Daniel Cormack, LLB (Hons) Scots Law Student, Law School, University of Strathclyde, UK. An earlier version of this paper was submitted as part of the “Competition Law and the Digital Economy” module within the University of Strathclyde Law School. Many thanks to Oles Andriychuk for his support in publishing this contribution. Responsibility for errors or omissions remains the author’s alone.

 

Introduction

President of the European Commission, Dr Ursula von der Leyen, tasked Margrethe Vestager in December 2019 with the challenge of ensuring that “competition policy and rules are fit for the modern economy”.^[1]The proposal by the Commission for a New Competition Tool (NCT) is a measure aiming to rise to this challenge by addressing perceived gaps in the current European Union (EU) competition rules.^[2]This proposal is based on the enforcement experience of both the Commission and National Competition Authorities (NCAs), in addition to a reflection on whether existing competition rules suffice in tackling certain structural competition problems.^[3]

Structural competition problems are those which relate to market characteristics that may have an adverse effect on competition or result in inefficient market outcomes, such as higher prices, lower quality, less choice and innovation.^[4]

The Commission has categorised structural competition issues as structural risks for competition and structural lack of competition.^[5]The former refers to instances where particular market characteristics and the conduct of companies operating in the markets concerned create a threat for competition^[6]. This arises from the creation of powerful market players with an entrenched market position.^[7]The latter refers to instances where the market is not working well or is failing to deliver competitive outcomes due to its structure.^[8]

This paper will consider whether Article 101 TFEU is a suitable and sufficiently effective instrument to address structural competition problems. It will explore relevant types of conduct and situations. It should be noted there is an array of structural competition problems related to Article 101 TFEU that could be discussed, however this paper addresses what the author considers to be the most pertinent issues.

We begin with the context of the Commission’s proposal for the NCT and then discuss oligopolistic markets with a high or substantial risk of tacit collusion, algorithmic-facilitated tacit collusion and platform-driven price-fixing.

The main aim of this paper is to highlight that Article 101 TFEU is largely insufficient and ineffective in addressing these structural competition problems, justifying additional ex-ante regulation in the form of the NCT.

  

Context

Overall, the Commission seeks to introduce the NCT to work alongside its existing enforcement powers under Article 101 TFEU.^[9]This aims to allow the initiation of market investigations without any prior finding and independent of any infringement of competition rules.^[10]It would also enable the Commission to impose remedies unilaterally to address any problems exposed.^[11]

Under this proposal, the Commission would not make findings based on competition law infringements but would envisage allowing markets with particular structural problems to function more efficiently. This would be done through imposing behavioural or structural remedies, despite the absence of an allegation of wrongdoing by any market participant^[12].

One could remark this would represent a “paradigm shift in EU competition law” from ex-post competition enforcement to including a combination of ex-post and ex-ante tools.^[13]In effect, this would result in there being a system where in the absence of violation of competition rules like Article 101 TFEU, the authority could intervene against a certain market structure or failure.

The Commission has yet to enunciate a precise form for the NCT, however in its impact assessment there are four separate options which are being considered^[14], and these will be considered in relation to certain structural competition problems later in this paper. 

Against this contextual backdrop, the focus of this paper will now turn to an evaluation of whether Article 101 TFEU addresses structural competition problems suitably and effectively, in order to clarify whether the NCT proposal is therefore justifiable.

 

I. Oligopolistic Markets with a high or substantial risk of tacit collusion

This section of the paper explores the specific structural competition problem of oligopolistic markets with a high and substantial risk of tacit collusion. It details how Article 101 TFEU is largely inadequate in addressing this situation.

This issue relates to a highly concentrated market where a few large oligopoly firms operate.^[15]The oligopoly firms operating in these markets may be able to behave in a parallel manner and gain benefits from the collective market power they hold, without explicitly entering into an agreement or concerted practice of a kind prohibited by Article 101 TFEU.^[16]This is evident particularly in oligopolies where there is a high level of concentration, similarity of goods and significant market transparency, which means the awareness of each other’s presence is automatically heightened and there is reaction to behaviour as a response to market circumstances.^[17]

A particular issue is that rivals may move together to raise prices or limit production at the same time and same extent without explicit coordination.^[18]This impacts similarly to a cartel, namely due to increased prices, lower choice and innovation, and begs the question of whether Article 101 TFEU is effective in dealing with this problem.

