for example from lower innovation. It concluded that neither merger control nor
antitrust enforcement are well designed for the intensive and ongoing work that needs to be done to
facilitate competition and entry by making it easier for consumers to move and control their data, and
for new digital businesses to interoperate with established platforms. It suggested that merger control
should be more forward-looking and take better account of technological developments and that, with
respect to antitrust, a greater use of interim measures to prevent damage to competition while a case
is ongoing, as well as an adjustment of the appellate standard, were necessary for the digital sector.

In the US in early October 2020 the Antitrust Subcommittee of the US House of Representative’s
Judiciary Committee released a staff report analyzing the challenges presented by the dominance of
Apple, Amazon, Google, and Facebook and their business practices. The report proposes a series of
radical remedies to restore competition in the digital economy, to strengthen antitrust laws, and to
reinvigorate antitrust enforcement. The report recommends structural remedies such as the
separation of digital platforms from commerce and discouraging mergers resulting in 30% or more
market share in order to prevent the increase in concentration. The report also discourages
acquisitions by dominant companies of direct competitor startups or startups operating in adjacent
markets. Its regulatory recommendations suggest rules requiring non-discrimination, data portability,
and interoperability. Setting the threshold for market dominance at 30% share, the recommendations
further call for a European-style offense of abuse of dominance. Additionally, the report calls for
special rules for dominant companies including nondiscriminatory access to their essential facilities
and a prohibition on tying products and services together.

Although they have very different perspectives these EU, UK and US reports have one point on which
they agree: all three consider that, as they are practiced today by competition authorities, antitrust or
competition law enforcement and merger control are inadequate or insufficient to deal with
competition issues in the digital sector. If the US report proposes to abandon the consumer welfare
standard altogether, a proposal which has little chance of success of being adopted, the other two
reports suggest that the difficulty is not with the standard applied by competition authorities but with
the tools that they use to carry on their analysis. Those tools need to be adapted or new tools need to
be adopted.

Electronic copy available at: https://ssrn.com/abstract=3776274

Firms and multi-sided platforms

The question of the adaptation of competition analysis tools to the digital sector requires further
elaboration.

Most of the tools of analysis used in traditional competition law enforcement rest on strong
hypotheses about the organization of economic activities. First, hierarchical firms are supposed to
operate on predefined markets for goods or services where they meet consumers. Second, those firms
are considered to operate on markets where other firms are competing with them by offering
substitutable products or services to those that they offer. Thus a market is the locus of competition.
Third, those firms sell their products or services for a price to consumers. Their goal is to maximize
their profits on the markets on which they sell the product or services they supply by producing up to
the point where their marginal revenue and their marginal cost are equal. Also, because they seek to
maximize profits, firms will never choose to charge a price which is below their average variable cost.
Their marginal revenue and ultimately the price they charge is therefore a function of the intensity of
competition they have on the market and the goal of antitrust is to ensure that the price of those
products or services remains competitive. An individual firm may have market power if it has a large
share of the market and is protected by barriers to entry and/or if consumers are unable to switch to
a different supplier. There are socially desirable and socially undesirable ways for firms to limit the
competition they are facing. On the positive side they can innovate and be more efficient than their
competitors. On the negative side dominant firm or firms with market power may limit competition by
erecting strategic barriers to entry or limiting competition through eviction of their competitors.
Another way firms on a market can limit competition is through a cartel or a merger (if there are
barriers to entry). Hence the customary focus of competition authorities on the horizontal price
dimension of competition, on exclusionary practices, and on merger control.

If we now turn to the digital sector, two major characteristics differ from traditional markets.

First, the internet abolishes distances and reduces the cost of interaction and transactions between
people and groups of people. Digital platforms are thus particularly well suited to facilitating the
interaction between several groups of users (for example between advertisers and users). As a result,
multi-sided platforms are prevalent in the digital sector. This means that the delivery of services is
organized differently in the digital sector than in one-sided firms operating on traditional markets.

Second, digital technologies allow firms and platforms, via digital interfaces, to gather real time data
on users and/or consumer choices, on the environment in which they operate and on the alternatives
they consider, as well as data on the products or services consumed and the reaction of consumers or
users to these products (through sensors), to store this data (on the cloud), and to analyze it (through
artificial intelligence algorithms). Digital technologies thus offer unparalleled opportunities for firms to
develop insights on individual customers and help firms develop products or services that are
customized to customer preferences(1). Access to data and use of data thus become strategic assets
for digital firms to increase their value proposition.

There are several types of platforms: transaction platforms where potential sellers and buyers can
meet and exchange via the platform (for example eBay); services platforms (where the platform
provides some services which attract users and allow suppliers or advertisers to reach these users, for
example Google) and social media (where people interact, for example Facebook). A platform wants
to maximize the value to its users of the interaction between them. This means, among other things,

1
Mohan Subramaniam , Digital ecosystems and their implications for competitive strategy Journal of
Organization Design 9 12 (2020)

Electronic copy available at: https://ssrn.com/abstract=3776274

that there is no longer a direct relationship between the price charged on one side and the cost of
servicing consumers on this side. Indeed, acquiring additional users on one side (for example users of
Google Search) , even non paying users, may be very profitable to the platform if this increases
sufficiently the value of the platform to another (paying) side (for example advertisers).

This type of platform may raise complex questions for competition authorities with respect to the
definition of the market on which these platforms operate, with respect to the assessment of market
power as well as with respect to the assessment of their strategy from a competition law standpoint
and with respect to the qualification of abusive practices. Each individual side of a platform may not
be a relevant market as the platform does not try to maximize its profits on one side independently of
the other side but takes into consideration the interaction between the users on the various sides (2).
If each side is not an independent market, the assessment of the market power of a platform is likely
to be more complex than the assessment of the market power of a firm operating on a traditional
market.

In addition, price competition is often meaningless on the side(s) of the platform where a service is
delivered free of charge in order to attract a large number of other (paying) users on other side(s) of
the platform. On the side or sides on which they do not charge a price, platforms compete on quality
and innovation. As the OECD secretariat observed(3) : “When the prevailing price in a market is zero,
the offering of improved functionality or new features may be one of the only ways for firms to capture
market share”. Indeed, in Google Shopping, the European Commission noted (4) (para. 268) that “In so
far as users expect to receive a service free of charge, an undertaking that decides to stop innovating
may run the risk of reducing its attractiveness, depending on the level of innovation on the market in
question””.

These characteristics mean that the competition law instruments, such as market shares or
concentration ratios, used for traditional markets in order to assess dominance cannot be easily used
when it comes to platforms because of the multiplicity of services they offer simultaneously to
different groups of consumers. It also means that market power indicators based on a comparison of
price and cost (such as price-cost margins or the Lerner index) cannot be used on each side of the
platform to assess its market power. It finally means that the characterization of an abusive practice
of a platform may be either different or more complex than in the case of traditional markets. For
example, pricing below cost (on one side of the platform) may be a profit maximizing strategy and
therefore not qualifiable as a predatory practice unlike what is the case on traditional markets). But a
degradation of quality may be an abuse of dominance by the platform(5) with the added difficulties
that quality is often not easily observable and that quality is multidimensional, which means that the
degradation of some elements of quality may in fact allow an improvement in other dimensions of
quality (for example the use by a platform of data to which it should not have had access may be a

2
On a discussion of the consequences of the mischaracterization by a competition authority of the market on
which a multi-sided platform operates, see for example: Sébastien Broos and Jorge Marcos Ramos “Google,
Google Shopping and Amazon: The Importance of Competing Business Models and Two-Sided Intermediaries in
Defining Relevant Markets”Competition Policy international, December 2015.

3
OECD Competition Committee “Quality considerations in digital zero-price markets” Background note by the
Secretariat 28 November 2018
4
European Commission CASE AT.39740 Google Search (Shopping) Decision, 27/06/2017
5
The decline in quality of the service may manifest itself on either side of the platform. For example, in
February 2019 The Japanese Fair Trade Commission opened an investigation on the practices of Amazon Japan,
Rakuten and Yahoo Japan, all suspected of having degraded the quality of the service they render to online
vendors by forcing them to shoulder an unfair burden in rewarding customers through loyalty programs.

Electronic copy available at: https://ssrn.com/abstract=3776274

degradation of the privacy of the users but also allow the platform to give a more relevant service to
those users).

Platforms have been studied extensively over the last two decades thanks to the fact that non digital
platforms such as payment platforms had attracted the attention of competition authorities. The fact
that they have been studied extensively in the economic literature does not mean that agreed upon
solutions have been found on how to adapt competition enforcement tools to platforms. The payment
platform cases are an area of antitrust law where we have seen contradictory decisions by competition
authorities on how they should be handled as well as much criticism of academics about the approach
taken by courts to these new problems, ‘ for example with respect to the Amex case in the US
Supreme Court (6).

Multi-sided platforms and ecosystems

The rapid evolution of digital platforms forces us to dig deeper into the issue and to consider new
layers of complexities. As platforms have developed a number of them have evolved into multi-product
or multi-services eco-systems and these forms create new challenges for competition law
enforcement. The concept of “ecosystem” comes from the strategic management literature (7). The
term digital eco-system typically designates a complex set (or a meta-organizations (8)) including a
generic modular technology (for example an algorithm) which can be used to supply different types of
products or services, and a platform which uses the modular technology to produce a set of services
which it combines with other services offered by third parties (complementors) in such a way that the
attractiveness of each service is enhanced by the offerings of the other services. The business strategy
literature distinguishes between two types of digital ecosystems.

Production ecosystems arise from the use of digital technology to better connect interdependent
activities and enhance a firm’s value proposition either through an increase in its value chain (9) or
through a modification of the distribution of the value among the firms in the chain. An example of
how a production ecosystem can enhance the value proposition of a firm suggested by M.
Subramanyan is insurance companies’ use of digital technologies to gather data allowing them to offer
new services, such as insurance contracts where premiums vary with the drivers’ recorded driving
performances, in addition to selling their traditional products.

The second type of digital ecosystem, known as consumption ecosystems, also use product or service
generated data to identify demand-related complementary or interdependent services or products
and to co-offer them either directly or through complementors. In other words, in consumption
ecosystems, the platform makes the value proposition of each product or service offered by the digital
ecosystem more valuable than if it were offered on a stand alone basis. Such eco-systems are often
based on modular technological algorithms which can be used to provide a wide variety of services.

6
US Supreme Court, OHIO ET AL. v. AMERICAN EXPRESS CO. Et AL No. 16–1454, June 25, 2018 . For a
comprehensive review of the academic criticism leveled at the American Express US Supreme court decision
see, for example, Jesse Leigh Maniff and Ying Lei Toh, Still on Trial? The Court’s Use of Economic Analysis in the
American Express Case, Kansas City Fed Payment System Research Briefing, April 1, 2020
7
See Jacobides, Cennarmo and Gawer, Towards a theory of ecosystems, Strategic Management Journal,
Volume39, Issue8, August 2018, pp. 2255-2276
8
See , for example, Tobias Kretschmer, Aija Leiponen, Melissa Schilling, Gurneeta Vasudeva, “Platform
ecosystems as meta‐organizations: Implications for platform strategies”, Strategic Management Journal ,
October 2020
9
Subramaniam et al. 2019

Electronic copy available at: https://ssrn.com/abstract=3776274

With respect to consumption ecosystems, we can think of Apple as a multiple product based
ecosystem which offers access to a multiplicity of devices and of services which can be accessed on the
various devices to the benefit of consumers. Another type of digital consumption ecosystem would be
Google, which is a service-based ecosystem built around the distribution of a free Android operating
system to independent OEMs and that provides multiple services (such as mail, office software, maps,
cloud services etc… supplemented by third party complementors’s through applications which can be
accessed on the Google play store. Google services are related to one another in the sense that Google
pools the consumer data it gets from its users to permit a high degree of service personalization across
all services. Amazon, which was originally a digital book retailer, complements its offering with the
distribution of numerous other products, as well as entertainment, video services, gaming services,
payment services, business solutions, a cloud storage service (AWS) and the offerings of a vast array
of independent suppliers selling through its market place. Each Amazon business segment plays a role
in the consumption of other types of products and services belonging to its ecosystem. For example,
third-party resellers bring in the volume the tech giant needs to leverage its investments in fulfillment
and allows it to experiment with innovations in retailing.

The participants in a digital consumption ecosystem may be third parties offering services
complementary to the services offered by the platform or third parties offering services competing
with the services provided by the platform who nevertheless have an interest in being part of the
ecosystem due to the fact their service is more valuable to consumers when it is offered as part of a
set of complementary services than if they offered it separately on their own. This means that within
an ecosystem there can be a complex combination of competition and cooperation between the
platform and the complementors (10).

The role of the core platform in a digital ecosystem is to provide, through the use of the modular
technology, the distinctive core services or products which will be the basis for an attractive offer of
complementary services or products, and to choose the additional products and services from
independent providers in a way that maximizes the value of the whole ecosystem, as well as to
establish the governance of the ecosystem, and to resolve conflicts between its members(11).

To better comprehend what makes an eco-system successful and what are the critical elements of
competition between platforms and eco-systems, it is interesting to compare the insights one gets
from the economic literature with those provided by business policy literature. The economic
literature has largely focused, first, on the determinants of market tipping and on the determinants
which may mitigate this tendency toward concentration and second on the dynamics of competition
among platforms. With respect to the determinants of concentration, economists point to increasing

10
See for example Miklos Dietz, Hamza Khan, and Istvan Rab “How do companies create value from digital
ecosystems?”, Mc Kinsey Digital, 7 August 2020. They state that “collaboration is critical for success within
an ecosystem” and describes the rise of the LEGO eco-system :”LEGO has extensively leveraged partnerships to
expand its brand and keep it relevant for evolving audiences. In 2009, it partnered with Disney to produce toys
and video games and recently, it partnered with Nintendo to launch LEGO Super Mario, featuring an interactive
LEGO Mario gure. The company also has longstanding partnerships with other well-known brands such as
Batman, Harry Potter, and Ninjago. In April 2020, the LEGO Group announced yet another collaboration—an
exclusive global partnership with Universal Music Group to launch a new suite of music-based LEGO products in
2021”
11
Miklos Dietz, Hamza Khan, and Istvan Rab “How do companies create value from digital ecosystems?”, Mc
Kinsey Digital, 7 August 2020 show that there are several possible strategies for a core platform to create
value through the development of an eco-system ( growing the core business through partnerships, Expanding
the network and portfolio on the platform, generating revenues from new products or building an end-to-end
solution to serve business customers and improve operational efficiency)

Electronic copy available at: https://ssrn.com/abstract=3776274

returns to scale for content and network effects(12). However, the economic literature points out that
concentration due to network effects is demand driven and efficient. With respect to the factors which
may prevent markets from tipping, economists point to platform differentiation, multi-homing on the
part of users, interoperability and congestion. With respect to the dynamics of competition, as Bruno
Jullien and Wilfried Sand-Zantman (13) point out: “Given that entry costs are small in most digital
markets, even if a market is prone to tipping there is the possibility for small niche entry exploiting
heterogeneous perceptions of network effects. For instance, a social network may focus on a small
circle of users. Although there is no threat to the large dominant network in the short run, these small
niche competitors will maintain some dynamic competitive pressure on the incumbent as they have
the potential to become future challengers”. This means that market dominance in the digital sector
does not have the same origin as in the non digital sector, does not necessarily have the same negative
consequences in terms of consumer welfare, and can be compatible with an active competition for the
market. On this last point, contrary to the common belief (of competition authorities) that network
effects give the dominant firm such an important advantage that no entrant may be able to replace it,
the introduction of dynamic competition analysis between platforms suggests that (10): ”an initial
advantage is not enough for a less efficient focal firm to maintain its dominant position”. Finally,
dynamic competition analysis shows that consumer beliefs are an important determinant of market
power.

If we now consider the business policy literature, Michael G. Jacobides, Nikolaus Lang, and Konrad von
Szczepanski (14), for example, focus their research on the factors of success for a digital ecosystem.
Their data shows that to explain the success of a platform it is necessary to go beyond the widely held
idea that “scale begets interest, which begets scale, leading to winner-takes-all markets”. “Being an
early mover, is not sufficient to ensure long-term success or dominance, for several reasons. First, the
ecosystem’s product or service may simply not be suited to some customers’ needs. Second, as
technologies and customer needs evolve, later entrants can jump in and take advantage. And third, in
their haste to be first, early movers sometimes build the wrong ecosystem or choose the wrong path”.
As counter examples to the idea that late comers are necessarily at a disadvantage they point to the
bankruptcy of Symbian, an early mobile operating system which had a 70% market share and to IBM
Watson a late comer smart-health space which in three years between 2014 and 2017 grew
explosively, thanks in part to an automated and streamlined onboarding process and developed an
ecosystem of 53 partners. The factors of success they identify are a strong user base (a leading share
in a market), a large number of partners bringing expertise from other industries (for example Amazon
has 67 core partners, roughly double the number of its e-commerce peers, thereby covering logistics,

12
However, some economists have called into question the importance of network effects, winner take all. For
example, Jordi Casanova (“Online search engine competition: why Bing vs Google Search does not support
entry barriers due to network effects, Published 20 june 2020, LinkedLn) argues that the advantage of large
platforms is mostly due to their ability to benefit from inter-market network externalities which allow them to
have more users and operate on more markets. Consequently they are able to better target advertising and
this allows them to get comparatively higher advertising revenues and to engage in more R&D. For the rest, the
evidence Jordi Casanova presents supports the idea that second movers can copy what first movers have done
and gain market share with lower investments than dominant players and that market leaders need to invest
and innovate more if they want to keep their technological leadership over time.
13
Bruno Jullien, Wilfried Sand-Zantman, The Economics of Platforms: A Theory Guide for Competition Policy,
July 2020 available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3502964 (accessed on January 14
2021)

14
Michael G. Jacobides, Nikolaus Lang, and Konrad von Szczepanski, What Does a Successful Digital Ecosystem
Look Like? Boston Consulting Group, 26 june2019

Electronic copy available at: https://ssrn.com/abstract=3776274

finance, media and telecommunications), a big geographical footprint, and a robust capacity for
cooperation because the core platform must manage dozens of partners, across multiple industries
and countries, and different types of relationships.

