Market Power in the Digital Markets. Do we really need a new market power definition?

by Alessandro Amoroso


The notion of market power is central to antitrust law. Such concept does not refer only to a dominant position but, actually, constitutes a general paradigm of the EU antitrust law.[1] A monopolistic conduct[2] certainly triggers the existence of market power, but also agreements[3] and mergers[4] relate to it.

Operationally, the assessment of whether a firm has market power or has reasonable prospects of achieving it, is the first task of any antitrust analysis.[5] Generally speaking, besides some special cases[6], the detention or acquisition of market power itself is not relevant for the enforcement of the EU competition rules applicable to enterprises, and has relevance only where it leads to anti-competitive behaviours.

In determining the firms’ market power, the first step to take is the assessment of the market definition and consequently the determination of the relevant market.[7] The relevant market is a tool to establish who competes against whom, by determining which goods or services potentially compete, to the exclusion of those that do not.[8]

Understanding the relevant market, is not a mechanical or abstract process but requires the analysis of previous behaviours in the market and a global understanding of the mechanisms of each sector, encompassing a two-stage examination[9].

In the first stage, it shall be taken into account, the substitutability of the demand[10] and, to a lesser extent[11], the substitutability of supply.[12] In the second stage, it shall be determined the geographic market, namely that in which the conditions of competition are sufficiently homogeneous to enable the judgement of the economic power of an undertaking.[13]

Once decided the relevant market, it shall be looked at the market shares (of the firm and of its rivals), balancing them with a global analysis of the economic features of such market, thus considering barriers to entry and bargaining power of clients and suppliers, and all the other structural characteristics of the market.

It is clear, therefore, that defining market power is not an easy task, since it is a notion that relies on disputed and shaded conditions.[14]

This said, we may now focus on data.

Issues in the digital market

In a world with zero distribution costs and zero transaction costs, the difference among online platforms is made predominantly by quality: consumers are attracted through the delivery of a superior experience, which draws modular suppliers, this improves the experience, attracting more consumers, and thus more suppliers. The consequence is a continuous virtuous circle[15], fostering the generation of data giants.

With the emergence of data as an asset for market players operating in the digital economy and as an increasingly important corporate currency, questions have risen about the relevance of data for competition enforcement and for the definition of market power. It is an issue currently under debate, hence, whether data collection should constitute a relevant market in which market power is assessed, and, consequently, if the possession of large amount of data by online platforms could constitute, just as such, an unfair competitive advantage and an unlawful barrier to entry for competitors that should be challenged by regulators in order to avoid effects entrenching market power.

Recently, indeed, a number of legal experts, practitioners and policy activists have called on antitrust regulators to incorporate the possession of data into their analysis of mergers and possible anticompetitive practices[16] and have suggested that data collection should form its own product market for the purpose of an antitrust analysis, by building a data-targeted market definition.[17]

According to such scholars, even though a platform may not have a large turnover, data involved could, indeed, entail a significant competitive advantage over rivals and end up by threating consumers, particularly in consideration of the economies of scale and network effects that characterize online platforms.[18]

The main antitrust questions that the new digital markets have raised therefore are two:

1) Should the aggregation of data be taken into consideration by competition law? Is data aggregation a harm per se to market competition?

2) Should action be taken? Is there a need to change the current legal norms and adapt the market power definition in order to deal with the digital economy and make express reference to data power? If so, how should it be done? [19]

Is data aggregation really a competition law issue?

The issue of data aggregation is two-folded.

On the one hand, although online data brokerage is emerging and some data marketplaces have actually had birth[20], we have to acknowledge that data yet mostly do not constitute a stand-alone product of a relevant market but more a parallel asset. Data, indeed, are primarily used by online platforms as an input for production in order to improve services and make users’ interactions more efficient; in most cases there is no competition between providers for the actual sale of data and no substitution, hence, no relevant market can be considered to exist. In most cases in which data processing is involved advertising services are the relevant product.[21]

On the other hand, the fact that control over large amounts of data could in some instances be used to thwart competition is out of any doubt.