In support of the contention that Article 101 TFEU is ill-equipped to address this situation, Michael Zymler presents an account of why it has not been robust enough.^[19]  He notes that in cases of tacit collusion there is the issue of whether the parallel behaviour as a result of undertakings responding rationally to market conditions actually amounts to a concerted practice.^[20]

This is true when considering the Wood Pulpcase, whereby the test for a concerted practice is that there ought to be reciprocal cooperation through either direct or indirect contact.^[21]It notes that where direct evidence of such conduct is not available, parallel behaviour may demonstrate circumstantial proof of an agreement or concerted practice, but only if it is not of the type which would be anticipated on the market or if there is no other reasonable justification for the conduct.^[22]

This is a meritorious observation because it links to key issues regarding the digital markets that the NCT is seeking to address, with NCAs strongly affirming that digital sectors were most prone to oligopolistic markets with a high or substantial risk of tacit collusion.^[23]This points to the need for there to be ex-ante regulation to complement the ex-post enforcement of Article 101 TFEU in order to keep pace with the developments in the digital economy.

Moreover, Patrick Andreoli-Versbach and Jens-Uwe Franck successfully develop the analysis of Zymler by explaining that the European Court of Justice is reluctant to infer any agreement between competitors from market conduct alone.^[24]Most noteworthy is that in the absence of any direct evidence of collusion, the Court is reluctant to presume any existence of an agreement, even if there has been proven observed parallel conduct as an expression of tacit collusion rather than oligopolistic interdependence.^[25]

This is insightful because it demonstrates there is inherent uncertainty and restrictions stemming from Article 101 TFEU’s treatment of tacit collusion. This emphasises the ineffectiveness of the instrument in addressing this structural competition problem, particularly when it should undeniably deal with tacit collusion as rigorously as Article 101 TFEU does with explicit collusion.

However, it has been noted that Article 101 TFEU has had some effect in dealing with facilitating practices as these can be caught as concerted practices under the instrument.^[26]Avinash B. Amarnath also notes that most academics concur that cases of pure tacit collusion are rare and involve in most cases at least some voluntary efforts or facilitating practices.^[27]

Whilst this has merit, the later discussion on algorithms and their facilitation of tacit collusion will emphasise why Article 101 TFEU is, and will prove to be, ineffective in dealing with structural competition problems relating to tacit collusion in terms of technological developments in the future.

Turning now to whether an NCT could better address this issue, a tool based on the success of market investigations in the United Kingdom clearly has scope to do this. The Commission could conduct in-depth analysis where oligopolistic markets are not working well, and this could enable action to correct features of the market, like the facilitation of tacit collusion, which causes an adverse effect on competition.^[28]

 

II. Algorithmic-facilitated Tacit Collusion

This section of the paper intends to demonstrate that algorithms facilitate structural competition problems in terms of the alignment of prices and less competition, which emphasises Article 101 TFEU is primarily ineffective and unsuitable in addressing this. This section will delve deeper into tacit collusion as explored in the previous section, albeit from the perspective of algorithmic facilitation.