There are a number of insights that one can derive for competition law enforcement from this brief
consideration of the economic and business policy insights on ecosystems.

First, as mentioned earlier, technical barriers to entry associated with the size of the ecosystem may
not always be the most crucial element explaining the dominance of a few highly successful platforms.
The design of these platforms and their organizational capacities may be equally important factors of
success. This means that potential competition among platforms may be more significant than what
the market shares or size of these platforms would suggest.

Second, understanding the economic factors which may determine the success of a platform may be
useful when analyzing the acquisition of a start-up by a major platform to establish the counterfactual
and get a sense of whether an alternative acquirer might have been a possibility or whether the start-
up would have been likely to develop into a successful platform had it not been acquired.

Third, the description of these ecosystems reveals that while production ecosystems may lead to a
redistribution of the allocation of the value of the chain between its various participing firms, the
situation is different for consumption ecosystems where the core digital platform seeks to develop the
combination of complementary services or products which would maximize the value of the ecosystem
for its users, given its technology. This has several implications for antitrust or competition law
enforcement.

First, the boundaries of a consumption ecosystem encompass a number of complementary products

or services. In addition, the boundaries of such complements are open-ended. Additional or different
combinations of complementary services or products can displace existing ones or non digital firms on
specific markets. Ecosystems can move fluidly from one market to a complementary market thanks to
their modular technologies and their analysis of consumption generated data. For example, an e-
commerce based ecosystem can acquire extensive information about the consumption habits of
consumers, their credit history, their location, their planned acquisition of a car, the kind of cars they
may be interested in etc. This information allows those ecosystems to have an edge allowing them to
add a loan function to the combination of services they already offer, thus competing with banks on
consumer loans (15).

The notion of relevant market does not define either the locus of activity of digital consumption
ecosystems or the locus of competition between ecosystems. An ecosystem providing a different set
of more attractive complementary products or services may displace an existing digital ecosystem. This
means that the market share that an ecosystem has on any particular market (and the existence of
barriers to entry on this particular market) may not be relevant indicators of its market power (16). This

15
For an example, see Subramaniam, M. Digital ecosystems and their implications for competitive strategy. J
Org Design 9, 12 (2020).
16
For example with respect to the Facebook/ Whats App merger, Eleonora Ocello, Cristina Sjödin and Anatoly
Subočs in “What's Up with Merger Control in the Digital Sector? Lessons from the Facebook/WhatsApp EU
merger case”, EC Competition merger brief 1/2015 write: “due to the fast-moving nature of the markets
concerned by the transaction, market shares fluctuate very frequently, sometimes even within weeks or
months. Such fluctuations can be explained by a variety of factors, including, for example, changes in the
perceived 'trendiness' of an app among users, emerging concerns about privacy, or even temporary service
outages. This means that past market shares do not necessarily truly represent the effective competitive force
of market players at the time of the Commission's decision, nor may they be valid for the two- to three-year

Electronic copy available at: https://ssrn.com/abstract=3776274

also means that when one assesses the competitive impact of a merger of consumer complementary
services ecosystems in the digital sector, one should consider potential competition to be the set of
alternative combinations of complementary products or services which could be offered in
competition with those offered by the merging ecosystems. This set is, by the way, not restricted to
ecosystems which have the same business model as the business model of the merging ecosystems.
For example, the Google ecosystem was developed to compete with the then dominant Apple
ecosystem even though the Google ecosystem is centered on a set of services which can be delivered
on terminals, whereas the Apple eco-system is based on a set of terminals which can deliver various
services. Thus, to understand who could be competing with whom in the digital sector one has to look
at a much wider set of possible candidates than in traditional antitrust cases where one can limit
oneself to the potential producers of a specialized but clearly defined product. Understanding the
trends of demand by platform users (so as to understand, for example, that a photo app on mobile
phones could compete with a social media on personal computers because social media users are
migrating toward mobile phone services and that they increasingly feel that exchanging photos is a
crucial part of socializing) is crucial to the definition of potential competition. Clearly, competition
authorities need to invest in understanding the consumption trends of a fast moving digital world.

Second, among digital platforms and ecosystems competition on innovation and on quality is the rule
rather than competition on prices, at least when it comes to consumption ecosystems. Also
innovations tend to be disruptive, that is they fundamentally change the value proposition of
established dominant platforms or ecosystems and replace them with more attractive new service
combinations. Quality is much more difficult to assess objectively than price and it is multidimensional.
This raises two difficult questions for competition authorities respectively related to the relationship
between competition and innovation and to the relationship between the size of ecosystems and the
quality of their services.

The relationship between competition and innovation is complex and not always well understood and
the relationship between competition and disruptive innovation is even less well understood. For
example, competition economists debate the question of whether the acquisition of small innovative
start-ups by large platforms (which under certain circumstances can reduce competition) accelerates
the rate of innovation in the digital sector by giving an incentive to start-ups to innovate in order to be
bought for a hefty amount of money or alternatively can reduce innovation (by allowing a large
platform to prevent the innovations from developing independently) (17). It has been suggested that
large platforms engage in killer acquisitions of app developing firms when they feel that those firms
could either compete with one of their products or services or become the basis of a new offer of
complementary services competing with their ecosystems. The allegation is therefore that they
suppress potential competition through those mergers. But in such cases, the description of the
counterfactual necessary to establish the anticompetitive nature of the transaction is particularly
difficult. What would have happened to the start-up had it not been bought by the merging ecosystem?
Would it have developed into a successful ecosystem or would it have remained an isolated specialized

time period normally considered by the Commission to assess the future effects of a merger”. The lower
informative value of market shares in these markets due to their volatility was also recently recognized by the
General Court in its judgment in Cisco v. Commission: "the consumer communications sector is a recent and
fast-growing sector which is characterized by short innovation cycles in which large market shares may turn out
to be ephemeral. In such a dynamic context, high market shares are not necessarily indicative of market power
and, therefore, of lasting damage to competition"
17
See , for example, Kamepalli, Sai Krishna and Rajan, Raghuram G. and Zingales, Luigi, Kill Zone (April 27,
2020). University of Chicago, Becker Friedman Institute for Economics Working Paper No. 2020-19, Available at
SSRN: https://ssrn.com/abstract=3555915 or http://dx.doi.org/10.2139/ssrn.3555915

Electronic copy available at: https://ssrn.com/abstract=3776274

platform? If the service offered by the start-up is going to be integrated into the offering of the large
ecosystem, is the merger good for consumer because it allows an innovation to reach the market or is
it detrimental to consumers because it eliminates a potential competitor? Unless competition
authorities have a good theory of what makes a digital start-up grow and become successful, their
assessment of the effect of such mergers will be controversial and they may have to turn to the
business economic literature to find clues.

Third, the quality of the services provided by an ecosystem can result from the network effects and/or
from the interaction between the complementary services it offers. The quality of service can also
benefit from the increased access of the platform to data necessary to feed the algorithms it uses to
produce the service offered. Thus, the increased size of an ecosystem may simultaneously allow it to
degrade the quality of its services (for example with respect to the quality of data protection it offers)
because of its increased market power while also allowing it to deliver higher value services to its
consumers. Deconcentration measures may simultaneously affect both dimensions thus leading to a,
uncertain results for users’ welfare.

Business models and competition law enforcement

Further to those general points, various economists and most notably Cristina Caffara and Fiona Scott
Morton (18) have argued that to apply competition law to digital ecosystems, it is necessary, first, to
consider the business model of each ecosystem and to analyze the incentives resulting from such
business models and, second, to consider the economies of scope in data that they can benefit from.

In particular, it is important to distinguish between the ecosystems which are funded by a subscription,
a licence fee imposed on users or by the sale of a product and the ecosystems which are funded by
advertising.

For example, Apple’s ecosystem is funded to a large extent through the sale of Iphones and a number
of other devices as well as through commissions on third parties’ sales. On transaction platforms such
as Amazon, third party sellers pay a commission to Amazon for the sales they realize through the
platform and Amazon also monetizes its services via its direct sales. On matchmaking platforms such
as Uber or Deliveroo, the platform gets a cut of the transactions which take place.

Aggregator ecosystems, such as Google or Facebook, provide free services to their users and the
aggregators fund these services (social networking, search etc….) by selling advertising or through data
collection. The power of the main aggregators comes from the control they have on demand which
leave suppliers with little choice but to accept their conditions to access their platform.

The reason it is important to analyze the business model of a particular platform is that the way it
monetizes its services will shape its incentives. This will have implications both on the extent to which
ecosystems will act in a direction consistent with the welfare of consumers and on the relationship
between the platform and third party providers of services on the platform.

18
See: Cristina Caffarra, Federico Etro, Oliver Latham, Fiona Scott Morton: “Designing regulation for digital
platforms: Why economists need to work on business models”, VoxEU/CEPR, 04 June 2020; Caffarra, C (2019),
“Follow the Money - Mapping issues with digital platforms into actionable theories of harm”, Antitrust Case
Laws e-Bulletin, e-Competitions Special Issue Platforms, August. Etro, F (2020a), “Device-funded vs Ad-funded
Platforms”, mimeo, University of Florence, Etro, F (2020b), “Product Selection in Online Marketplaces”, mimeo,
University of Florence; Hagiu, A and J Wright (2015), “Marketplace or reseller?”, Management Science 61(1):
184-203; Hagiu, A, T-H Teh and J Wright (2020), “Should Amazon be allowed to sell on its own marketplace?”,
mimeo, Boston University

10

Electronic copy available at: https://ssrn.com/abstract=3776274

Roughly speaking the intuition is that ecosystems which charge users for their services will have more
incentive to deliver high quality services to users since they can increase their revenues by charging
more for those high quality services. Thus self preferencing by these platforms to foreclose competing
third parties offering high quality services is unlikely. Indeed it would be against the interest of the
platform.

Platforms which offer their services for free to users will benefit from their superior ability to provide
digital goods which users value and their ability to prevent users from switching to other competing
platforms (thus preserving the ability of the platform to monetize the users’ eyeballs). Such platforms
will be less able to directly monetize the quality of their services and more intent to fund themselves
through the exploitation of the data provided by users. The ability of ecosystems to exploit consumer
data will of course depend on whether they can see what consumers do. If they can, the data to which
they have access can be used in a variety of ways which are not always aligned with the interest of
consumers. This data can be used by the ecosystem to keep users engaged or to develop in adjacent
markets (thanks to the exploitation of the data’s economies of scope) or to preempt competition or to
enhance the effectiveness of the digital advertising through which the ecosystems finance themselves.
It can also be sold to third parties. Such platforms may thus have an incentive to self reference even if
it degrades the quality of their offering or to hoard the data they get to limit competition.

First, the fact that in the digital sphere the business models of ecosystems are varied and that to a
large extent they determine the competitive strategies of ecosystems means that there cannot be a
one size fits all approach to competition law enforcement (or regulation) with respect to ecosystems.
For example, platforms such as Amazon Web Services or Microsoft Azure do not have the ability to see
what their users do unlike platforms such as Facebook or Instagram. Thus some of the competition
concerns associated with the potential exploitation of user’s users’ data may not be relevant for all
platforms.

Second, in the digital sector competing platforms may have different business models and those
business models may evolve over time. This is not unique to the digital sector and there are similar
examples in the non digital sector of the economy. For example pay TV has competed with free to air
TV for a long time.

Yet, until the emergence of the internet, in many non digital sectors all of the competitors shared the
same business model and this business model was relatively stable over time. For example, if we think
of steel making or of automobile manufacturing, all the competitors in each one of these industries
have shared the same business model for decades. This means that it was fairly easy to assess
competition on the merits in these industries and to have general ideas about the practices which
might be problematic form a competition law enforcement standpoint.

In the digital sector, however, it is more frequently the case than in non digital sectors that direct
competitors may have different business models. For example, music streaming services may be
offered both by ecosystems which charge a fee for users (charging, for example, for a premium version,
such as Spotify) or by ecosystem delivering free services (such as Napster did originally). Furthermore,
digital ecosystems may compete with brick and mortar firms (for example in the retail sector or Uber
competing with taxis). This means that, on the same market, the potential competition concerns may
be different from one competitor to the other.

Furthermore, over time the business model of an ecosystem may change and with this transformation
the nature of the potentially problematic competition issue may also change. For example, there is
speculation that the Apple ecosystem may change from one in which the monetization of the
ecosystem largely rests on selling different types of hardware (Iphones or other devices such as tablets

11

Electronic copy available at: https://ssrn.com/abstract=3776274

or computers) that are complementary to the services offered to an ecosystem in which the
monetization will rest more importantly on the payment for services. If this were to happen, the Apple
incentives could shift from maintaining the quality and the complementarity of the services accessible
through its ecosystem to engaging in strategies to foreclose access to competing services. Similarly
there is speculation about the fact that the Amazon ecosystem (based on commissions paid by third
parties selling through the Amazon platform and on the profit margin of Amazon when its acts as a
seller) is migrating toward an ad-funded business-model. This means, first, that, competition concerns
may not necessarily be the same for all the competitors on the market. It also means that to assess
what the competition concerns are, competition authorities need to consider the dynamics of the
business models of the competing players.

Of course, the realization that the business models of competitors on a market can be heterogenous
and evolutive and that the specific features of business models need to be taken into consideration
both to identify potential competition issues and to propose remedies implies that attempts to
complement competition law enforcement with across the board ex ante regulations may be
problematic as some practices (such as, for example, data portability or interoperability) may be pro-
competitive or pro-efficiency in certain ecosystem environments and be potentially anti-competitive
in other ecosystem environments. We will come back to this issue in the section devoted to the EU
proposal for a Digital Market Act.

Lessons from the field

It is useful to assess the extent to which the analytical frameworks developed by economists and
business policy specialists apply in real life situations. Without pretending to exhaustively test their
analysis, it is nevertheless interesting to discuss the meteoric rise of the new social media TikTok
between 2018 and 2020, in the world and in the US, in the light of the academic literature.

TikTok is a 15 second video app used all around the world. It was launched in 2016 in China under the
name Douyin by the Chinese company ByteDance, founded in 2012, and which owns several other
popular networking apps. Today the company value of ByteDance is $75 billion. CB Insights considers
this company “the current-times most successful Unicorn”.

Among the Byte Dance apps, Jinri Toutiao (“today’s headlines”) became China’s largest news site, with
more than 700 million users. Toutiao’s underlying technology learns about readers through their usage
– taps, swipes, time spent on each article, time of the day the user reads, pauses, comments,
interactions with the content and location – but doesn’t require any explicit input from the user and is
not built on their social graph. Today, each user is measured across millions of dimensions and the
result is a personalized, extensive, and high-quality content feed for every user, each time they open
the app. The success of Jinri Toutiao prompted acquisition offers from Baidu, Alibaba, and Tencent, all
of which the owner of BytedDance declined.

Within a year of its launch, the 15 seconds video app Douyin soared to 100 million users in the world
and, in 2017, ByteDance launched iOS and Android versions in international markets under a new name
– TikTok. In just two years, TikTok has emerged to rival companies like Netflix, YouTube, Snapchat ,
and Facebook, with more than one billion downloads in 150 markets worldwide and in 75 languages.

TikTok has a specific identity. It presents itself as a destination where the world’s creativity, the
treasured moments and knowledge thrive. Thanks to the platform, users can transform their ideas or
passions into interesting videos, with just a few clicks.

12

Electronic copy available at: https://ssrn.com/abstract=3776274

Tik Tok’s artificial intelligence network works differently from the algorithms of other social networks.
Unlike Facebook, Instagram, and SnapChat, Tik Tok doesn’t rely on social connections to figure out
what to show you when you open it. TikTok decides what videos to show by analyzing the faces, voices,
music, or objects in videos that users watch the longest and the signals they send when watching these
videos (such as liking, sharing, or commenting).

Although the app initially took off in India and Southeast Asia, it struggled to attract users in the U.S.
and Europe. However, since 2018 it has grown rapidly in the US. Tik Tok had 18.8 million users in the
US in 2018, 37.2 million users in 2019 and 45.4 million users in 2020.