If data were considered a relevant market, as some suggest[22], access to a large volume or variety of data would have to be avoided as such, since data power could cause autonomously a harm to competition. However, merely having data more than one’s rivals has, should not constitute by itself a risk for the market or, in any case, should not be an antitrust issue.[23] Unduly curbing data collection and data use would depress innovation and technological advancement, while, at the contrary, policymakers should be encouraging it. As Joe Kennedy’s sustains any potential harm driven by the concentration of data is most likely outweighed by the benefits of scale; he deems that “the focus should be on abusive behavior and not on structural issues, such as how much data a company holds”, since, actually, an abusive conduct can occur in industries with higher as well as moderate concentration ratio and “data-rich companies are not an economic threat, but rather an important source of innovation” [24].

Kennedy endorses his opinion referring to the features of data: data is a non-exclusive and non-rivalrous good, that it is often highly substitutable, easily and cheaply accessible, and that in many cases may lose its worth very quickly. Due to such features, data market dominance in digital markets does not convey lasting market dominance; actually, as it has been said about Google’s dominance or loss of it, dominance in the digital markets is “only one click away’[25].

Today’s regulation; do we need a new definition of market power?

The second issue faced by the regulators has been whether a new antitrust regulation for digital markets is actually needed, i.e. a definition of market power that takes into account the characteristics of the digital market .

Some eminent scholars[26], and Joe Kennedy himself, in contrast with Stucke and Grunes[27], deem that if data-rich companies thwart competition, the regulators already have the tools to address such harms. According to such scholars: (i) the main detriment that data aggregation poses to consumers may be more a “degradation of privacy” than a competition issue[28]; (ii) competition law should not persecute mere relevant market power, but should intervene only when other interests, predominantly consumers’ interests, are at stake; (iii) market power shall already be assessed acknowledging a global analysis of the relevant market – “market power is not a standalone concept, but one that can only exist in relation to the market(s) in which the company interacts and importantly how it interacts”[29]. Indeed, the legal definitions of market power settled by the European Court of Justice, namely the “power to behave to an appreciable extent independently of its competitors, customers and ultimately its consumers”[30] and the power to “if not to determine, at least to have an appreciable influence on the conditions under which that competition will develop”[31], are flexible and broad, encompassing many measuring tools.

Furthermore, since digital economies have shown to influence strongly traditional markets and are a very dynamic and inconsistent economy, fragmenting the understanding of the unitary notion of market power would be an impracticable approach with more flaws and pitfalls than benefits.

In light of the above, there is no need for a new market power definition for the digital economy, at the contrary, the legal framework should simply be interpreted acknowledging the particular features of the digital markets, namely that:

1) there is a tendency to compete for the market more than in the market; the competition is a “winner takes all competition”, it is a different type of competition;

2) network effects are strong and with these come economies of scale;

3) multi-sided markets are common, and both customer groups need to be considered, even if one group receives the product for free;[32]

4) innovation is fast-paced, hence time scale of dominance – generally 5 or more years – is a relevant clue for market power.[33]


In conclusion, it is out of any doubt that the challenges posed by the digital economy and the future developments at stake are high, but the point is, partly approving the cited authors’ view, that EU Law and EU authorities should intervene, not by changing the definition of market power or its regulation, but rather by getting to grips with the digital markets, answering, in practice and effectively, to the challenges posed.

Certainly, it is much more difficult to assess market power in the digital environment (market share is a relatively poor proxy for power when the product at issue is information[34], prices are generally homogenous and often services are free), but are we sure that traditional competition policy is inadequate to deal with harmful conducts in it?

The digital market is, yet still, in its infancy and is an ever changing and a very fast-paced market. A new definition of market power and of relevant market for the data sector would be risky and, mostly, is not needed, since the existing theories that support such intervention conflict with the realistic characteristics of data and, also, because the current definition of market power is broad and flexible enough to apply in the cases of market abuse. Moreover, limiting the collection and use of data would probably harm competition and innovation more than foster it. Even though we could recognize that large amount of data enables market dominance, bigness is not an antitrust offence, thus, it would be wrong to consider the sole fact of a vast collection or use of data as a question of market power.