Algorithms may allow companies to align their behaviour, particularly in the instance of retail pricing, without any explicit coordination through reliance on technology.^[29]For example, pricing algorithms are automated and allow for frequent changes to prices.^[30]  A key issue is that of algorithmic tacit collusion, namely the practice of firms in an oligopoly coordinating their actions but without explicit cartel agreement.^[31]

Such an issue pertains to a structural lack of competition because the growing use of algorithms collecting price data leads to an improved market transparency.^[32]With this market transparency, competitors can use pricing algorithms to adjust their prices in real time and this facilitates price monitoring and reacting to these in real time.^[33]

In support of the contention of this paper, Manon van Roozendaal expertly illustrates the way in which algorithms are not, and may not be, sufficiently addressed by Article 101 TFEU. In her paper she distinguishes independent algorithms, where unrelated companies in the same market use an algorithm to determine their prices, from others, such as implementing algorithms.^[34]She notes that whilst competition authorities have found infringements in situations like implementing algorithms, the issue of independent algorithms speaks to tacit collusion which has not been qualified as an infringement of Article 101 TFEU for several years.^[35]

This analysis has merit in consolidating the argument in this paper. In terms of qualifying as an agreement under Article 101 TFEU, tacit collusion generally does not qualify as an agreement because of the lack of concurrence of wills between any parties.^[36]Algorithms allow companies to easily align their behaviour, particularly in terms of retail prices, and this can lead to prices being kept artificially high and result in distorted market conditions.^[37]As such, Article 101 TFEU has, and will have, limited effectiveness in dealing with the anti-competitive nature of algorithms and their increasing prevalence in the digital economy.

Moreover, Ezrachi and Stucke expand on this analysis by noting that the case law accepts that without any proof of express collusion or communication, parallel action and tacit collusion may be the only explanation of the market outcome.^[38]The result of this is higher prices for consumers and yet the law is unable to render the parallelism as illegal.^[39]  This links to their previous work regarding the ineffectiveness of Article 101 TFEU which noted the distinct issue of the “predictable agents” category of collusion, in that each rival firm unilaterally adopts an algorithm that allows for conscious parallelism which monitors price changes and reacts accordingly.^[40]

This emphasises the importance of the previous issue alluded to in the case law regarding agreement, because without evidence of direct or indirect communication between colluders, the enforcement of Article 101 TFEU in these circumstances is challenging and renders it ineffective in dealing with the structural competition problems identified.

In contrast to the contention of this paper and the aforementioned academic literature, the majority of NCA respondents to the Commission’s consultation assert that Article 101 TFEU is suitable and sufficiently effective in addressing algorithms and the structural competition problems they pose.^[41]The contention is that nothing in Article 101 TFEU itself or in case law prevents the Commission from intervening in algorithmic collusion cases and that enforcement could be improved through better technological analysis of algorithmic behaviour.^[42] 

This is a perfectly laudable position, however, the significance of an OECD paper which analysed the effects of algorithms and collusion holds significant weight.^[43]This paper illustrated that whilst there are many pro-competitive aspects linked with automation from algorithms, there is a risk of algorithms distorting digital markets by creating mechanisms to collude which might not have existed otherwise.^[44]The OECD took the position that competition law enforcers should consider their current approach to the legal treatment of tacit collusion under Article 101 TFEU, primarily because algorithms might actually succeed in governing collusive structures better than humans due to their greater accuracy in detecting price changes.^[45]

This OECD paper is noteworthy because it highlights to the author the need for an NCT to deal with the difficulties raised with the current legal standard. As has been demonstrated in previous analysis, the case law itself presents challenges to bringing algorithmic collusion under the Article 101 TFEU framework and this contrasts the majority view of EU NCAs who suggest there is, or will be, little issue with enforcement.

As a result, it is clear Article 101 TFEU does not, and will not, sufficiently address the issue of algorithmic collusion and requires urgent ex-ante action in the form of the NCT, particularly due to the continuous growth of algorithms on the digital marketplace.

Having established thus far how Article 101 TFEU fails to suitably and sufficiently address the structural competition problem of algorithms, it is worth considering the proposition that an NCT could better address this.

The response to consultation from BEUC is compelling, as they strongly advocate the introduction of new measures that can help prevent, rather than belatedly attempt to cure, the resulting harm of issues like algorithmic tacit collusion to consumers.^[46]They believe an NCT is necessary at an EU level to deal with the problematic conduct and structures identified previously in this paper, as this will allow for disruptive and innovative competition from new technologies.^[47]This is insightful because Article 101 TFEU has the intention of preventing competition from diminishing. The market feature of algorithms means there is a greater need to proactively promote increased competition.