In order to grow in the US, Tik Tok pursued a strategy of active acquisitions. It entered the US market
in November 2017 after buying, for $800 Million, Musical.ly a Shanghai based company with more
than 200 million users worldwide, which in 2014 had already launched an application in the U.S.
allowing users to create 15-second lip-syncing music videos. Over the next few years, ByteDance
snapped up Los Angeles–based Flipagram, a video and photo creation app set to music clips, and
invested $50 million in Live.me, a live streaming app in Los Angeles that is majority owned by Chinese
mobile app developer Cheetah Mobile. Additionally, ByteDance acquired News Republic, a global
mobile news aggregation service based in France, from Cheetah Mobile for $86.6 million. ByteDance
also attempted unsuccessfully to buy a major stake in U.S. social news aggregator Reddit.

Besides its strategic acquisitions, TikTok invested heavily in advertising. It paid its future competitors
Facebook and Google, which together account for 80% of online advertising, billions of dollars for ads
designed to drive mobile app installs.

In 2018 the ByteDance advertising campaign amounted to more more than $1 billion. This included
nearly $300 million on Google advertisements for TikTok. While ByteDance’s ad spending was across
numerous platforms, the majority of its spend was on Facebook. The initial focus on Facebook was
logical. Facebook had helped to pioneer the app install ad, and the amount of time their users spent
on mobile made them the top choice for app install ads in 2018. According to some sources, there
were so many ads that you could find threads on Reddit and other places on how to get rid of the
TikTok ads users were seeing. Reuters reported that “nearly 22% of all ads seen by U.S. Apple device
users on Facebook ad network came from TikTok and its Chinese counterpart Douyin” while Bloomberg
noted in 2019 : “13 percent of all the ads seen by users of Facebook’s Android app were for TikTok.”
Even though Facebook was benefitting from the advertising bought by Tik Tok, it perceived the growth
of Tik Tok as a threat and tried to launch its own short-video app, Lasso, in the US in November 2018.
However, Lasso was only downloaded by 70,000 US users within four months of its launch, in
comparison to about 40 million users for TikTok during the same period of time.

In mid-2018, ByteDance’s U.S. advertising focus shifted to Snapchat and ByteDance/TikTok became
Snapchat’s biggest advertiser. The Wall Street Journal ( 19) reported that inside Snapchat, a debate
erupted over taking the competitor’s ads. Many employees felt they were aiding a rival trying to lure
away its users. Snapchat, under increased pressure from investors to increase profits, faced a difficult
choice with respect to whether it should restrict its largest advertiser from trying to acquire part of its
network. Snapchat decided to keep selling the ads.

Tik Tok did not develop a profitable business model until it had a stable and large following throughout
the world but many companies have used influencers and sponsored hashtag challenges on TikTok to

19
Georgia Wells, Yang Jie , Yoko Kubota TikTok’s “Videos Are Goofy. Its Strategy to Dominate Social Media Is
Serious”, Wall Street Journal, June 29, 2019

13

Electronic copy available at: https://ssrn.com/abstract=3776274

get their message out to young people, and have started taking an interest in showing ads on TikTok
due to its growing popularity.

TikTok brought out a self-serve ad platform in July 2020, which allows marketers to purchase ads
without having to go through the TikTok sales team. TikTok does not use the Google ecosystem and
charges its advertisers on a cost per thousand impressions basis.

Another source of income for TikTok is virtual coins through which both TikTok and TikTokers can earn
money. TikTokers that have at least 1,000 followers can earn virtual gifts by broadcasting a live video.
If users like their content, they can send virtual gifts as tips to them, but each gift is worth a certain
amount of coins and the users have to purchase these coins with real money through the app. Then
they can use these coins to send gifts to their favorite content creators.

This short summary of the growth of TikTok illustrates a number of points made in the economic or
business literature on platforms but also suggests a few possible critical comments.

Tik Tok was developed through the combination of an original modular technology first used for a
news service (the algorithm which allows Tik Tok to identify the tastes of the users) and a technology
allowing users to create and disseminate videos. Tik Tok was a new service differentiated from the
services offered by traditional social networks. The platform had a specific identity (the promotion of
the creative talents of its users) and developed following a niche strategy (appealing to young people)
at first. It maintained a competitive pressure on the more established social networks such as
Facebook and Snapchat and the dominant social network was aware of the looming threat it
represented. Clearly Tik Tok followed a strategy of building its following rather than maximizing its
profits for a number of years and it is only when it had a stable following and a global reach that it
started to maximize its profits. Having an initial advantage was not sufficient for Facebook to fight the
growth of Tik Tok among other reasons because it was unable to come up with an equally efficient
technology in the creation of short videos.

What this example also reveals is that barriers to entry can be very substantial with respect to digital
platforms. Unlike what the previously mentioned quote of Bruno Jullien and Wilfried Sand-Zantman
suggests, entry costs can be very high in digital markets even for niche entrants. As we saw, to build
its service and its following Tik Tok needed to invest a significant amount of resources (at least $2
billion) of which half was an investment in technology and the other half was an investment in an
advertising campaign to build up the expectation among users that Tik Tok would become a core
platform for a particular type of social interaction. Thus upstart investment costs (some of which are
sunk costs) can be very high and act as a barrier to entry.

Two sets of questions are raised with respect to the Tik Tok advertising campaign. First, one may
wonder whether Tik Tok could have successfully developed without such a campaign or whether there
is some small chance that it would have eventually developed through words of mouth among the
users of social networks in a relevant time frame. Because a profit maximizing business model cannot
be implemented by the entrant before its pool of users is sufficiently large to be attractive to
advertisers, and because it is faced with the need to finance significant upstart investment costs in
technology, it seems inevitable that the entrant must follow an aggressive strategy to develop its
following as early as possible and cannot wait for the market to adjust spontaneously. Second, the
incumbents’ advantages are not only due to the fact that they benefit from more substantial network
effects because of the following they have built in the past but also due to the fact that they may be
the gateway through which the entrant needs to go to reach potential users for its innovative service
which could develop into a service competing with that of the dominant platform. As we have seen
this created a dilemma for Facebook and Snapchat. They were both aware of the competitive potential

14

Electronic copy available at: https://ssrn.com/abstract=3776274

of Tik Tok services for their own services but could not resist selling it advertising space given the scale
of Tik Tok advertising(20). One might also argue that had Tik Tok not been backed by a powerful
financial group, it is not clear that Facebook and Snapchat would have felt compelled to accept Tik Tok
advertising.

Finally, it remains to be seen whether Tik Tok will develop into a full fledged social network competing
with Facebook or other social networks by adding additional functionalities to its video creation
capacities.

Spotify

In 1999 Napster marked the beginning of online music sharing by enabling pirates to transfer mp3 back
and forth through peer-to-peer communication channels. Besides the fact that Napster encouraged
piracy, it durably changed the listening habits of music lovers. The public started moving towards music
streaming, away from ownership of records or CDs, and towards subscriptions, rentals and access on-
demand. This evolution toward streaming signaled the beginning of a severe crisis for the global music
industry which had, up to that point, relied on the sale of records and then of CDs.

In the early 2000s the two most popular streaming services were iTunes and Pandora. iTunes was
launched in 2001, when the first generation iPod MP3 player - which stored 1,000 tunes - transformed
the world of digital music. Pandora, an American subscription-based music streaming service owned
by Sirius XM Holdings was launched in 2000 as a business to business company licensing the Music
Genome Project to retailers as a recommendation platform. In 2005, Pandora shifted its focus to the
consumer market by launching Pandora as an internet radio product.

Both services had limitations. With Pandora, you could not listen to the song you wished to. For iTunes,
you had to purchase the songs to listen to them. Other platforms had various drawbacks such as limited
available music content and unreasonable pricing models but they could achieve a scale allowing them
to compete with Pandora, and iTunes.

In 2006 two Swedish entrepreneurs, Daniel Ek and Martin Lorentzon, set out to find a legal and better
way for people to find and to listen to music than Napster or other platforms had been able to deliver.

Besides the fact that Napster encouraged piracy, it took several minutes to download a single song.
The audio quality of the pirated tracks varied wildly. Even popular music pieces were infested with
viruses and malware. As Daniel Ek stated : “ I just really believe that if we create the right product,
which is better than piracy, that people will come.” They also wanted to create a music platform which
would sit between Napster which was free and Apple’s Itunes (launched in 2001) which sold songs
individually for as much as $2 per track.

They set out to providing an interactive service where consumers would be in complete control of their
choices of music, where the music would be delivered instantly with high quality audio and without
the necessity to download. They also envisioned that the platform should have a social sharing
functionality allowing users to create and share playlists of songs with friends.

20
In July 2020 , Facebook’s Instagram intoduced a TikTok style feature ‘Reels’ on which users can create and
share short videos. On 23 November 2020 Snapchat announced the launch of a new section of Snapchat called
Spotlight that will surface vertical video content from users that’s more meme-like and jokey instead of the
day-in-the-life content Snap previously encouraged. Spotlight was launched in 11 countries, including the US,
UK, France, Germany, and Australia. In order to entice users of its new application it committed to distributing
US$ 1 million a day until the end of 2020 (or upward of a total of nearly $30 million) between the most popular
creators on the app, per day, through the end of 2020.

15

Electronic copy available at: https://ssrn.com/abstract=3776274

Their business model rested on a combination of two services : a free service and a subscription fee
based premium service with the hope that a substantial number of users of the free service would
upgrade to the premium service. The free ad-supported service was to allow users to have limited on-
demand online access to the company's music catalog and unlimited online access to its catalog of
podcasts. The premium advertising-free service would allow users to have unlimited access to the
Spotify catalogues for a monthly subscription fee. To attract customers for the pay premium service
Ek and Lorentzon needed a very high quality service allowing users to enjoy a superior audio quality
access to any music instantly and easily. To compete with free music, the experience of using Spotify
had to be so good that users would happily pay $10 per month even though they could easily download
torrents of their favorite music for free.

Achieving this goal required, first, a costly investment in engineering as the quality of the service
delivered was crucial for the success of the Ek and Lorentzon business plan. While other online services
offered 64k AAC+ and/or 128kbps, Spotify delivers three different bit rates, all the way up to 320 kbps–
essentially the output of a CD. This attention to the music experience as well as the technology
experience made Spotify’s product one of the best, if not the best, on the market.

Second, the commercial success of Spotify required users to have access to an attractive and diverse
content to satisfy music lovers.

The founders of Spotify succeeded in convincing the “Big Four” American record labels—EMI, Sony,
Universal, and Warner Music—and a couple of smaller labels who had seen their sales of music
decrease precipitously (the total revenue of the global music industry in 2008 was US$ 16,9 billion, a
33% decline compared to the 1999 revenues of that industry ; sales continued to decrease and in 2009
were half of what they had been in 1999) to agree to share their content for use outside the U.S. on a
limited basis in return for an aggregate 20% stake in Spotify.

The development of Spotify’s initial product, which was known as Spotify AB, began in Sweden in 2006.
Less than a year later, in 2007, Spotify AB was tested by a number of selectively chosen bloggers. Daniel
Ek and Martin Lorentzon launched their service in October 2008 in Scandinavian countries, the United
Kingdom, France, and Spain, barely two years after having started working on it.

Unlike many startups which choose to scale quickly, Spotify, as mentioned earlier, chose to focus on
developing a solid product and decided to grow slowly. To find its initial users, Spotify decided not to
use TV ads and to depend on word-of-mouth, PR, and co-marketing. Spotify contacted influential music
bloggers in Sweden, inviting them to try the new product. Spotify’s (beta) testers were struck by just
how good the company’s product was, even at such an early stage in its development. In just one year,
Spotify had built a product that music bloggers were excited about.

By March 2011, Spotify had more than 10 million tracks in its catalog, 6,7 million users and more than
1 million subscribers in Europe using its service. It needed to scale up and, to this end, to establish itself
on the US market, the largest music market in the world.

Spotify was launched in the US in July 2011, nearly a year after the originally planned launch. The delay
was due to the initial reluctance of the major record labels to grant access to their content in the US.
This reluctance was due to several factors. The major record labels objected to the Spotify business
model and in particular to the free offer which Spotify considered to be an essential offer to attract
users who would eventually migrate to the premium offer. They also considered that the rate of
conversion of users from the free service to the premium pay service (which stood at 7%) was

16

Electronic copy available at: https://ssrn.com/abstract=3776274

insufficient. Finally, they were fearful of the reaction of their brick and mortar distributors (21).
However, the sales of CDs were fast declining (22) and after a hard bargaining the major record labels
decided to grant access to their catalogue in the US. Spotify also partnered with some of America’s
biggest brands such as Chevrolet, Coca-Cola, Motorola, Reebok, and Sprite, among others. It ran
display ads for its American launch partners on its freemium product. In exchange, the American
advertisers gave away exclusive Spotify invitations to their social audiences. By November 2011, the
number of Spotify paid subscribers had more than doubled to 2.5 million. One year later, in December
of 2012, the service had grown to 20 million users and 5 million paid subscribers.

While these developments were taking place, Spotify’s engineering teams continued to build product
features that helped it grow and made its product ever more attractive for users. One of the important
moves during this period was when Spotify officially integrated with Facebook, becoming the official
music player of the social network. Spotify’s integration with Facebook allowed Facebook users to
create a free Spotify account in seconds. The integration with Facebook netted Spotify 1M new
Facebook-connected users in just four days.

By 2013, Spotify had more than 6M paid subscribers, but this number was insufficient to offset its
costs. It needed to increase the number of paid subscribers and decided to make its freemium more
attractive to woo a larger base of users who could then migrate to its paying premium service.

In order to make its freemium offer more attractive, in 2013 Spotify made mobile-streaming, a feature
which up to then had only been available for its premium service users, available to all users and in
2014 it eliminated all time restrictions on its freemium offer. By the end of 2014, Spotify had more
than 50M active users, of whom 12.5M were paying subscribers—the equivalent of a 25% conversion
rate.

By that point it had become clear that the movement toward streaming was here to last and
competition intensified in the market.

In 2012 Google launched its Google Play Music offer of unlimited music streaming for $9,99 and in
2015 Apple launched its Apple Music streaming offer and started to aggressively seek to gain market
share in the US where it overtook Spotify in terms of US subscribers (28 Million for Apple Music versus
26 million for Spotify). However, Spotify had double the number of subscribers world wide compared
to Apple.

In 2016 Amazon launched its long-rumored on-demand music streaming service available to Amazon
Prime members for $7.99 per month or $79 per year, which was cheaper than the premium options
from Spotify or Apple Music. In addition, Amazon announced that owners of one of its voice-controlled
Echo devices would be able to get the service for just $3.99 per month. Amazon was touting Music
Unlimited’s tight integration with its Echo devices and Alexa voice assistant as the real difference with
other streaming services. Users could request songs from Music Unlimited in a variety of ways just
using their voices. Over time, Amazon reduced the price of its music streaming service, eventually
making Amazon Music free for anyone — Echo owner or not — across a range of devices in 2019.
However the Amazon catalogue is much smaller than that of its competitors, and the Amazon service
is not nearly as advanced in terms of its personalization technology.

21
Tower records, one of the largest and the best know retail music chain hod gone into bankruptcy in 2004 and
ceased operations in 2006
22
The total sales of the global music industry declined from $14.6 billion in 1999 to $6.7 billion in 2015 (a 68%
drop in inflation adjusted terms)

17

Electronic copy available at: https://ssrn.com/abstract=3776274

In terms of competitive advantage, Apple has a large active base of over one billion connected iPhones
which gives it a head-start in establishing a connection with non-music streaming customers and a
significant advantage in Customer Acquisition Cost (CAC). Like Apple, Amazon enjoys device
advantages due to its Alexa smart speakers as well as Alexa design integration on non-Amazon
hardware. Despite their advantages, each of Spotify’s competitors’ ability to scale globally is
constrained. Amazon is only present in 45 national markets. Apple, is unable to develop significantly
outside of its iOS operating system, which currently hovers around 25% on a global basis.

Facing an increasingly competitive environment, in July 2015, Spotify engaged in a major push to
improve the quality of its offers.

It introduced the first of its algorithmically curated playlists, Discover Weekly. This feature, uses an
artificial intelligence algorithm to produce a playlist of 30 songs for each user every Monday, based on
their preferences and listening habits. Discover Weekly was immediately very popular with users. By
May 2016, it was being used 40 million Spotify users. Although Apple Music also uses customer data
to suggest and curate music, the Spotify algorithm is much more sophisticated. Most importantly, the
development of Discover Weekly gave Spotify the technological capacity to engage in deep
personalization of a large number of personalized playlists. For example, Spotify added a playlist which
makes music recommendations based on the time of day and activities a user might be doing at a given
time. Users can choose from a selection of time-specific activities: Morning, Morning Commute,
Workday, Evening, and so forth. These playlists are auto-generated based on a blend of the user’s
history and what the service thinks what would be most appropriate music for that user’s context and
likely mood.

That same year Spotify updated its service to make Spotify more of a one-stop-shop for listening by
including spoken, radio-style audio, and short clips from video providers like Comedy Central, Vice,
BBC, Nerdist, TED, and Conde Nast among others. Some of these partners were to produce exclusive
content just for Spotify. Other providers, such as Nerdist, were to make their content available to
Spotify listeners early, with new episodes landing a few hours before they arrive on other platforms.