The solution, therefore, is, not to impede large data collections or to provide ad hoc rules for data mergers and data controls, but rather to create a substantial favourable background, in privacy and data law[35], to permit an effective competition, allowing the market to adjust itself. In essence, indeed, if the structure of the market is such that there is little potential for consumers to be harmed, we do not need to be especially concerned with how firms behave because the presence of effective competition will provide a powerful antidote to any effort to build up digital monopolies or to harmfully exploit consumers.[36]

The new General Data Protection Regulation[37] (GDPR) may already contain some relevant provisions capable to affect competition and weaken the data giants. Such Regulation empowers consumers to take back some control of their data, and Internet giants face stiff fines if they abuse of their dominance.[38]

A relevant provision could be, for instance, the new “right to data portability”[39] provided by Article 20 of the GDPR. The right to data portability provides for mandatory interoperability and this could have a significant impact on facilitating market access, preventing high switching costs and consumers’ lock-in effects as well as alleviating network effects that threaten potential competition in a marketplace.[40]

Concluding, is market power really different in the digital markets? Is it really so elusive that a new definition of market power is needed? Maybe not; maybe it would be sufficient to empower competition and act on other fields of law different from antitrust law, because no doubt that, as Janelle Orsi affirmed, “Internet platforms are like the wizard of Oz, all they have is your devotion”[41]. And devotion, with the right conditions, can easily move away.

[1] The entire reference to such statement would be too long, and this is not the purpose of this article, thus for further reading: Blanco Luiz Orthiz, Market power in EU antitrust law, Oxford, Portland, 2012.

[2] Art. 102 TFEU.

[3] Art. 101 TFEU.

[4] Merger Regulation, EU Reg. 139/2004.

[5] Hay, George A., Market Power in Antitrust, Cornell Law Faculty Publications. Paper 1158, 1992. Available at:

[6] ECJ, Europemballage Corporation and Continental Can Company Inc. v. Commission, C-6/72; Microsoft Corp v Commission (2007) T-201/04 – “Microsoft impaired the effective competitive structure on the work group server operating systems market by acquiring a significant market share on that market” .

[7] Shmalensee Richard, A presentation on Market Definition, Competition Policy International, Mexico Competition Day, April 4, 2011. Available at: It shall be noticed, however, that whilst market definition plays an absolutely central role in EU law, it plays a different lesser role in the US, where the former relevance is given to the behaviour of the theoretical monopolist –  Baker J., Market Definition: An Analytical Overview, Antitrust Law Journal, 129-74, 2007.

[8] In general, the more narrowly the market is defined the more likely a firm or firms will be found to have market power.

[9] Whether for the purposes of Art. 101 TFEU, Art. 102 TFEU or with respect to European Merger control.

[10] ECJ Hoffman-La Roche (1979); ECJ United Brands (1978); ECJ L’Oreal (1980).

[11] Only on condition that its effects are equivalent to those of the substitutability of demand in terms of efficiency and immediate response.

[12] E.g. Commission decision, Tetrapak I (BTG License), 1988; ECJ Continental Can. (1973) para. 33.

[13] ECJ Alsatel 1988, paras. 14-15.

[14] Economists usually define market power as the ability to price above short-run marginal cost and, in the long run, above average total cost – Blanco Luiz Orthiz, Market power in EU antitrust law, Oxford, Portland, 2012.

[15] It’s the Aggregation Theory. Ben Thompson, Antitrust and Aggregation, Stratechery, April 2016.

[16] Google Dominates Search. But the Real Problem Is Its Monopoly on Data, The Guardian, April 19, 2015;; Bernasek A. and Mongan D.T., Our Massive New Monopolies: Amazon, Google and Facebook Have the Power to Move Entire Economies, Salon, June 7, 2015.