However, the analysis of Massimo Motta et al also has some weight. In their paper they note that whilst the NCT is an encouraging approach to generate evidence in terms of whether a certain sector’s decentralised pricing leads to supra-normal prices, the challenging issue of finding effective remedies subsists.^[48]This is certainly true, but this paper contends that the ex-ante NCT would enable the Commission to intervene to make digital markets less prone to tacit collusion and have far more effect than Article 101 TFEU does or could have on its own.

 

III. Platform-Driven Price-Fixing

This section of the paper turns to consider the structural competition problems persistent in sharing economy platforms and emphasises how Article 101 TFEU is ineffective and unsuitable in addressing the key issues which stem from them.

The “sharing economy” refers to the economic activities carried out in markets characterised by the use of internet platforms to facilitate the sale to an individual of some asset owned by another individual.^[49]Uber is an example of such a platform, whereby consumers are connected with individuals using their own personal cars to provide a taxi service. The individual providing the service receives the fair, with commission payable to Uber.^[50]

Nowag has explored how platform-driven price-fixing defies traditional antitrust thinking, particularly with regards to Article 101 TFEU, and how this has given rise to new policy and legal challenges.^[51]A key aspect of his article which underlines how Article 101 TFEU is ineffective is through analysis of the concept of the single economic entity, with there needing to be two or more separate firms or entities or undertakings for the instrument to apply.^[52]For example, the concept of single economic units means that agreements within the same economic unit are not subject to Article 101(1) TFEU.^[53]  This is because a single economic entity only exists where there is a unitary organisation of personal tangible, and intangible elements, which pursue a specific economic aim on a long-term basis.^[54]

This analysis is relevant to this paper’s discussion regarding Article 101 TFEU’s effectiveness because according to the jurisprudence, the sellers on these platforms cannot be considered to form one single company. As a sharing platform’s role is to connect a seller to a buyer, it is unlikely that a sharing economy platform and the seller would be considered a single company for the purposes of competition law.

Furthermore, a situation observed by Nowag whereby price-fixing may be shielded from antitrust intervention, particularly in terms of Article 101 TFEU, is through the labour relationship between the platform and sellers^[55]. This is relevant for discussion of Article 101 TFEU’s effectiveness in dealing with this structural competition problem because whether for example, Uber’s drivers are employees, affects the application of competition law. Nowag’s discussion is complemented by the case of Poucet et Pistrewhich saw the Court rule that employees cannot be considered undertakings according to Article 101 TFEU because they are a part of the company.^[56]

However, Friso Bosteon presents a compelling case that Article 101 TFEU has been effective in dealing with the structural competition problems inherent in the platform economy.^[57]This is achieved through discussion of agency and how it has traditionally been seen as a way to escape the grip of Article 101 TFEU.^[58]He notes that as a principal and agent are deemed as unable to compete, the cartel prohibition in Article 101 TFEU is not applicable to this relationship.^[59]Usefully, he notes that an argument based on agency will not normally be a success for platforms and their peers, and he achieves this through a stellar examination of Uber and its agents. He questions whether the agents and peers bear risk sufficient to meet the criteria to form such a relationship.^[60]

It is undeniable that there are elements of effectiveness with Article 101 TFEU’s treatment of this structural competition problem as identified by Bosteon. However, it is clear the discussion surrounding the single economic entity doctrine highlights fundamental flaws with the way Article 101 TFEU deals with these competition concerns.

Having detailed how Article 101 TFEU partially fails to suitably and sufficiently address the structural competition problems inherent with sharing platforms, it is essential to consider how an NCT could better address this.

This paper contends that an NCT based on market structure could sufficiently address these problems. This is because ex-ante intervention without an infringement identified would be possible. In turn, loopholes present in Article 101 TFEU allowing sharing platforms to fix seller’s prices (as discussed by Nowag) could be plugged by the NCT and negate the anti-competitive effects which arise from these practices.