The introduction of these features led to a strong increase in the number of users starting in 2015.

In order to try to compensate for the advantage of its competitors, Spotify entered into an agreement
with Samsung in 2018. In August 2018, Spotify became Samsung’s new go-to music service provider.
Deepening their partnership Spotify and Samsung announced in March 2019 that Spotify will be pre-
installed on the newest Samsung mobile devices, so music and podcast lovers can start listening as
soon as they turn on their new phone.

Spotify also filed an antitrust complaint in March 2019 against Apple with the European Commission,
alleging that the iPhone maker was harming consumer choice and stifling innovation via the rules it
enforces on the App Store.

Ek argued that Apple which is both the owner of the iOS platform and the App Store—and a competitor
to services like Spotify — operates a platform that, for over a billion people around the world, is the
gateway to the internet and that Apple essentially acting as both a player and referee to deliberately
disadvantage other app developers thus making its competition with Spotify unfair.

More precisely, Spotify was denouncing the fact that if it chose to give its users the in-app payment
option on the App Store, Apple required it to pay a 30% commission on purchases made via the App
store or on upgrades from Spotify Free Service to its Premium service (the tax drops to 15 percent in
the years after the year of the subscription).

18

Electronic copy available at: https://ssrn.com/abstract=3776274

Spotify observed that numerous other apps on the App store and, in particular, Apple Music, its direct
competitor, were not subject to the Apple commission and that paying this commission, would force
it to artificially inflate the price of its Premium membership well above the price of Apple Music.

If on, the other hand, Spotify chose not to use the Apple in-app payment option to sell its service, Apple
applied a series of technical and experience-limiting restrictions such as denying it the right to tell its
Apple users that other ways of upgrading were available or blocking experience-enhancing upgrades
including locking Spotify (and other competitors) out of Apple services like Siri, HomePod, and Apple
Watch.

Thus Spotify asked the Commission to ensure that the same fair set of rules and restrictions—would
apply to all competitors including Apple Music. It also requested that consumers should have a real
choice when it came payment systems, and not be “locked in” or forced to use systems with
discriminatory tariffs. Finally it argued that platforms such as Apple should not be allowed to control,
directly or through its App store, communications between services and users or to place unfair
restrictions on their marketing and promotions.

Two year later, on March 5, 2021, there was speculation in the press that the EC would act on the
complaint in the “near future”( 23). Simultaneously, the United Kingdom CMA announced that it had
launched an investigation into Apple following complaints that its terms and conditions for app
developers are unfair and anti-competitive. The probe has been prompted by the Competition and
Markets Authority’s (CMA) own work in the digital sector, as well as several developers reporting that
Apple’s terms and conditions are unfair and could break competition law. The CMA press release
stated : “All apps available through the App Store have to be approved by Apple, with this approval
hinging on developers agreeing to certain terms. The complaints from developers focus on the terms
that mean they can only distribute their apps to iPhones and iPads via the App Store. These complaints
also highlight that certain developers who offer ‘in-app’ features, add-ons or upgrades are required to
use Apple’s payment system, rather than an alternative system. Apple charges a commission of up to
30% to developers on the value of these transactions or any time a consumer buys their app. The
CMA’s investigation will consider whether Apple has a dominant position in connection with the
distribution of apps on Apple devices in the UK – and, if so, whether Apple imposes unfair or anti-
competitive terms on developers using the App Store, ultimately resulting in users having less choice
or paying higher prices for apps and add-ons”.

The CMA press release ( 24) also referred to the need for the CMA to coordinate with the EC
investigation. This press release stated that : “The European Commission (EC) currently has four open
antitrust probes into Apple, which were launched prior to the end of the UK’s Transition Period. These
include three open investigations into Apple’s App Store. The CMA continues to coordinate closely
with the EC, as well as other agencies, to tackle these global concerns”.

Parallel to these developments Spotify also diversified its regional footprint. For example, it launched
its service across 13 countries in the Middle East and North Africa In 2018. It launched its service in
India in February 2019 and picked up two million additional users within two months of its launch
there.

By 2020 Spotify offered its services globally with a presence in 92 countries and territories. It had 320
million monthly active users (MAUs) and 144 million Premium Subscribers as of September 30, 2020.
In 2021 Spotify had a 35% share the global streaming market whereas Apple’s market share was 19%,

23
See Javier Espinosa “ Apple faces EU charges on music streaming », Financial Times , 5 March 2021
24
4 March 2021 Press release “CMA investigates Apple over suspected anti-competitive behavior”

19

Electronic copy available at: https://ssrn.com/abstract=3776274

Amazon’s market share was 15%, Tencent market share was 11% and YouTube 6%. Half of Spotify's
revenue originated from the U.S (about 30%) and the UK (about 20%). The other 50% originated in
other countries throughout the world.

Spotify was listed on the New York Stock Exchange in April 2018 and its initial value was $29.5 billion.
Its value reached $50 billion in June 2020. Since its creation Spotify has raised $2.6 billion in total
funding to finance its growth. For the first seventeen years of its life Spotify has consistently lost money
and the amount of its yearly losses increased with the number of its users until 2017.

Year Revenues Operating

losses
Million $ Million $
2013 98
2014 1.085 191
2015 1.929 236
2016 2.952 349
2017 4.090 378
2018 5.259 43
2019 6.764 73

Such figures raise the question of whether Spotify can ever be profitable. The jury seems to be still out
on the question of whether the Spotify business model is viable. On the one hand, Spotify has been
able to reduce its engineering costs in the last few years, as well as to renegotiate its contracts with
two of the major music majors in 2017, thus lowering the amount of its payments to the music industry.
On the other hand, price competition with regard to user’s service is particularly stiff as a number of
Spotify competitors sell speakers, home assistants and tablets which allow people to use their
streaming apps directly with the brands' own products (which Spotify cannot do) and are willing to
incur losses on the streaming of music in order to subsidize the sale of their other products or services
(more phones for Apple, more ads for Google, deeper engagement to its ecosystem in the case of
Amazon). In addition, it seems that after a user transitions from Spotify’ freemium tier to a
subscription, it takes Spotify about 12 months to recoup the cost of having served music to that user
for free. This could mean that the cost of acquisition of paying users is structurally too high for Spotify.

This quick overview of the development of Spotify suggests several comments.

First, as mentioned earlier, the successful development of an application of this type requires
considerable financial means. Therefore barriers to entry on the market for music or video streaming
applications are very significant and result both from the engineering cost associated with the building
of a high quality product and from the cost of the acquisition of a sufficient number of paying users to
incite the major music companies to contribute to the catalogue which can be accessed by users.

Second, the success of Spotify and its acquisition of a preeminent position in the global streaming
music market is in large part due to its ability to anticipate the shift in demand for music consumption
and to provide both a higher quality audio experience to its users than its competitors and a new
business model. Thus the source of the initial appeal of Spotify was the disruptive nature of its
innovation.

Third, in a second stage of its development Spotify was able to use artificial intelligence algorithms to
personalize its offers in a way which greatly enhanced the value of its product for users compared to

20

Electronic copy available at: https://ssrn.com/abstract=3776274

what its competitors could achieve. This modular technology was then used to develop a number of
music and video playlists offering complementary personalized services to consumers.

Third, the fact that in a digital ecosystem the same firm may combine two functions as a supplier of a
digital service but also as the rule setter of a market place on which third party app developers sell
services that compete with the service of the firm controlling the market place may raise competition
issues when such a firm can control the exclusive access to a large number of potential users. Such a
gatekeeping function may come from the fact that the firm bundles the access to its market place and
its operating system.

Data issues

Data play several simultaneous roles for platforms and consumption ecosystems. They can be an input
exhibiting increasing returns if used in an artificial intelligence context; they can be a strategic asset
allowing the platform to enhance its economic power and prevent entry, they can be a valuable by-
product of the activity of the platform which can be commercialized and the respect of privacy can be
a non price parameter of competition ( 25).

Competition authorities have had only limited experience with the economics of data in the non digital
world since in the non digital world the link between the accumulation of data and the quality of a
service provided using this data is less pronounced than it is in the digital world(26).

One possibility which is of concern to competition authorities is that large ecosystems or platforms
enjoy information advantages which allow them to have market power. As Geoffrey Parker, Georgios
Petropoulos and Marshall Van Alstyne (27) observe: “Through the data they collect from other market
participants, platforms have superior information over the ecosystem which they can use to create
ecosystem benefits by increasing the value of their intermediation services. As a platform facilitates a
larger number of interactions, users can have greater challenges when switching to substitute
intermediation services. Network effects favor match variety and match quality on larger platforms as

25
See for example Eleonora Ocello, Cristina Sjödin and Anatoly Subočs : “What's Up with Merger Control in the
Digital Sector? Lessons from the Facebook/WhatsApp EU merger case”, EC Competition merger brief 1/2015. in
its Microsoft/LinkedIn decision, the EC held that data privacy is ‘a significant factor of quality’ in the market for
Professional Social Networks. Similarly, in Google/DoubleClick, the US Federal Trade Commission ("FTC")
acknowledged that mergers can ‘adversely affect non-price attributes of competition, such as consumer
privacy’.

26
However there have been antitrust cases in the non digital sector where the issue of access to data has been
a competition issue. For example, in the context of the deregulation of sectors which were formerly
characterized by a state monopoly, the refusal of the former state monopoly, now competing with private
firms for after sales services, to give access to its data base of clients has been held an abuse of its dominant
position in various jurisdictions. The French Autorité de la concurrence had to deal with this issue in 2014,
when it received a complaint against GDF Suez. The measures requested by the complainant included ordering
GDF Suez to give competing suppliers of natural gas access to certain customer data including the customers'
names, addresses, telephone numbers and consumption profiles. (Autorité de la concurrence, Décision n° 14-
MC-02 du 9 septembre 2014 relative à une demande de mesures conservatoires présentée par la société Direct
Energie dans les secteurs du gaz et de l’électricité, p. 52-53). The fact that such data were protected under the
French Loi Informatique et Libertés did not prevent the Autorité from ordering GDF Suez to grant access to this
data. In line with the recommendations made by CNIL (the French data protection authority), the Autorité
merely required GDF Suez to inform its customers that competitors would be able to request access to their
personal data and that they had the possibility to refuse such access.
27
Geoffrey Parker, Georgios Petropoulos Marshall Van Alstyne “Platform Mergers and Antitrust”, January 10,
2021, SSRN …….

21

Electronic copy available at: https://ssrn.com/abstract=3776274

illustrated by search and e-commerce”. Further, platforms having market power can strategically deny
access to their data to potential competitors (whether a competitor of the platform within the
ecosystem or a competing ecosystem) when access to this data is necessary for the competitors to
match the level of quality of the offers of the large platform ecosystem.

In competition law one classical solution to this problem has been to facilitate data transfers from the
dominant firm to its competitors. For example, in the EU refusal to grant access to information was
considered abusive in the Microsoft case (28). A dominant firm must grant third parties access to data
it has collected if (i) the data constitute an "essential facility" to the activity of the third party asking
for access (i.e. the data is indispensable for the business activities of the third party), (ii) there are no
objective considerations that would justify a refusal to grant access and (iii) it is likely that a refusal
would exclude all competition in the other market.

However, in practice such an approach is limited to correcting abuses and requires having the answer
to several complex questions. The first question is whether the data gathered by the “dominant”
ecosystem could be duplicated (gathered though other means) by the competitors and, if so, at what
cost disadvantage, if any. The second question concerns the conditions (organization, format
(processed data or raw data), timing) under which the data must be made available to competitors.

In merger control cases among digital platforms competition authorities may have different types of
concerns with respect to data. First, they may fear that the merger will allow the acquirer to get access
to the data gathered and accumulated by the target, thus giving the merged firms an advantage which
could not be duplicated by competitors. Second, they may fear that the merged firms may degrade the
privacy protection of their offerings, thus degrading the quality of their service to the detriment of
consumers or by implicitly increasing the price of their service.

In its examination of the Google/DoubleClick and Facebook/WhatsApp mergers, the European

Commission (EC) has analysed in detail for each case the effects of the combination of the two sets of
the merging companies' data and whether these mergers could have led to the foreclosure of
competitors. The Commission, however, decided that these mergers would not create a competitive
advantage which could not be replicated by competitors of the merging firms.

In the case of the Facebook/WhatsApp merger, the Commission examined both whether the potential
data concentration was likely to strengthen Facebook's position in the online advertising market and
also its effects on the ‘zero priced’ markets for social networking and consumer communications
services.

With respect to the consumer communication services market, the Commission considered that the
merger was unlikely to produce unilateral effects on the social networking and consumer
communications market, as Facebook and WhatsApp were not close competitors and as consumers
tend to multi-home and would be capable of switching to competitive alternatives post-merger.

With respect to the online advertising market, It noted that only Facebook was active in the provision
of this service and that WhatsApp did not, previous to the merger, collect data about its users or store
messages.

It then analysed two main possible theories of harm.

First, it examined the possibility that Facebook could use WhatsApp as a potential source of user data
for the purpose of improving the targeting of Facebook's advertising activities outside WhatsApp. The

28
Case No COMP/C-3/37.792 – Microsoft, 24 March 2004.

22

Electronic copy available at: https://ssrn.com/abstract=3776274

Commission noted, however, that this would require Facebook, to match each user's WhatsApp profile
with her/his Facebook profile, provided she/he has both and that the Notifying Party had submitted
that there were major technical obstacles thereto. The Commission also noted that such a move on
the part of Facebook might prompt some users to switch to different consumer communications apps
that they perceive as less intrusive and that therefore Facebook might lack the incentive to engage in
such a move.

Second, the Commission also engaged in an ‘even if analysis’ exploring what would happen if the
merged entity integrated both parties’ datasets despite the technological difficulties and the risk of
losing consumers. It concluded that even in this case, there were a sufficient number of alternative
platforms providing online advertising space and collecting user data services (such as Google., Apple,
Amazon, eBay, Microsoft, AOL, Yahoo!, Twitter, IAC, LinkedIn, Adobe and Yelp, among others). Thus,
the Commission considered that the transaction did not give rise to serious doubts as to its
compatibility with the internal market as regards the market for the provision of online advertising
services.

A different and more radical approach to the problem of data based platform market power has been
proposed by Geoffrey Parker, Georgios Petropoulos and Marshall Van Alstyne ( 29), at least for
platforms which can be considered gatekeepers. They propose to impose an ex ante obligation which
would require “gatekeeper platforms to allow third party access to a user’s data upon that user’s
request”.They argue : “An in situ rights regime grants users all the benefits of data portability but
confers several additional benefits. Context is preserved rather than lost, as in the case of friends’ posts
that do not belong to a user. Data do not grow stale but rather include both stocks and flows of activity.
And data remain actionable such that one might reach a friend or make a purchase based on that data.
Giving users control of data where it resides allows them to invite third parties to compete to create
benefits with the host site, prompting greater sharing of value. Absent access to the infrastructure,
certain benefits cannot be created”.

Similar issues are being heatedly discussed in the EU in the context of the debate on connected cars.
Automobile manufacturers of these cars favor the “ extended car” concept whereby the data
generated by connected cars will be gathered on the external server of the car manufacturer and then
selectively passed on to interested business users rather than being directly accessible in real time on
the connected cars by competing providers of after sales or complementary services (such as insurance
companies or independent car repair shops).

An additional question (which we will address in discussing the European Ex Ante Regulatory Proposal
in the Digital Market Act) is whether the asymmetrical obligations imposed (either ex ante or ex post)
on large platforms to give access to their data could backfire by reducing their incentive to gather
data which could be used to increase the quality of the services to the benefit of users or consumers.

Privacy Issues

An interesting aspect of the EC Facebook/WhatsApp merger decision is the refusal of the Commission
to consider the potential effect of the merger on privacy protection issues of the data collection. The
Commission states in paragraph 164 of the decision that “(…) Any privacy-related concerns flowing
from the increased concentration of data within the control of Facebook as a result of the Transaction
do not fall within the scope of the EU competition law rules but within the scope of the EU data
protection rules”. It was on this point following the 2006 precedent of the November 2006 ECJ

29
Geoffrey Parker,Georgios Petropoulos Marshall Van Alstyne “Platform Mergers and Antitrust”, January 10,
2021, SSRN

23

Electronic copy available at: https://ssrn.com/abstract=3776274

judgment in the ASNEF-EQUIFAX and Administracion del Estado case(30) in which the Court stated that
“(….) any possible issues relating to the sensitivity of personal data are not, as such, a matter for
competition law, they may be resolved on the basis of the relevant provisions governing data
protection”.

However, there is no consensus within the EU on whether privacy-related concerns should be

considered to be outside the scope of competition law. In its Facebook decision(31), the
Bundeskartellamt noted that many users were not aware of the fact that under its terms of service
Facebook was able to collect user data from third party sources such as Facebook-owned services
(Instagram or WhatsApp) and third party websites including interfaces such as the “Like” or “Share”,
allocate these data to the users’ Facebook accounts and use them for numerous data processing
processes. The Bundeskartellamt observed that Facebook’s terms of service and the manner and
extent to which it collected and used data were in violation of the European data protection rules and
were an exploitative abuse of the dominant position of Facebook in the market for social networks.
This approach was based on the case-law of the Federal Court of Justice of Germany under which not
only excessive prices, but also inappropriate contractual terms and conditions constitute exploitative
abuses.