[17] European Parliament, Competition & Privacy in Markets of Data, Speech by Pamela Jones Harbour, former Comm’r, Fed. Trade Comm’n (Nov. 2012), files/us/images/publications/121127PJHarbourEUParliamentCompetitionPrivacy.pdf. Stucke A. & Grunes S., Big data and competition policy, Oxford University Press, 2016.

[18] Shelanski distinguishes three uses of data as an asset. Customer Data as: (i) commodity; (ii) input of production; (iii) strategic asset. Shelanski H.A., Information, Innovation, and Competition Policy for the Internet, 161 U. Pa. L. Rev. 1663-1678, 2013.

[19] Dzodiak Natalia, EU Asks: Does Control of ‘Big Data’ Kill Competition?, The wall Street Journal, 2 Jan. 2018; European Commission, Online Platforms and the Digital Single market Opportunities and Challenges for Europe, COM(2016) 288.

[20] Barth Paul, The Rise of the data marketplace, CIO, 27 March 2017;

[21] The platforms use data to identify interests and behaviours of the users and to achieve tailored matching of individuals and advertisers. Advertisers benefit from the data because of the higher possibility to reach their intended audience, users benefit from it because of a more interesting advertisement, better quality of the functionalities of the platform, delivered results, suggestions and recommendations. It is a win-win system.

[22] Supra n. 17.

[23] Kennedy JOE, The Myth of Data Monopoly: Why Antitrust Concerns About Data Are Overblown, Information Technology & Innovation Foundation, March 2017. Available at:

[24] Kennedy JOE, ibidem, p.1.

[25] Helft Miguel, Google Makes a Case That It Isn’t So Big, N.Y. TIMES, June 28, 2009, at B1 (quoting Dana Wagner, Google senior competition counsel, as saying “Competition is a click away”); Schmidt E. Testimony, (“[I]f consumers don’t like what one website is providing them, they can switch to another website with just one click.”) at:

[26] Sokol Daniel &Comerford Roisin, Antitrust and Regulating Big Data, Geo Mason L. Rev., 1026.

[27] Stucke A. & Grunes S., Ibidem.

[28] Manne Jeoffrey A. & Sperry R. Ben, The law and economics of data and privacy in antitrust analysis, TPRC Conference, March 1, 2015. Available at SSRN:

[29] Schmidt Hedvig K., ibidem, p. 13.

[30] United Brands Corp. v. Commission, Case 27/76.

[31] Hoffman-La Roche, Case 85/76.

[32] Russo Francesco and Stasi Maria Luisa, Defining the Relevant Market in the Sharing Economy, Internet Policy Review, vol. 5(2), 2016.

[33] In the Microsoft case the Commission dismissed Microsoft’s initial attempt to plead that its dominance was “fragile” in the new operating systems, looking at the fact that Microsoft had held steadily increasingly market shares in the market for clients’ PC operating systems in the last decade – paras.430-435; European Commission Press release, Google Shopping, Brussel, 27 June.

[34] Patterson R. Mark, Google and Search Engine Market Power, Harvard Journal of Law and Technology, July 2013.

[35] Ohlausen Maureen K. & Okuliar Alexander P., Competition, Consumer Protection, and The Right to Privacy, Antitrust Law Journal 80, 2015.

[36] Hay George A., Market Power in Antitrust, 60 Antitrust L.J. 807 (1992)

[37] Regulation (EU) 2016/679 of 27 April 2016.

[38] Siobhan Kenny, Why the Internet needs to start taking regulation seriously, Campaign, Dec. 2017. Read more at:

[39] Such right has two key elements: the right of the data subject to obtain a copy of personal data from the data controller and the right to transfer that data from one data controller to another.

[40] Aysem Diker Vanberg and Mehmet Bilal Ünver, The right to data portability in the GDPR and EU competition law: odd couple or dynamic duo?, European Journal of Law and Technology, Vol 8, No 1, 2017. Available at:

[41] Orsi Janelle, Platform Cooperativism: The Internet Ownership and Democracy, Convention, San Francisco, 13-14 Nov. 2015.