In addition, the structural competition problem of data accumulation and sharing could also be pursued through the NCT.^[61]This is because sharing economy platforms, in seeking to match offer and supply, have to make available and easily searchable suppliers’ pricing data and commercially sensitive information.^[62]As a result of the increased transparency, this carries the risk of facilitation of anticompetitive exchanges of sensitive information between suppliers and this could raise concerns regarding anti-competitive price signalling or coordination.

 

Conclusion

This paper has considered Article 101 TFEU in terms of its suitability and effectiveness in dealing with structural competition problems.

We have discussed three distinct conduct and situations which Article 101 TFEU has largely failed to sufficiently and effectively address, namely oligopolistic markets with a high risk of tacit collusion, pricing algorithms and their facilitation of tacit collusion and an analysis of platform-driven price-fixing.

In light of this, this paper has provided insight regarding how an NCT could remedy the structural competition problems identified and strongly endorses the general principles underlying the Commission’s proposal.

Overall, the author contends that ex-ante regulation in the form of the NCT, combined with ex-post enforcement of the Article 101 TFEU prohibition, could better remedy the structural competition problems identified.

 

Daniel CORMACK

 

 

[1]Ursula von der Leyen, ‘Margarethe Vestager: Executive Vice-President for a Europe fit for the Digital Age’ (Brussels, 1 December 2019)

<https://ec.europa.eu/commission/commissioners/sites/comm-cwt2019/files/commissioner_mission_letters/mission-letter-margrethe-vestager_2019_en.pdf> accessed 14 November 2020

[2]Commission, Impact Assessment for a possible New Competition Tool(European Commission, 26 October 2020) <https://ec.europa.eu/competition/consultations/2020_new_comp_tool/index_en.html> accessed 14 November 2020

[3]Commission, Questionnaire for the public consultation on a New Competition Tool(European Commission, 2 June 2020) <https://ec.europa.eu/eusurvey/runner/New_Competition_Tool?surveylanguage=EN> accessed 14 November 2020

[4]ibid

[5]ibid

[6]ibid

[7]ibid

[8]ibid

[9]Commission (n 2)

[10]Commission (n 3)

[11]ibid

[12]ibid

[13]Mark Powell et al, ‘The European Commission eyes the addition of a market investigation tool to its 60-year-old toolbox – but is it a chisel or a sledgehammer?’ (Kluwer Competition Law Blog, 5 June 2020) <http://competitionlawblog.kluwercompetitionlaw.com/2020/06/05/the-european-commission-eyes-the-addition-of-a-market-investigation-tool-to-its-60-year-old-toolbox-but-is-it-a-chisel-or-a-sledgehammer/> accessed 14 November 2020

[14]Commission (n 2)

[15]Commission, Summary of the contributions of the National Competition Authorities to the impact assessment of the new competition tool(European Commission) <https://ec.europa.eu/competition/consultations/2020_new_comp_tool/summary_contributions_NCAs_responses.pdf> accessed 14 November 2020

[16]ibid

[17]Michael Zymler, ‘Tacit Collusion: A Solution in Search of a Problem’ (2010) SSRN Electronic Journal <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1751545> accessed 14 November 2020

[18]Commission (n 16)

[19]ibid

[20]ibid

[21]Joined Cases 89/85, C-104/85, C-114/85, C-116/85, C-117/85 and C-125/85 to C-129/85 A. Ahlström Osakeyhtiö and others v Commission of the European Communities [1993] ECR I-1307 

[22]ibid

[23]Commission (n 16)

[24]Patrick Andreoli-Versbach and Jens-Uwe Franck, ‘Econometric Evidence to Target Tacit Collusion in Oligopolistic Markets’ (2015) 11(2) J. Competition Law Econ. <https://academic.oup.com/jcle/article/11/2/463/872320> accessed 14 November 2020