In the US the FTC also cleared the Facebook/WhatsApp merger in 2014 (32) without an in depth
investigation.

The Facebook/WhatsApp merger decisions, however, turned out to be controversial on both sides of
the Atlantic.

Indeed, in August 2016, after WhatsApp announced updates to its terms of service and privacy policy,
including the possibility of linking WhatsApp users' phone numbers with Facebook users' identities
contrary to what the parties had said, the Commission fined Facebook (but did not rescind its decision
to allow the merger based on the fact that it had established though the ‘even if ’ analysis that this link
would not lead a competition problem on the online advertising market).

This led to a controversy over whether, and if so why, the EC commission had missed a crucial
dimension of the case by refusing to look at the implications of the merger for privacy protection.

Thus, for example, Elias Deutscher (33) wrote : “Taking into consideration the economic role of privacy
would have enabled the Commission to accurately assess the merging parties’ incentive to integrate
their datasets post-merger. It would have revealed that the potential benefits arising from the
matching of Facebook’s and WhatsApp’s databases would have exceeded potential losses as a
consequence of consumer switching in response to a deterioration of WhatsApp’s privacy protection.
A privacy-related theory of harm would also have directed the focus of the Commission’s analysis of
potential anticompetitive effects from the online advertising to direct consumer harm in the form of
lower privacy protection on the consumer communications apps market.

30
European Court of Justice Case C-238/05 Asnef-Equifax ECLI:EU:C:2006:734 [63
31
See Bundeskartellamt Press Release: “Bundeskartellamt prohibits Facebook from combining user data from
different sources” 07.02.2019
32
Federal Trade Commission, Press release: “FTC Notifies Facebook, WhatsApp of Privacy Obligations in Light of
Proposed Acquisition”, April 10, 2014
33
Elias Deutscher, “How to measure privacy-related consumer harm in merger analysis? A critical
reassessment of the EU Commission’s merger control in data-driven markets”, EUI Working papers,
Department of Law, LAW 2018/13

24

Electronic copy available at: https://ssrn.com/abstract=3776274

While the Commission had concluded that the combination of databases would not cause any
competitive harm on the online advertising market, the Commission simply omitted to assess whether
this would entail any consumer harm on the consumer communications markets. (…) The
Facebook/WhatsApp merger thus shows that even though competition authorities already account for
the competitive relevance of personal data, the role of privacy or data protection remains an important
blind spot in their antitrust and merger analysis”.

This view was echoed by BEUC the European consumer organization which stated in its press release
of 18 May 2017 (34) “It is very disappointing that the Commission decided not to revise its original
decision on the Facebook merger with WhatsApp. It is crucial in our data economy that competition
bodies more closely scrutinise the potential consumer harm of a merger between data-heavy
companies. The Commission failed to do so when it gave the go-ahead to the Facebook/WhatsApp
takeover”.

Elias Deutscher contends that: “The EU Commission‘s failure to assess and identify the potential
consumer harm arising from the matching of Facebook’s and WhatsApp’s databases can be explained
by the fact that antitrust scholarship and enforcers have so far fallen short of putting forward a
coherent theory of privacy-related consumer harm.(….) thus far, commentators and antitrust
authorities have not assessed to what extent the accumulation of personal data not only harms
competitors and advertising customers, but might actually also have immediate negative effects on
consumers on the ‘free’ user side of online platforms” He attributes this failure to “the assumption
that data protection and efficiency or consumer welfare, as predominant goals of antitrust, constitute
incommensurable values which cannot be adequately balanced within the framework of antitrust
analysis” a view which has been stated by the US FTC, the EU Commission and the Court of Justice of
the European Union (35). Even if he acknowledges that competition authorities lack adequate tools and
methodologies to account for privacy as a non-price dimension of product quality, he suggests as an
alternative to consider privacy as a non-monetary price paid by consumers of free services.

In the US, on the other hand, in September 2020, forty eight state attorney generals brought an action
to redress violations of Section 2 of the Sherman Act, 15 U.S.C. § 2, and Section 7 of the Clayton Act,
15 U.S.C. § 18, against Facebook arguing, among other things, that following the acquisition, of
WhatsApp, Facebook degraded Instagram users’ privacy by matching Instagram and Facebook Blue
accounts so that Facebook could use information that users had shared with Facebook Blue to serve
ads to those users on Instagram. The complaints argues that:

“(179) Thus, Facebook Blue users who had declined to give their phone numbers to Facebook suddenly
found their phone numbers connected to their Facebook Blue accounts anyway. Facebook was able to
use that additional data in its recommended friend (“People You May Know”) ranking, leading to
growth of its social graph. This harm to users’ privacy resulted from Facebook’s acquisition of
WhatsApp

34
BEUC Press release 18 May 2017 Facebook/WhatsApp merger: BEUC welcomes fine but regrets failure to
tackle consumer data concerns accessed on January 1st 2021 at
https://www.beuc.eu/publications/facebookwhatsapp-merger-beuc-welcomes-fine-regrets-failure-tackle-
consumer-data/html

35
European Court of Justice Case C-238/05 Asnef-Equifax ECLI:EU:C:2006:734 [63]. Case COMP/M.4731
Google/DoubleClick (n 27) [368]. Statement of Federal Trade Commission Concerning Google/DoubleClick (n
30) 2. Case COMP/M.7217 Facebook/Whatsapp (n 2) [164].

25

Electronic copy available at: https://ssrn.com/abstract=3776274

180. Facebook’s acquisition of WhatsApp thus substantially lessened competition and further
entrenched Facebook’s monopoly power in the Personal Social Networking Services market.
Moreover, Facebook’s subsequent degradation of the acquired firm’s privacy features reduced
consumer choice by eliminating a viable, competitive, privacy-focused option”.

Time will tell whether the attorney generals are eventually successful in pushing a third approach to
allow consideration of privacy within the context of antitrust or competition law (degradation of
privacy as a reduction of consumer choice).

What is clear at this point is that there is considerable controversy on how data and privacy issues
could be integrated into the competition assessment of ecosystems and whether they should be. There
is strong resistance in many jurisdictions to using competition law to deal with privacy issues but as
there are cases pending in several jurisdictions, time will tell in which direction the courts will go.

Competition issues within ecosystems

In transaction ecosystems and in multi-service or multi products ecosystems, within ecosystem ( i.e.
competition issues arising among the participants of an ecosystem) competition issues may also arise
particularly when the platform is competing with some of the business users of the platform or some
the providers of complementary services. The core platform of the ecosystem may refuse access to
applications that competes with its own applications or, if access of third party applications is allowed,
the platform may discriminate in favor of its own services to the detriment of the services offered by
independent providers of complementary services or by business users of the platform. Furthermore
the platform may use some of the information it receives through the activity of independent
businesses using the platform to piggy back on their successful offers.

In those cases, competition authorities tend to focus on the fact that the central platform of the
investigated ecosystem has a dominant position on a service and on the fact that the platform abused
its market power either by restricting access to the market for this application or a complementary
application or by restricting the ability of its competitors to compete on these markets.

In the 2017 Google shopping case(36), where Google was fined € 2.42 Billion, the European Commission
found that Google dominated the market for general algorithmic online search in national markets
throughout the EEA and that it leveraged this dominance into the market for comparison shopping
services by favoring its own shopping comparison service, Google Shopping, in general search results
over third-party comparison shopping services. The Commission considered that Google
systematically gave prominent placement to its comparison shopping service at or near the top of its
general search results, while demoting rival comparison shopping services. As a result, the comparison
shopping service of Google was much more visible to consumers in Google’s search results than those
of its competitors, which led consumers to click more often on Google Shopping. Given the importance
of Google’s search engine as a source of traffic, Google’s preferential treatment of its own comparison
shopping service resulted in a substantial and durable loss of traffic for its competitors’ competing
shopping services. Google was thus found to have infringed its special responsibility as a dominant
firm, by implementing an unjustified practice capable of foreclosing competition in the comparison
shopping market.

The EC Decision orders Google to comply with the principle of giving equal treatment to rival
comparison shopping services and its own offering.

36
CASE AT.39740 Google Search (Shopping) 27/06/2017

26

Electronic copy available at: https://ssrn.com/abstract=3776274

On 10 November 2020, the EU Commission charged Amazon with violating competition law, alleging
that the company used nonpublic data it gathers from third-party sellers to unfairly compete against
them.

There are numerous other examples of this nature in many jurisdictions.

As an example, Russia has dealt with three cases of this nature.

In the Russian Yandex-Google case(37), the Federal Antimonopoly Service (FAS) concluded that Google
had a dominant position in the market of pre-installed app stores for smartphones running on Android
operating system (OS), based on the facts that that on the one hand smartphone manufacturers using
the Android operating System represented more than 50% of the market for mobile phones in Russia
since 2012 and, on the other hand the Google Play Store which allowed the downloading of application
was hugely popular and therefore pre-installed on nearly all mobile phones using the Android
operating system. Similarly, in its case against Apple, it concluded that Apple had a dominant position
in the distribution market for iOS applications.

On 18 February 2015, Yandex lodged a complaint against Google on the basis of a violation of Russian
competition law. In its complaint Yandex argued that Google conditioned the pre-installation of its Play
Store application to the purchase of the entire Google Mobile Service application package, the
compulsory pre-installation of Google search as a default search engine, the imposition of a
preferential placement of Google applications on device screens and the prohibition of pre-installation
of Google competitors’ applications. The FAS noted that there were no technological reasons requiring
those restrictions which limited Google competitors from accessing the GMS apps market and could
lead to their exclusion from this market due to a combination of factors such as a guaranteed presence
of the Google apps on a large number of devices, the high usage frequency of the Google apps and
users’ “passivity”(38).

Altogether it was found that that Google, which has a dominant position on the market of pre-installed
app stores for Android OS localized for distribution on the territory of the Russian Federation, had
restricted competition on adjacent product markets (app stores), and was in violation of Part 1 Article
10 of the Law on Protection of Competition. The case was appealed to the Court of Moscow, which
affirmed the FAS decision.

On 10 August 2020 the FAS completed its investigation of Apple (39) following the complaint from
Kaspersky Lab, (a developer of antivirus programs and parental control applications). The FAS
concluded that Apple had abused its dominant position in the market for distributing mobile apps on
the iOS operating system for having, in the fall of 2018, restricted the tools and capabilities of third-
party parental control applications, as a result of which such applications lost some of the important
functionality while at the same time, having pre-installed a Screen time app which has a parental
control functionality, thanks to the use of iOS technology capabilities which are not available to third-

37
See FAS Yandex vs Google 27.06.2016 accessed on January 5th 2021
http://en.fas.gov.ru/documents/documentdetails.html?id=14677
38
The use of consumer biases to establish the possibility of foreclosure of the independent application
producers which was used in the Google Yandex case and in a number of these cases outside of Russia is
discussed in the next section
39
See FAS Russia Ruling and remedies on the case No.110110-242019 against Apple Inc. (abuse of dominant
position in the market for distributing mobile apps on the iOS operating system) 11.09.2020 accessed at
http://en.fas.gov.ru/documents/documentdetails.html?id=15363 on January 5 2021

27

Electronic copy available at: https://ssrn.com/abstract=3776274

party developers. In addition, FAS found that Apple reserved the right to reject any third-party
application from its App Store, even if it met all the company's requirements.

FAS imposed remedies on Apple requesting it to remove from its documentation provisions that give
it the right to reject third-party apps in the App Store even if they meet all the requirements, to ensure
that in-house apps do not take precedence over third-party apps, and that developers of parental
control apps can distribute apps to the App Store without loss of the important functionality.The
remedy also requires Apple Inc. to ensure that in-house apps do not take precedence over third-party
applications.

In 2017, following a complaint by Kaspersky Lab, the FAS investigated whether Microsoft had favored
its own antivirus application and encouraged users to abandon third-party antivirus applications.

The FAS of Russia considered that Microsoft had a dominant position on the market for operating
systems for stationary devices (computers and laptops) and on the market for the software to allow
the connectivity and interdependence of the application to its operating system. Microsoft was alleged
to have pre-installed its antivirus software on Windows 10 OS, called Windows defender, to have
deleted antivirus software which it considered to be incompatible with Windows 10 and to have
abusively limited the ability of independent antitrust software to ensure the compatibility of their
software with Windows 10, thereby effectively "squeezing out" third-party antivirus application
developers, while actively promoting its own antivirus program built into the operating system.

As a result of this case Microsoft significantly changed its company policy. Microsoft modified the
"Antimalware Platform Requirements" that govern the relationship between the corporation and
independent antivirus software vendors and eliminated all calls for the abandonment of third-party
software.

On 9 November 2020 the Competition Commission of India opened an in-depth investigation into the
claims of whether Google prominently promotes Google Pay during the setup of an Android
smartphone and if Play Store’s billing system is designed “to the disadvantage of (…) apps facilitating
payment through UPI, as well as users.”

Another type of case aims at practices designed to protect the position of a key service of a platform.
The platforms may want to limit the strategic freedom of business users or of independent providers
of complementary services active on their platform in order to protect themselves against the
potential competition of other ecosystems or platforms.

For example in Europe, in March 2019 the European Commission fined Google (€ 1.49 billion) for
having abused its dominant position on the market for online search advertising intermediation
services. Google’s online search advertising intermediation service AdSense, which has a 70% market
share in the EEA, is integrated free of charge by a vast array of websites seeking to monetise their
content by earning revenue through engagement with advertising facilitated by AdSense. The service
includes traditional text/hyperlink-style ads in addition to images, animated graphics, and even
imbedded videos (for example Google subsidiary YouTube monetizes its content by displaying video
ads through AdSense). The European Commission found that Google has abused its dominant position
on this market by imposing restrictions on third party websites using the AdSense such as forcing them
to exclusively use AdSense for advertising, or to reserve premium placement for Google search, and
take a minimum number of AdSense ads. The Commission found that such practices, which did not
have any objective justification were designed to exclude Google’s competitors (such as Microsoft and
Yahoo) from the market for online search advertising.

28

Electronic copy available at: https://ssrn.com/abstract=3776274

Those cases raise several perplexing issues.

First, in most of these cases, each application is considered to belong to a specific market and the
question raised is whether competition on this specific market has been restricted.

For example, in the FAS Russia Microsoft case the FAS considered the restriction of competition on the
antivirus application market for operating systems for stationary devices. Similarly, in the EU Google
Shopping case the EU Commission was concerned with the market for comparison shopping services
and found that Google inflicted durable loss of traffic for its competitors shopping competing services
etc…. In other words, the application is not considered in its relationship with the other applications
offered on the platform.

Yet we know from the business policy literature that the choice of the complementary services offered
on a platform is a crucial choice for the ecosystem both in order to establish its identity and to increase
its value for users in the competition among ecosystems. Therefore the value of one particular service
to an ecosystem may be much more important than the value of the service if it were offered on a
stand alone basis.

If we assume that the platform has a choice between several applications which provide the same kind
of service whose value would, thanks to the complementarity with the other services offered on the
platform, increase the value of the platform over and beyond the value of the service offered on a
stand alone basis, it is difficult to understand why a platform would not choose the application which
best matches the other applications and services offered on the platform thus maximizing the overall
value of the ecosystem. For example if we go back to the FAS Russia Microsoft case, it would seem that
in order to maximize the value of its ecosystem, Microsoft would have an incentive to choose the best
anti-virus available irrespective of whether or not it is its own anti-virus. Similarly, Google has made
the point that Google search is better than other search engines not because it is pre-installed but that
it is pre-installed because it is a superior search engine to those of its competitors.

Thus, first, the choice of which app will be chosen for pre-installation on a platform will have
consequences not only on competition between the applications considered but also on competition
between the ecosystem it participates in and other ecosystems. Therefore, the consequences of the
choice of an application made should be considered, by competition authorities, from the standpoint
of competition on the application market and competition among ecosystems. Separating the two
analyses does not adequately reflect all of the competition implications of the choice of a particular
application.

Second, when an application that provides a service for users is denied access to a platform (or is a
victim of the self-preferencing of the platform for its own application), it is only if we assume that the
possibility to pre-install the competing application (and to have the platform treat this competing
application in a non-discriminatory way) would either increase the quality of the platform application
or the quality of the competing application over and beyond the initial quality of the platform
application that there would be the possibility of a consumer welfare gain from increasing the
competition for particular applications in the ecosystem. This possibility, however, cannot be assumed
to exist in all cases. Indeed when the quality of an app is related to the intensity of its use (for example
through the better training of artificial intelligence algorithms) there is a possibility that the quality of
both the platform application and the competing application(s) on the same platform may be
decreased by the fact that they compete for users on an equal footing ( and that each one has fewer
users than would be the case if there was only one application pre-installed).

29

Electronic copy available at: https://ssrn.com/abstract=3776274

Thus, a case by case analysis of the quality dimensions of the applications considered and their
relationship with the quantity of data collected by the platform is warranted.

Third, the usefulness of competition between applications on a platform (assuming they have equal
access to the platform and are treated in a non discriminatory way) will very much depend on whether
or not users can make and are actively making competition work between these platforms. In other
words it will depend on the desire and ability of consumers to reward the providers of best quality
applications(40). If, as in the Russian Yandex-Google case consumers are considered to be passive or
inert, increasing their choices may not result in an increase in their welfare. We come back to this issue
later on.

Competition between ecosystems

When we consider competition among multiproduct ecosystems such as Apple or Google the standard
analyses used in antitrust become even more difficult to use to assess competition issues among
ecosystems. A number of merger cases are relevant to this issue.

In the digital markets, large platforms have engaged in a buying spree often acquiring digital start-up
firms with very low turnover at very high prices (41). A number of those firms were platforms with one
or a few innovative functionalities using a specific algorithm which after the merger were integrated
into the ecosystem of the acquiring platform. A few turned out to be enormously successful for their
acquirers such as the acquisition by Facebook of Instagram (a photo application with 30 million
followers) in 2012 for 1 billon U.S. dollars, followed by the acquisition of WhatsApp (with 450 million
followers) in 2014 for 19 Billon U.S. dollars. Today Instagram has more than 1 billion followers and
Facebook has more than 2 billion followers.

From the competition standpoint, the concern about mergers in the digital sector is based on the idea
that the acquired start-up firms can provide competitive pressure because they are able to siphon off
or “cream skim” customers and collect valuable data. Thus the acquisition of start-up companies by
large firms may benefit the incumbent by reducing competitive pressure of potential entrants on the
periphery of the market or by preventing future entry and expansion by such start-up firms that could
undermine the incumbent’s dominance.

Nevertheless, an important question for competition authorities is how to analyze such mergers.

It should be recalled, for example, that the Facebook Instagram merger was investigated by the OFT in
2012. In its decision the OFT stated :

“14. The OFT considered two unilateral effects theories of harm: actual competition in the supply of
photo apps and potential competition in the supply of social network services.

40
The ability of consumers to make competition work between platforms may depend, for example, on
whether or not data portability is possible; the desire of consumers to actively make competition work may
depend on whether they have a behavioral bias such as, for example, a status quo bias
41
Google has acquired around 200 companies since 2000, including Android, YouTube, and Waze and Tenor
(2018).
Microsoft has acquired more than 100 companies in the last ten years, including Skype, Nokia Devices,
LinkedIn (2017) and GitHub.
Amazon has also acquired more than 100 companies, including Whole Foods (2017).
Facebook has acquired ninety companies, mainly start-ups; in May 2019, Tim Cook, declared in an NBC
program that his firm bought on average one company every two to three weeks including Surreal Vision (
2015) and Shazzam (2017).

30

Electronic copy available at: https://ssrn.com/abstract=3776274

15. Instagram allows users to take photos, apply digital filters to those photos, and then share those
photos on the Instagram network or via other social networks. Facebook launched its mobile photo app
in May 2012, weeks after it had announced that it would acquire Instagram. Facebook’s app has similar
functionality to Instagram’s. It allows users to apply filters, tag photos, comment on photos, and post
the photos to Facebook.

(…)

21. To conclude, there are several relatively strong competitors to Instagram in the supply of camera
and photo editing apps, and those competitors appear at present to be a stronger constraint on
Instagram than Facebook’s new app. The majority of third parties did not believe that photo apps are
attractive to advertisers on a stand-alone basis, but that they are complementary to social networks.
The OFT therefore does not believe that the transaction gives rise to a realistic prospect of a substantial
lessening of competition in the supply of photo apps.

(…) In summary, the evidence before the OFT does not show that Instagram would be particularly well
placed to compete against Facebook in the short run. In addition, there are other firms that appear to
be presently able to compete against Facebook for brand advertising. For these reasons, the OFT
believes that there is no realistic prospect that the merger may result in a substantial lessening of
competition in the supply of display advertising”.

Simultaneously in its letter to Facebook on 22 August 2012, the FTC stated that “upon further review
of this matter, it now appears that no further action is warranted by the Commission at this time.
Accordingly, the investigation has been closed”.

However, less than five years after the merger, Instagram had more than 500 million active users, and
had become the second most popular social media network in the world, behind only Facebook (You
Tube excluded) , and the number of advertisers on Instagram had grown to more than one million, in
part due to the fact that Instagram was then considered to be the best social media platform for
customer engagement.

It is thus interesting to mention the fact that the FTC has decided in 2020 to revisit the issue of the
acquisition of Instagram by Facebook.

In its 2020 complaint , the FTC writes:

“12. Facebook initially tried to compete with Instagram on the merits by improving its own mobile
photo-sharing features. But by September 2011, Mr. Zuckerberg saw that Facebook had fallen far
behind, writing internally: “In the time it has taken us to get ou[r] act together on this[,] Instagram
has become a large and viable competitor to us on mobile photos, which will increasingly be the future
of photos.”

13. (…) In February 2012, Mr. Zuckerberg acknowledged that if left independent—or in the hands of
another acquirer like Google or Apple—Instagram threatened to leave Facebook Blue “very behind in
both functionality and brand on how one of the core use cases of Facebook will evolve in the mobile
world.” Emphasizing that this was a “really scary” outcome for Facebook, Mr. Zuckerberg suggested
“we might want to consider paying a lot of money” for Instagram.

14. Mr. Zuckerberg recognized that by acquiring and controlling Instagram, Facebook would not only
squelch the direct threat that Instagram posed, but also significantly hinder another firm from using
photo-sharing on mobile phones to gain popularity as a provider of personal social networking. As Mr.
Zuckerberg explained to Chief Financial Officer David Ebersman in an email, controlling Instagram

31

Electronic copy available at: https://ssrn.com/abstract=3776274

would secure Facebook’s enduring dominance around one of the few social mechanics that could
provide a footing for a competing personal social networking provider:

[T]here are network effects around social products and a finite number of different social mechanics to
invent. Once someone wins at a specific mechanic, it’s difficult for others to supplant them without
doing something different. It’s possible someone beats Instagram by building something that is better
to the point that they get network migration, but this is harder as long as Instagram keeps running as
a product. [Integrating Instagram’s functions into Facebook] is also a factor but in reality we already
know these companies’ social dynamics and will integrate them over the next 12-24 months anyway.
The integration plan involves building their mechanics into our products rather than directly integrating
their products if that makes sense. . . . [O]ne way of looking at this is that what we’re really buying is
time. Even if some new competitors spring[] up, buying Instagram, Path, Foursquare, etc now will give
us a year or more to integrate their dynamics before anyone can get close to their scale again. Within
that time, if we incorporate the social mechanics they were using, those new products won’t get much
traction since we’ll already have their mechanics deployed at scale.

15. On April 9, 2012—the day Facebook announced it was acquiring Instagram—Mr. Zuckerberg wrote
privately to a colleague to celebrate suppressing the threat: “I remember your internal post about how
Instagram was our threat and not Google+. You were basically right. One thing about startups though
is you can often acquire them.”

Three things are striking in this excerpt from the FTC complaint.

The first one is the acknowledgment by Mr Zuckerberg that there are a “number of different social
mechanics to invent”. This confirms that ecosystems compete by offering differentiated combinations
of services which are complementary in an effort to maximize the value to users of the interaction
between the various sides of their core platform rather than by competition head to head offering the
same combination of services. Mr Zuckerberg saw early on (as early as 2011) that photo sharing apps
could become an essential element of a combination of services which would make a social media
platform successful for mobile users. Having tried to develop such an app but recognizing that
Instagram was superior to its own photo app, the Facebook founder decided to acquire Instagram to
integrate its dynamic by offering a photo application based social media for mobile users. The
acquisition made sense in the context of an evolving dynamic competition among ecosystems in which
a new offering integrating an element which hitherto had not been a crucial part of social media
offerings could become highly successful given the migration of a large part of social media users from
personal computers to mobile phones. Furthermore, being the first to build a social media offering for
mobile users offering an already developed but not yet widely distributed photo sharing application,
would, allow Facebook to benefit first (and more effectively) from the network effects and give it an
advantage over potential competitors.

Mr Zuckergerg suggested that, without the acquisition, Facebook could be displaced and Instagram
including the functionalities of a photo sharing application and of communication among its users
could become the dominant social media. It interesting to note that this comment implies that, as
previously mentioned, competition among ecosystems is for the market rather than for a market share
of the market. Thus, in this digital world of ecosystems dominance is not an indicator of lack of
competition but rather the result of it but dominant forms are threatened by other ecosystems able
to offer a more attractive combination of communication services to users.

The second striking element of this excerpt is that it helps us understand why the traditional tools of
merger control used by competition authorities are inadapted to grasping the competitive implications
of such a merger. Typically, competition authorities think in terms of existing markets narrowly defined

32

Electronic copy available at: https://ssrn.com/abstract=3776274

at the time of the transaction and ask themselves whether competition on these markets might be
diminished through a unilateral or coordinated effect. Thus the OFT looked at the social media market
as it was (and as it was assumed to remain) at the time of the merger and at the photo app market as
it was at the time of the merger and concluded that there were competitors to Instagram on the
market for photo applications and that there were competitors to Facebook on the market for social
media. It also concluded that a digital photo application in itself was unlikely to become a social media
in the short run. What the OFT missed (because it was thinking in terms of restriction of competition
on existing markets) was the fact that the acquisition was the sign of possible transformation of the
business model of social media with the emergence of an innovative social media product based on a
combination of a photo application functionality and other communication functionalities among
users.

The third striking element is the fact that there was a debate within the Facebook management about
who was the most important competitor to Facebook: Google + or Instagram. Thus within the
company, understanding where the competitive threat was coming form was open to debate.

The US FTC, after conducting a nonpublic investigation of the US$ 1 billion Facebook Instagram merger
in August 2012, cleared it at the time (while it decided to challenge it in 2020) possibly because, at the
time, it feared that it would not prevail in court. Indeed, the Supreme Court of the United States in the
Marine Bancorporation case of 1974 (42), established that “two preconditions must exist” before an
actual potential competition theory “establishes a violation of § 7 of the Clayton Act. The first condition
is that the potential competitor could enter the market at issue absent the merger; the second
condition is that such entry would produce a likelihood of deconcentration or other significant
procompetitive effects” (43). Moreover, with respect to the first condition, the Court implied that
“unequivocal proof” of actual future de novo entry is required. Arguing on the one hand that a
photo sharing application could become a social media and that therefore the merger was restricting
potential competition (at a time when no such move from the photo sharing application market to the
social media market had ever taken place) might have been difficult given the high standard of proof
required by US jurisprudence. Alternatively, arguing that Facebook and Instagram were actual
competitors on the market for the attention of users of social communication services and that their
merger would significantly alter the structure of this market would have been equally difficult. Finally,
convincing a judge that the merger was anticompetitive even though Instagram and Facebook would
still be free for users after the merger, as they were before the merger, might have been equally
difficult as US judges are used to thinking that transactions should be opposed if and only if they are
likely to raise prices and thus reduce consumer surplus.

Similarly (44), in 2014, the US FTC cleared Facebook’s $19 billion acquisition of the messaging
application WhatsApp, though the FTC did send both companies a letter reminding them of their
obligation to maintain privacy practices in accordance with the WhatsApp user agreement in place at
the time that user data was collected.

There is something disturbing about the fact that the US FTC is now trying to challenge a merger it
cleared by arguing, inter alia, that at the time it took place the goal and the effect of the merger was
to eliminate the main source of competition that Facebook faced and to deprive its competitors from

42
United States v. Marine Bancorporation, Inc., 418 U.S. 602 (1974)
43
See Mark Glick, Catherine Ruetschlin & Darren Bush , BIG TECH’S BUYING SPREE AND THE FAILED IDEOLOGY
OF COMPETITION LAW , Forthcoming : Hastings Law Journal

44
FTC, FTC Notifies Facebook, WhatsApp of Privacy Obligations in Light of Proposed Acquisition, April 10, 2014

33

Electronic copy available at: https://ssrn.com/abstract=3776274

the ability to challenge its dominance(45). If the FTC is right today, it must be that its analysis was wrong
at the time of the merger.

Overall, the static analysis of relevant markets, the narrowness of the interpretation of the consumer
welfare standard, the high burden of proof when it comes to the risk that potential competition may
be lessened and the difficulty to integrate into competition analysis the inherent instability of business
models of digital ecosystems make merger control as it is traditionally implemented an ineffective tool
to identify problematic mergers in the digital world.

If our traditional tools of antitrust and merger control need to be modified when it comes to assessing
competition among digital ecosystems, this is not to say that the underlying goal of competition, i.e.
the protection of consumer welfare, is inadequate. There is no argument that consumers should not
be protected against possible abuses of market power by large digital ecosystems. But we need to be
able to reliably assess how consumer welfare can be defined for the users of the platforms who are
offered services at a zero price and we must be able to distinguish between pro and anticompetitive
behaviors or transactions. Unfortunately the competition analysis tools which work satisfactorily for
the rest of the economy do not perform well in the digital world.

Consumers status quo bias

Two European cases against Google, the 2017 Google shopping case where Google was fined € 2.42
billion and the 2018 Google Android case where Google was fined €4.34 billion, and the Google, share
a common feature.

In the Google Shopping case the European Commission alleged, first that (216) “Contrary to what
Google claims, there is (…) limited substitutability between comparison shopping services and
merchant platforms, such as Amazon Marketplace and eBay Marketplaces”, and second, that “Google
systematically gave a prominent placement to its own comparison shopping service and demoted rival
comparison shopping services in its search results”.

According to the Commission, by giving prominent placement to its own comparison shopping service
and by demoting competitors, Google has given its own comparison shopping service a significant
advantage compared to rivals because: “The evidence shows that consumers click far more often on
results that are more visible, i.e. the results appearing higher up in Google's search results”. Google’s
practice was thus found to be an anticompetitive abuse of its dominant position in online search on
the ancillary online comparison shopping market. Thus a behavioral bias of consumers was alleged to
explain why the behavior of Google was anticompetitive.

Similarly, in the EU Google Android Case, the Commission considered, first, that Google held a
dominant position in the worldwide market (excluding China) for the licensing of smart mobile
operating systems and, second that the tying of the Google Search app and Google Chrome with the
Play Store resulting in the fact both were pre-installed and impossible to uninstall on mobile phones
which had installed the Play Store were capable of restricting competition because they provided
Google with a significant competitive advantage that competing general search services providers and

45
In the US, on Feb. 11, 2020, The Federal Trade Commission ordered big tech companies to provide detailed
information about their acquisitions of fledgling firms over the past 10 years, seeking to determine whether the
deals harmed competition. On Dec. 9, 2020, the US FTC and 46 states sued Facebook, claiming that Facebook
had unlawfully maintained monopoly power in personal social networking services in part by acquiring
potential competitors before they grew into larger threats and seeking to unwind Facebook’s acquisitions of
Instagram and WhatsApp.

34

Electronic copy available at: https://ssrn.com/abstract=3776274

competing non OS-specific mobile web browsers could not offset. To justify this competitive
advantage, the Commission argues (in paragraphs781 and 782) that:

(781) “ The reason why pre-installation, like default setting or premium placement, can increase
significantly on a lasting basis the usage of the service provided by an app is that users that find apps
pre-installed and presented to them on their smart mobile devices are likely to "stick" to those apps.
HP described the creation of a "status quo bias" in the form of premium placement and default setting
as follows: "Premium placement and default settings give applications and services located in those
positions the advantage of being the first things users see when they start to interact with their device.
Users are more likely to try these applications/services based on their prominent visibility and once they
are using them, they usually continue to do so. It is an easy way to obtain new users and deliver almost
automatic stickiness for an application or service.”

(782) Users are unlikely to look for, download, and use alternative apps, at least when the app that is
pre-installed, premium placed and/or set as default already delivers the required functionality to a
satisfactory level”. Thus again, a behavioural bias of consumers is alleged to explain why the practice
is anticompetitive.

These two cases follow in the footsteps of the EU Microsoft decision of 2007 where Microsoft was
fined € 492 million by the EU Commission for two abuses of dominant position, one of which was
Microsoft’s tying of Windows Media Player with Windows, its operating system. In its decision, the
Commission stated that : “(979) Through tying WMP with Windows, Microsoft uses Windows as a
distribution channel to anti-competitively ensure for itself a significant competition advantage in the
media player market. Competitors, due to Microsoft’s tying, are a priori at a disadvantage irrespective
of whether their products are potentially more attractive on the merits” and “(981) This shields
Microsoft from effective competition from potentially more efficient media player vendors which could
challenge its position. Microsoft thus reduces the talent and capital invested in innovation of media
players, not least its own and anti-competitively raises barriers to market entry. Microsoft’s conduct
affects a market which could be a hotbed for new and exciting products springing forth in a climate of
undistorted competition”.

Even though there is no reference to a behavioral bias on the part of consumers to explain why the
tying gave Microsoft a significant competitive advantage, it is implicitly clear that this significant
competitive advantage for Microsoft exists, in spite of the alleged superiority of other media players,
if and only if Windows consumers just use the media player which is installed on their computer and
do not bother to look for a “better” media player.