[25]ibid

[26]Avinash B. Amarnath, ‘The Oligopoly Problem: Structural and Behavioural Solutions under Indian Competition Law‘ (2013) 55(3) JILI <https://www.jstor.org/stable/pdf/43953671.pdf?refreqid=excelsior%3A5a7ae9af6c764509a23d8d88e39acd51> accessed 14 November 2020

[27]ibid

[28]Peter Freeman, “Market Investigations and Oligopolistic Markets” in Barry Hawk (ed),   

International Antitrust Law & Policy: Fordham Competition Law 2007(Juris Publishing 2008)

[29]Commission (n 16)

[30]ibid

[31]Gregory S. Crawford et al, ‘An Economic Evaluation of the EC’s proposed “New Competition Tool”’ (European Commission, 16 October 2020) <https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0320680enn.pdf> accessed 14 November 2020

[32]Marcin Mleczko, ‘Feedback from: Marcin Mleczko’ (European Commission, 22 June 2020) <https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12416-New-competition-tool/F525459> accessed 14 November 2020

[33]ibid

[34]Manon van Roozendaal, ‘Algorithms: Teenage Troublemakers of EU Competition Law’ (European Law Institute, 2018) <https://www.europeanlawinstitute.eu/fileadmin/user_upload/p_eli/YLA_Award/Submission_ELI_Young_Lawyers_Award_Manon_van_Roozendaal_2018.pdf> accessed 14 November 2020

[35]ibid

[36]Case T-41/96 Bayer AG v Commission[1996] ECR II-3378, 67 and 176

[37]ibid

[38]Ariel Ezrachi and Maurice E Stucke, ‘Sustainable and Unchallenged Algorithmic Tacit Collusion’ (2020) 17(2) Nw J Tech & Intell Prop 217 <https://scholarlycommons.law.northwestern.edu/njtip/vol17/iss2/2> accessed 14 November 2020

[39]ibid

[40]Ariel Ezrachi and Maurice E Stucke, Virtual Competition: The Promise and Perils of the Algorithm-Driven Economy(Harvard University Press 2016)

[41]Commission (n 16)

[42]ibid

[43]OECD, Algorithms and Collusion: Competition Policy in the Digital Age (June 2017)

<www.oecd.org/competition/algorithms-collusion-competition-policy-in-the-digital-age.htm> accessed 14 November 2020

[44]ibid

[45]ibid

[46]BEUC, Digital Services (Ex Ante Rules) and New Competition Tool(7 October 2020) <https://www.beuc.eu/publications/beuc-x-2020-079_digital_services_act_and_new_competition_tool.pdf> accessed 14 November 2020

[47]ibid

[48]Massimo Motta et al, ‘Expert Advice for the Impact Assessment of a New Competition Tool’ (European Commission, October 2020) <https://ec.europa.eu/competition/consultations/2020_new_comp_tool/kd0420575enn.pdf>accessed 14 November 2020

[49]Guy Lougher et al, ‘EU Competition Law in the Sharing Economy’ (2015) 7(2) J. Competition Law Econ. <https://academic.oup.com/jeclap/article/7/2/87/2357789> accessed 14 November 2020

[50]ibid

[51]Julian Nowag, ‘When sharing platforms fix sellers’ prices’ (2018) 6(3) J. Antitrust Enforc <https://academic.oup.com/antitrust/article/6/3/382/4999859> accessed 14 November 2020

[52]ibid

[53]ibid

[54]Case T-112/05 Akzo Nobel NV v Commission[2007] ECR II-5049, 57-58

[55]Julian Nowag (n 52)

[56]Joined Cases C-159/91 and C-160/91 Christian Poucet v Assurances Générales de France and Caisse Mutuelle Régionale du Languedoc-Roussillon [1993] ECR I 637

[57]Friso Bostoen, ‘Competition Law in the Peer-to-Peer Economy’ in The Platform Economy(Intersentia, 2019) <https://ssrn.com/abstract=33299802> accessed 14 November 2020

[58]ibid

[59]ibid

[60]ibid

[61]Guy Lougher et al (n 50)

[62]ibid