In the US, on 20 October 2020, the US Justice Department (DOJ) filed an antitrust lawsuit against
Google, alleging it uses anticompetitive tactics to preserve a monopoly for its search engine and related
advertising business. On Thursday 17 December 2020 a group of 38 US states and territories sued
Google claiming that it illegally maintained a monopoly in general search and search advertising
through anticompetitive conduct and contracts. Colorado Attorney General Phil Weiser indicated that
the group would file a motion to consolidate the case with the Justice Department’s recent lawsuit
against Google.

The DOJ lawsuit argues that Google has entered into a series of exclusionary agreements that require
that Google Search be set as the preset default general search engine on billions of mobile devices and
computers worldwide and, in many cases, prohibit pre-installation of a competitor search engine.
Google’s pre-installation agreements condition the distribution of Google Play and GPS to the
distribution of Google Search. According to the DoJ Complaint this tie reinforces Google’s monopolies
by foreclosing distribution opportunities to rival general search engines, thereby protecting Google’s

35

Electronic copy available at: https://ssrn.com/abstract=3776274

monopolies. Thus the DoJ Google complaint is fairly similar to the EU Google Android case. However,
the DOJ complaint does not explain why it believes that the pre-installation agreements have
foreclosed distribution opportunities to rival search engines given the fact that these rival search
engines can be easily installed on mobile devices through Google Play.

These cases raise a number of troubling issues.

In order to find an abuse on the part of the investigated firms, the EC Commission has to rely on the
behavioral economic literature and argue a) that when it comes to the digital products considered
there is a failure of consumers to rationally choose what is in their best interest and b) that the practice
of the firm investigated was a deliberate attempt by the dominant firm to exploit this demand side
market failure to the detriment of its competitors and ultimately of consumers. But there is no
explanation of what causes this demand failure or whether the consumer biases are exogenous
(observable whatever the characteristics of the competing products) or endogenous to the variety of
products available (observable only when the competing products are so similar as to be
indistinguishable).

It could be that consumers do not have the necessary information about what their options are.
Consumer protection laws or privacy laws could help remedy this type of situation. But alternative
explanations are also possible. For example, it could also be that consumers do not think that the
minor qualitative differences between the products offered are sufficient for them to bear the cost of
the search of an alternative.

If that is the case, it is difficult to see how the alleged violations decreased the welfare of consumers
who chose not to look for an alternative to the pre-installed product in spite of the fact that this
alternative was readily available or to understand how forcing them to exercise a choice that they do
not feel they need to make would increase their welfare. It is equally difficult to see why suppliers of
these alternative products would invest more in minor innovations if they had better possibilities of
access to consumers if consumers did not value minor innovations in these products (and therefore
did not base their choice on the technical differences between these products). Finally, it is not
established that competitors of the dominant firms would be deterred by the pre-installation from
investing in R&D for major innovations if such major innovations changed the attitude of consumers
with respect to the benefit they could derive from seeking an alternative to the pre-installed product.

In addition, the lack analysis of the importance of the assumed biases of consumers is an obstacle to
determining whether such biases could result in a foreclosure of applications which are not pre-
installed or are given preferential treatment. Indeed, invoking a status quo bias of consumers is
insufficient if we do not know how important this bias is. Does it result in the fact that consumers never
compare the alternatives they have? Does it result in the fact that consumers do not compare
alternatives as often as they would if they were perfectly rational but still compare sufficiently
frequently for the pressure of their choices to be felt by the suppliers of digital services? Those
empirical questions need to be answered to get a sense both of the reality of the possible
anticompetitive effect of the practices and of their severity.

Thus understanding the reasons for the demand side failure as well as the nature and the severity of
such failures are crucial both for the qualification of the practice if one believes that the goal of
competition is to protect consumer welfare and for the design of remedies, as the Commission found
out in the Microsoft case by forcing Microsoft to offer consumers the option of a choice of media
players which was a resounding failure.

36

Electronic copy available at: https://ssrn.com/abstract=3776274

Gatekeepers

To tackle some of the challenges of competition law enforcement in the digital sector on 15 December
2020 the European Commission unveiled a proposal for a regulation on contestable and fair markets
in the digital sector (Digital Markets Act). This proposal is an attempt to complement its existing
enforcement powers which it considers are neither sufficient to address some of the competition
concerns raised in digital markets nor have sufficient deterrent effect on the largest digital players.

One important part of the proposed regulation deals with the within ecosystem relationship between
some platforms (integrated or not) and third parties suppliers of complementary services.

The regulation would impose a number of constraints both in terms of positive obligations and in terms
of prohibited practices on “ Gatekeepers”, defined as providers of “core platform services” (online
search engines, online intermediation services, online social networking services, video-sharing
platforms, operating systems, interpersonal communication services, cloud computing, and
advertising) that: (i) have a significant impact on the internal market; (ii) serve an important gateway
for business users to reach end users; and (iii) enjoys an ‘entrenched and durable position’ either at
present or foreseeably ‘in the near future.’ Threshold levels with respect to the turnover of these
platforms, the number of EU countries they serve, their number of end users and number of business
users are proposed to establish the rebuttable presumption that a platform qualifies as a “gatekeeper”.

The obligations are designed to protect the businesses using the platform either to provide
complementary services to the core services of the platform or as a market place against the possibility
that the gatekeeper will use their data to outcompete them or will self preference its own services to
the detriment of the services these businesses offer by prohibiting the gatekeeper from discriminating,
or imposing business restriction on the users of the platform and by facilitating the transferability of
data and the interoperability of their services with the services offered by independent businesses.

The obligations imposed on the gatekeepers would be either obligations that would apply to them
irrespective of their specificity or obligations which would be ‘susceptible to be further specified’ on a
case by case basis.

The obligations would prevent the gatekeepers from combining personal data from their core platform
services with data from other sources (including other services offered by the gatekeeper) or from
restricting business users from contracting with end users outside of the gatekeepers’ eco-systems or
from requiring business users of the platform to use, offer or interoperate with any identification
service of the gatekeeper in the context of providing its services via the relevant gatekeeper’s core
platform services. Other self executing obligations would protect the advertisers and publishers to
which gatekeepers provide advertising services by imposing transparency obligation on the gatekeeper
with regard to the price paid by the advertiser and publisher, as well as the amount of remuneration
paid to the publisher, for the publishing of a given ad and for each of the relevant advertising services
provided by the gatekeeper.

The second category of obligations ‘susceptible to be further specified’ would include an obligation
not to use data acquired by the platform in relation to business users to then compete with those
business users, unless the data is publicly available, obligations to allow end users to un-install any pre-
installed software applications on its core platform service, an obligation to allow installation and
effective use of third party software applications or software application stores (subject to certain
carve-outs) and not to technically restrict end users from switching between and subscribing to
software applications and services accessed under a gatekeeper’s operating system; an obligation to
allow business users and providers of complementary services access to and interoperability with the

37

Electronic copy available at: https://ssrn.com/abstract=3776274

same operating system, hardware or software features that are available or used in the provision by
the gatekeeper of any ancillary services; an obligation to refrain from using any aggregated or non-
aggregated data, which may include anonymized and personal data that is not publicly available to
offer similar services to those of their business users ; an obligation to provide to any third party
providers of online search engines, upon their request, with access on fair, reasonable and non-
discriminatory terms to ranking, query, click and view data in relation to free and paid search
generated by end users on online search engines of the gatekeeper, subject to anonymization of the
query, click and view data that constitutes personal data; and an obligation to apply fair and non-
discriminatory general conditions of access for business users to the gatekeeper’s software application
store.

In short, in proposing this kind of ex-ante regulation the European Commission wants to be able to
prohibit some of the practices it has examined in the past without having to define relevant markets,
to assess market dominance or to bear the burden of establishing that these practices are capable of
restricting competition. This is definitively a move away from an economics based approach to
competition law enforcement. It is noteworthy that Ms Vestager has publicly justified this proposal on
the grounds that many of those practices were unfair to the business users of platforms.

It is somewhat difficult to follow the logic of this proposition from a competition standpoint.

The definition of “Gatekeepers” is, firstly, limited to providers of “core platform services”. According
to the Commission, core platform services include: (i) online intermediation services ( for example,
marketplaces, app stores and online intermediation services in other sectors like mobility, transport
or energy) (ii) online search engines, (iii) social networking (iv) video sharing platform services, (v)
number-independent interpersonal electronic communication services, (vi) operating systems, (vii)
cloud services and (viii) advertising services.

Furthermore, not all providers of those services will be considered to be “Gatekeepers”. The
Commission states in its proposal that: “Such gatekeeper status can be determined either with
reference to clearly circumscribed and appropriate quantitative metrics, which can serve as rebuttable
presumptions to determine the status of specific providers as a gatekeeper, or based on a case-by-
case qualitative assessment by means of a market investigation”.

With respect to the quantitative metrics, the proposal defines three conditions for a platform to be
presumed to be a gatekeeper:

- first, the core platform services provider must have a significant impact on the EU internal market.
Such a significant impact will be presumed to be met where the undertaking that the provider forms a
part of has achieved more than EUR 6.5 bn turnover in the EEA in each of its last three financial years
or the average market capitalization or the equivalent fair market value of the undertaking was at least
EUR 65bn in its last financial year and it also provides a core platform service in at least three EU
Member States;

- second, the provider must operate a core platform service which serves as an important gateway for
business users to reach end users. This condition will be presumed to be met by providers of core
platform services that have more than 45 million monthly active end users established or located in
the EU and more than 10,000 yearly active business users established in the EU in the provider’s last
financial year;

- Finally, the provider must enjoy an entrenched and durable position in its operations or it is
foreseeable that it will enjoy such a position in the near future. The existence of such an entrenched

38

Electronic copy available at: https://ssrn.com/abstract=3776274

and durable position will be presumed to be met if the rebuttable presumption thresholds for the two
previous conditions were met in each of the last three financial years.

This definition of “Gatekeepers” is well adapted to activities which are dominated by very large
platforms such as digital retail (with Amazon) or social networking (with Facebook). This is not
surprising since the EU Commission states in its proposal: “A few large platforms increasingly act as
gateways or gatekeepers between business users and end users and enjoy an entrenched and durable
position, often as a result of the creation of conglomerate ecosystems around their core platform
services, which reinforces existing entry barriers. As such, these gatekeepers have a major impact on,
have substantial control over the access to, and are entrenched in digital markets, leading to significant
dependencies of many business users on these gatekeepers, which leads, in certain cases, to unfair
behaviour vis-à-vis these business users. It also leads to negative effects on the contestability of the
core platform services concerned”.

However, its applicability raises some questions for sectors in which a number of competitors each
operate a platform which serves as an important gateway for business users to reach the users of the
platform.

For example, it is not obvious that the definition applies to automobile manufacturers of connected
cars. Yet, in the extended vehicule model, which is the reference model that automobile
manufacturers have pushed, each car manufacturer has a privileged and exclusive in-time access to
the data provided by the connected cars of its make since this data is automatically transmitted by
automobiles and stored on the manufacturer’s server. Because the car manufacturer is in potential
competition with the independent providers of a number of after sales services (such as car repair or
maintenance shops or insurance companies) or complementary services (such as mapping services or
audio visual services), the competitors are entirely dependent on the transmission by the
manufacturer’s server of the information generated by the connected car in real time to be able to
offer those services competitively with the car manufacturer.

Thus, although there is reason to believe that non discriminatory access to the data gathered by
manufacturers of connected cars is crucial for the independent after sales service providers competing
with them, it is not clear that that each one will meet the quantitative thresholds set by the
Commission for the rebuttable presumption of “Gatekeeping” or that these car manufacturers will be
(qualitatively) considered as having a significant impact on the EU internal market in the assessment
of a market investigation. Much will depend on the interpretation given by the Commission of the
relevant EU “internal market”.

Secondly, the applicability of the regulation is limited to platforms providing core services which have
an ‘entrenched and durable position’ either at present or foreseeably ‘in the near future.’

But the quantitative elements on which the presumption of “ an entrenched and durable position” is
founded ( i.e. their turnover in the EU or the number of active end users or business users in each of
the last three years) tell us nothing about whether they will have an entrenched or durable position
in the future. Furthermore, as we saw, competition authorities have a difficult time understanding the
logic of competition between ecosystems, and it is not clear that the EU Commission will be able, in a
market study, to assess qualitatively which platforms have an entrenched and durable position and
which do not. As mentioned earlier, the rapidity with which Instagram became a major social network
was not recognized by competition authorities at the time of the merger between Facebook and
Instagram.

39

Electronic copy available at: https://ssrn.com/abstract=3776274

Similarly, back in 2005 when MySpace was acquired by NewsCorp, (46) “millions of teenagers across
the world used MySpace, spending hours every day connecting with each other online and fine-tuning
personal profile pages that reflected their tastes and personalities”. Yet within three years Facebook,
which had been created in December 2004, had become a serious competitor and within five year
years ( i.e. in December 2009) MySpace’s share of the social networking market has tumbled from 66
per cent to 30 per cent, with MySpace losing US $ 100 million in 2009. Facebook had expanded its user
base, articulating a qualitative product differentiation between itself and its competitors, and
integrating new ways of engaging users into its suite of social networking functions by offering new
features and functionalities(47). Facebook surpassed its main rival, MySpace, to become the most
popular website in the United States in 2009, just five years after its founding(48). There is little doubt
that had the proposed EU regulation been in force in 2005 MySpace would have been considered to
have an entrenched and durable position. Yet it was all but gone within four to five years of its
acquisition by NewsCorp.

As Newscorp found out and, as the CMA demonstrated, in the previously discussed Facebook
Instagram merger, it is particularly difficult to assess whether a platform has an entrenched position
on a market and what seems to be a “durable” entrenched position can, in fact, evaporate very rapidly.

What the EU Commission’s approach seems to ignore is that competition among ecosystems comes
about through disruptive innovation and a fight among ecosystems for the market rather than in the
market. In such circumstances , the past is a poor predictor of the future.

Third, some of the obligations to be imposed on the “gatekeepers” platforms contemplated by the EU
proposal seem to misunderstand the function of an ecosystem and could actually restrict competition
or innovation of such ecosystems in the name of fairness or of protecting competition within an
ecosystem.

For example, preventing gatekeepers from combining personal data from their core platform services
with data from other sources (including other services offered by the gatekeeper) is likely to restrict
the information a platform can use to improve its value proposition to consumers. This proposal seems
to ignore, on the one hand, the fact that in an ecosystem it is the provision of complementary services
which increases the value of the ecosystem for its users and, on the other hand, the fact that for a
number of services offered, the quality of the application is, ceteris paribus, increased by the amount
of data used to train the artificial intelligence algorithms which allow the provision of these
complementary services .

If it seems sensible to argue that the consumers should have the right to oppose the use of their data
for other services than the ones for which they provided their data, a prohibition per se of the
combination of data from gatekeeper core services with the data from other services is very likely to
limit the quality of the services they render and hence the welfare of the users.

Similarly, the obligation for a platform which is also a market place not to use data acquired by the
platform in relation with business users to then compete with those business users, unless the data is

46
Matthew Graham, The rise and fall of MySpace , Financial Times Magazine, 4 December 4 2009
47
Mark Glick, Catherine Ruetschlin, & Darren Bush : “BIG TECH’S BUYING SPREE AND THE FAILED IDEOLOGY OF
COMPETITION LAW”
48
Mark Glick, Catherine Ruetschlin, & Darren Bush : “BIG TECH’S BUYING SPREE AND THE FAILED IDEOLOGY OF
COMPETITION LAW”

40

Electronic copy available at: https://ssrn.com/abstract=3776274

publicly available is clearly likely, in the name of fairness, to decrease the intensity of competition
among the participants to an ecosystem to the detriment of the welfare of the users.

Asymmetric data sharing regulatory obligations or interoperability obligations imposed on platforms

seen as “gatekeepers” may allow smaller competitors to upgrade the quality of their service offering
and overcome their disadvantage of size or allow independent service providers to offer services to
users of the platforms in competition with the platform itself. But such obligations may also decrease
the incentive of platforms to accumulate data and have a deleterious effect on the quality of services
offered to consumers and on innovation.

As Wolfgang Kerber and Heike Schweitzer ( 49) argue “interoperability obligations can lead to a higher
level of homogeneity among platforms. To the extent that uniform standards and interfaces are used,
the possibilities of firms to develop their own specific products and services are limited, because they
have to comply with these standards and interoperability requirements. This will limit the scope for
innovation and therefore the extent to which specific consumer preferences can be fulfilled by way of
product differentiation. Although greater interoperability may lead to more innovation and
competition with regard to complementary products, it also can lead to less innovation and competition
with regard to the standards and interfaces themselves, which may have the characteristics of natural
monopolies (with all their negative consequences). Furthermore, the openness of products and
platforms for complementary products can lead to higher risks for consumers, if the complementary
products offered by other firms are not monitored closely with a view to their interoperability, quality,
and safety. Through a generally higher level of interconnectedness in a digital economy, more
interoperability may lead to higher risks regarding reliability, security, and privacy. Considering these
(potentially large) costs of interoperability, the policy objective should not be full or maximum
interoperability, but rather an optimal degree of interoperability that balances benefits and costs”.
They conclude that “:“The complex trade-offs between benefits and costs of a higher degree of
interoperability suggest the need for a careful and separate analysis of each specific interoperability
issue, caution regarding a (top down) imposition of mandatory standards and interoperability
obligations, and a greater focus on unilateral solutions of interoperability problems, such as adapters
or converters. Within the framework of Art. 102 TFEU, EU competition law may be better advised to
develop a workable test to address hurdles for unilateral interoperability solutions created by dominant
firms, than to continue focusing on the essential facilities doctrine to mandate interoperability”.

If data portability is ineffective to counter the network effects that dominant platforms may enjoy, it
is sometimes considered useful to alleviate any data-induced lock-in effect. However, even this
solution may not be effective in all cases. For example, Gabriel Nicholas and Michael Weinberg ( 50)
note : “However compelling in theory, few have investigated whether competitors can actually use
ported data to create or grow competing platforms. This gap is particularly troublesome because we
found no competitive products built on ported data, despite the fact that many large platforms have
enabled users to export their own data for years. For example, Facebook has allowed users to download
their data since 2010—well before current competition concerns emerged, and long enough ago for a
competitor built on ported Facebook data. Still no such competitor has emerged”. Indeed the results
of their empirical investigation suggest that ported Facebook data is simultaneously insufficient to
replicate Facebook and too tailored to Facebook to be useful for anything else. They conclude by

49
Dr. Wolfgang Kerber,Prof. Dr. Heike Schweitzer” Interoperability in the Digital Economy” Journal of
Intellectual Property, Information Technology and E-Commerce Law, 2017
50
Gabriel Nicholas and Michael Weinberg “Data Portability and Platform Competition: Is User Data Exported
From Facebook Actually Useful to Competitors?”, Engelberg Center on Innovation Law and Policy, New York
University, November 2019

41

Electronic copy available at: https://ssrn.com/abstract=3776274

stating : “that regulators should not assume that competitors will be able to use ported data to build
innovative products and services. An overreliance on data portability may distract from more effective
tools for addressing concerns with large platforms”.

The regulatory framework for gatekeepers proposed by the European Commission as a complement
to competition law enforcement raises three questions which deserve further elaboration respectively
concerning its scope, the underlying analysis of platform competition and the remedies proposed.
These questions no doubt will be (and should be) further discussed in the next few months during the
consultation phase on this proposal.

First, the proposal does not seem to aim at solving the competition issues raised by gatekeepers in the
digital sector in general but limits itself to the sub-set of these problems raised by a small number of
very large platforms without providing a clear rationale for this choice. Thus, it is difficult to avoid the
impression that this proposal is driven more by the political desire to act against these large platforms
than by the desire to promote competition and innovation in the digital sector in general.

Second, the proposed regulation seems to have, at least partially, overlooked a number of the specific
features of competition within and between ecosystems and in particular, the essential role of core
platforms in monitoring the complementarity between the services they offer and the services offered
by their third party complementors within an ecosystem in order to promote competition between
them and other ecosystems, the fact that restricting access to the data which the core platforms of an
ecosystem can use to train their algorithms may have the undesirable consequence of lowering the
quality of the services they offer to their users, as well as the fact that competition among ecosystems
is not a competition for market shares among firms in preexisting markets but competition for the
market through disruptive innovations.

Third, with respect to remedies, the proposal does not seem to sufficiently take into consideration
both the trade-offs involved in the interoperability of platforms and the limits of data portability.

Merger control

With respect to merger control, article 12 of the proposed regulation creates an obligation for
“Gatekeepers” to notify any intended merger or acquisition that involves, within the meaning of the
EU Merger Regulation, another provider of core platform services or of any other services provided in
the digital sector. The obligation, which applies to all transactions whether or not they would trigger a
notification under the EU or the national merger law, is designed to inform the Commission on whether
there is any need to revise the individual gatekeeper designation and to help the Commission monitor
the general trend in the digital sector. Thus, the EC proposal is different from proposals in other
jurisdictions which envisage a derogatory merger control review process applying to certain digital
firms.

However this does not mean that the EU Commission is not intent on controlling mergers by large
firms with start-ups having a very low turnover in order to prohibit killer acquisitions.

On 11 September 2020 Ms Vestager, the EC Commissioner, in a speech to the International Bar

Association, rejected the idea of an amendment to the EU Merger regulation in order to be able to
control possible killer acquisitions. However, Ms Vestager stated that from mid-2021, the EC
Commission will accept referrals, made under Article 22 of Regulation (EC) No. 139/2004 (known as
the “Dutch clause”), by Member States of mergers affecting trade between Member States but which
do not have Community dimension and which the Member States Competition Authorities are not
competent to review under their national merger control rules.

42

Electronic copy available at: https://ssrn.com/abstract=3776274

If it thus likely that the EU Commission will be able to review most mergers which could be killer
acquisitions in the digital sector.

However, at the substantial level, competition authorities in general and the EU Commission in
particular face a number of challenges in controlling mergers in the digital sector.

First, as mentioned earlier, the definition of the relevant markets on which a platform operates is much
more complex than in the non digital world, in particular on the sides where services are offered by
the platform at zero price to consumers. Yet it is crucial in a merger examination to understand what
are the competitive pressures faced by the merging parties. If the traditional instruments such as the
evaluation of cross price elasticities or the hypothetical monopolist test can be useful on the sides of
the platform where users (such as advertisers) have to pay a positive price, these instruments are
useless if prices are equal to zero. New tools have thus to be developed to assess substitutability
between free services. Brynjolfsson E. and A. Collis (51) developed a method for measuring the benefits
associated with the digital services using large-scale well-designed surveys by exploring how much users
would need to be paid to give up a given digital service for a certain period of time. It has been
suggested that this sort methodological tool could be used to map the preferences of consumers with
respect to alternative digital services. This kind of approach combined with the traditional
methodologies to assess substitutability on the sides of the platforms for which users pay could give
competition authorities a more accurate view of the competitive environment of the merging parties.
So far competition authorities have had little experience in running methodologically sound large scale
surveys of consumers.

Second, merger control assessment is a prospective exercise and competition authorities must
therefore compare the expected effect of a merger with the expected developments on the identified
relevant market if the merger did not take place over a period of a few years (usually 3 to 5 years). The
questions raised by this prospective exercise require the competition authority to assess what would
have happened to the target barring the merger, what potential entry into the market could take place
within the time frame considered and eventually how competition on the market would play out if the
merger took place and how it would have played out barring the merger. This predictive exercise is
always difficult (which is the reason why competition authorities tend to focus on the effect of the
merger in the few years following the merger). But the exercise is made even more difficult in dynamic
sectors in which competition on quality rather than on price plays a large role and which are
characterized by disruptive innovations and unstable business models.

These characteristics imply that competition authorities cannot rely on the past to predict the future.
Indeed, we have seen that dominant firms like MySpace can be displaced in a few years, that a
comparatively small platforms like Instagram can develop very rapidly and compete with the most
established social media networks thanks to a shift in the pattern of consumption of users or that a
unicorn like TiK Tok can develop an innovative technology to produce and exchange videos, which even
the more successful social media cannot emulate, and capture a large segment of the social media
market in just a few years. To assess whether the target firm in a digital merger is a potential
competitor of the acquiring platform or whether other potential entrants could materialize
competition authorities have to look beyond the characteristics or structure of the digital markets at
the time of the merger and evaluate the dynamics of the ecosystems and platforms. This means that
competition authorities must develop a specific intelligence function with respect to digital markets

51
Brynjolfsson E. and A. Collis. “How should we measure the digital economy?” Harvard Business
Review, 2019

43

Electronic copy available at: https://ssrn.com/abstract=3776274

allowing them to have a more dynamic approach to merger control. There have been calls for
competition authorities to set up digital units within their institutions ; those units must clearly
understand the business side of the ecosystems and the evolving characteristics of digital service
users’ demand (while the competition authority itself must be vigilant to avoid being captured by the
digital firms) .

A third issue relevant to merger control of digital platforms is that consumption network effects (or
demand economies of scale) mean that size allows an increase in the quality of the service offered by
a platform. Furthermore, the more data a platform has access to in order to train its artificial
intelligence algorithms, the higher is the likelihood that it will be able to increase the value of its
services to users (data synergies effect). This, combined with the fact that quality is a crucial dimension
of competition for users of free services, means that size begets efficiency together with the risks
associated with market power. In traditional sectors, mergers can increase the size of a firm and allow
it to benefit from enhanced scale economies, thus reducing its average cost, without the beneficiary
necessarily passing on to consumers those efficiency benefits. In the digital sector, on the other hand,
the efficiency benefits associated with the ability to train artificial intelligence algorithms on larger
data sets increase the qualities of the service provided by platforms and are thus necessarily passed
on to consumers. This means that competition authorities must consider the efficiency benefits
associated with digital mergers simultaneously with the possible restrictions to competition that those
mergers entail.

A fourth issue is that of remedies. One of the consequences of our previous discussion on the possibility
of data efficiciencies in mergers between platforms, is that structural remedies, such as divestitures,
on which competition authorities have traditionally relied to solve competition problems associated
with anticompetitive mergers may have the unintended consequence of lowering the quality of digital
services to users and reducing consumer welfare if they are not complemented by measures to
facilitate data portability between the merged firms and the divested businesses. But, as we saw
earlier, measures to facilitate data portability may not always be effective or may have some negative
consequences on competition.

Furthermore, given the difficulties in understanding the dynamics of the digital sector, given the fact
that new or different business models are constantly appearing, and given the fact that the behavior
of users of digital services is not always well understood or anticipated, it is very difficult for
competition authorities to formulate behavioral remedies in digital merger cases. As Geoffrey Parker,
Georgios Petropoulos, and Marshall Van Alstyne (52) state : “we should strengthen the ex-post
evaluation of merger analysis for big platforms to better understand the validity of analysis at the time
of the merger and whether the proposed remedies are the appropriate ones. Mistakes in this analysis
should receive a particular attention and have a didactic function when the same big platform comes
forward with a notification of its next merger.

We should be ready to impose remedies that are contingent on specific future outcomes. If it becomes
clear that the remedies attached to the past approval of a merger do not have the desired effects, there
should be flexibility such that remedies could be modified accordingly. It would be helpful if remedies
are periodically reviewed to assess whether they have the desired effect and are then revised or
updated. The specific targets in terms of the welfare impact of a merger as well as authorities’ concerns

52
Geoffrey Parker, Georgios Petropoulos, Marshall Van Alstyne “Platform Mergers and Antitrust” January 10,
2021, SSRN accessed on January 23 2021 at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3763513

44

Electronic copy available at: https://ssrn.com/abstract=3776274

should be clearly communicated at the time of the approval of the merger. Remedies should be flexible
to change in order to ensure that the specific targets are reached, if needed”.

Conclusion

Overall, the US underenforcement of competition law in the digital sector cannot be sustained much
longer, both for political and for technical reasons. But the assertive enforcement of competition laws
in the digital sector by other jurisdictions has so far been based on shaky analytical grounds, a poor
understanding of competition within and between ecosystems and of their interaction and also by
confusion about the goals of competition law. To be effective, the EC Commission proposal to regulate
the digital sector, which purports to give the EU commission new means to intervene in the digital
sector, needs to be refined to avoid the risk that it may backfire and ultimately restrict both
competition and innovation in the digital sector.

In order for competition law to become more relevant in the digital sector, competition law
methodology and instruments need to be adapted to the complex specificities of this sector. This
adaptation has not taken place for a number of reasons.

First, competition authorities are generally reticent to say that the instruments they have relied on in
the past and the methodologies they have used have limited applicability in the digital sector for fear
that the use of a new conceptual apparatus and new instruments for this sector would meet the
skepticism of judges who value the stability of jurisprudence more than the reliance on new economic
thinking or tools.

Second, it is difficult for non-specialist regulators to accept the explanations of the business leaders
of the digital sector on why and how competition in the digital world is different from what it is in non-
digital sectors. Competition authorities have a long experience of listening to business leaders in all
sectors of activities claim that competition law provisions should not be applicable to them the way
they are applicable to other businesses because of the specificity of their activity. These authorities
are, in general, rightfully skeptical of such claims and so they are biased toward relying on their past
successful theories of harm rather than trying to adapt to a new economic logic.

Third, the fast growing academic literature on the digital sector started with business policy articles
discussing the business models of digital firms and has only relatively recently attracted the attention
of market competition specialists. There is thus a genuine difficulty in understanding what this
literature implies for competition law enforcement.

The review of past decisions and the consideration of the economic and business policy literature on
the digital sector, however, do suggest a few urgently needed improvements to competition law
enforcement in the digital sector.

First, there is a need to deepen our understanding of barriers to entry in the digital sector. The
economic literature has largely focused on the role of network effects, scale economies, learning
economies and data accumulation as constituent elements of the protection that the largest platforms
enjoy against entrants or smaller players. However, as we saw, some economists have called into
question the importance of these factors ( 53) and found that second or third movers actually benefit
from an advantage over the first-movers. But these approaches seem to ignore the importance of

53
See for example Jordi Casanova “Online search engine competition: why Bing vs Google Search does not
support entry barriers due to network effects, Published 20 june 2020, LinkedLn

45

Electronic copy available at: https://ssrn.com/abstract=3776274

commercial barriers to entry for second or third movers, as evidenced in the case of the challenge of
Tik Tok to Facebook or of SnapChat‘s reaction to Tik Tok.

Second, there is a need to find new indicators of market power for platforms and ecosystems than
those used for single-market firms (indicators which are largely based on cost price comparisons),
taking into account the fact that platforms are multi-sided, that users on various sides interact, that,
on some sides, the services are offered for free and that, on these sides, competition on quality is
crucial to attract the maximum number of users which is the objective of the platforms.

Third, there is a need to recognize that competition between ecosystems is, to a large extent, the
result of differentiation of the combination of services that the ecosystems offer to their users and
that this competition has a winner take all dimension. This has a number of implications. It means, first
that market shares (however they are calculated) at any one time may not be an indicator of
dominance in the future since a completely different combination of services may replace the currently
existing dominant ecosystem. This also means that to assess potential competition among ecosystems
one has to look, first, at the extent to which the technology of various specialized platforms (even if
they offer very different services than the ecosystem considered) could be used for the distribution of
a number of other complementary services and, second, at the evolving demands of platforms users
for a combination of services. Finally, this means that there is a need to use the teachings of the
business policy literature to understand what are the factors of success of a digital startup in order to
assess whether the acquisition by a platform at the center of a large ecosystem of a small application
producer is likely to restrict competition or is likely to spur more innovations.

All of this requires investment in research staff which have not only a technological understanding but
also a business understanding of digital ecosystems, including an understanding of the dynamics of
users’ demand for ecosystem service – i.e. the demand trends.

Fourth, when dealing with competition issues among applications within an ecosystem there is a need
to understand and take into consideration the interaction between these issues and the competition
between ecosystems in order to be better able to differentiate what is an anticompetitive practice
from what is a procompetitive practice.

Fifth, there is a need to further research the role of data in competition among platforms and of
innovation in digital services. In particular, it would be useful to explore the characteristics of cases
where access to the data gathered by an ecosystem or by a platform are necessary for competitors of
this ecosystem to be able to compete effectively as well as to define/explore the kind of data we are
talking about. One should also explore the conditions under which interoperability and data portability
would unambiguously enhance competition (54), the cases where they could actually make it more
difficult for entrants to grow and, finally the conditions under which the quantity and quality of data
gathered by ecosystems would be affected by the conditions put on the use of these data. There is
also the related question of the scope for convergence on the issue of whether privacy is a relevant
element of quality which can or should be dealt with by competition authorities.

54
See, for example, Jordi Casanova “Online Search Competition and the Risk of Unintended Consequences of
Data Access”, Competition Policy International, January 2021 in which the author argues that mandatory
access to digital platforms’ data is likely to undermine competition on the merits and the incentives to offer
services to consumers free of charge. Furthermore, it could distort companies’ incentives to innovate with user
data.

46

Electronic copy available at: https://ssrn.com/abstract=3776274

Sixth, there is a need to better assess the origin, the nature and the importance of the biases of
consumers digital which may help dominant platforms in an ecosystem to foreclose the producers of
competing applications. In many cases the vague mention of the existence of such biases is insufficient
to assess the competitive implications of these biases or to design appropriate remedies. In particular,
it would be useful to ask what we know about the implications of consumer biases in various sectors,
how competition authorities have dealt with this issue in non-digital sectors and what lessons we have
learned from these experiences (or from the disregard for this issue in other sectors).

Seventh, there is an urgent need to improve our methodologies with respect to merger control in the
digital sector. We need a merger control system that will be more relevant, more dynamic, with a
better grasp of the interaction between competition and efficiency in the digital sector and more
flexible when it comes to remedies. This will require changes in perspective as well as new instruments.

Fulfilling this (admittedly large) research agenda is a prerequisite if we want to ensure that competition
law enforcement and the design of the regulatory environment of digital services are well founded
analytically, relevant and predictable. An exigent intellectual effort is the only way to ensure that
competition authorities will avoid the risks of inadvertently giving in to the political pressure of
economic populism or ideology or issuing misguided decisions which may be ineffective or, even
worse, restrict competition or innovation.

47

Electronic copy available at: https://ssrn.com/abstract=3776274