Skip to main content

Advertisement

SpringerLink
Log in

The Applicability of Art. 101 TFEU to Horizontal Algorithmic Pricing Practices: Two Conceptual Frontiers

  • Article
  • Published:
IIC - International Review of Intellectual Property and Competition Law Aims and scope Submit manuscript

Abstract

As the use of pricing algorithms in electronic commerce has become ubiquitous, competition authorities have become alert to the potential anti-competitive outcomes that might arise from the use of such computerised tools by market players, particularly those who compete directly against each other in digital markets. This paper seeks to explore the extent to which existing European competition law principles are applicable to market players who pursue algorithmic pricing strategies to achieve sub-competitive parallel pricing outcomes, diminishing the degree of price competition that might otherwise exist between them. In particular, we will focus on situations where no agreement exists between the market players, starting from the premise that their actions are non-collusive in the sense that they are not implementing some sort of cartel arrangement via electronic means. The interesting legal question which emerges is whether, in the absence of a specific anti-competitive agreement, plan or scheme that has been concluded between competing undertakings, an infringement of the existing competition law rules can nevertheless be established against those whose deliberate algorithmic pricing activities produce stifling effects on price competition in online marketplaces, resulting in harms to competition similar to what they would have achieved had they actually reached such an agreement beforehand. Two conceptual frontiers will be scrutinised in greater detail: (1) whether or not the concept of a “concerted practice” can be applied to such algorithmic pricing conduct; and (2) whether or not a credible theory of competitive harm can be constructed in such scenarios.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1.
Fig. 2.

Similar content being viewed by others

Notes

  1. Vestager (2017); OECD (2017), pp. 18–32.

  2. There are two infamous antitrust cases involving the use of price-setting algorithms by sellers using Amazon’s e-commerce marketplace: (1) how algorithmic pricing was used to push the selling price of a book, “The Making of a Fly”, to over US$23million (Solon 2011); (2) how specific pricing algorithms were adopted by a group of sellers of posters to implement a pricing agreement between them (United States v David Topkins 2015); see also parallel case in the United Kingdom: Competition & Markets Authority (2016).

  3. For example, see OECD (2017), p. 36; in the US context, see Harrington (2018), pp. 358–359.

  4. For example, see Bourreau and Streel (2019), pp. 7–8; Han et al (2018), p. 9.

  5. See Honoré and Fabre (2019), pp. 42–43.

  6. Ezrachi and Stucke (2016).

  7. For example, see OECD (2017), pp. 18–19; Calvano et al. (2019).

  8. The difficulties with using the “collusion” label, in this context, have been manifested elsewhere by the widespread use of the phrase “tacit collusion” by economists to describe the supra-competitive pricing outcomes associated with oligopolies – the same sort of outcomes which algorithmic usage might facilitate. See Whish and Bailey (2018), p. 573.

  9. Harrington (2018), p. 7.

  10. Calvano (2019), p. 168.

  11. As illustrated, for example, in the cases brought by the US and UK antitrust authorities against a group of online sellers of posters. See supra note 2. See also Ezrachi and Stucke (2017), p. 1782, where the authors describe this category of behaviour at the extreme end of the spectrum of illicit conduct – “The Computer as Messenger” – where computers simply “execute the will of humans … to assist in implementing, monitoring and policing the cartel”.

  12. For a brief discussion of regulatory challenges posed by algorithmic decision-making, see Bayern (2018), p. 144.

  13. For a brief discussion of the anti-competitive potential of “self-learning” or “deep learning” algorithms, see OECD (2017), at [4.3.4].

  14. Such algorithms might conceivably deploy technologies similar to those used by price-comparison websites used by e-commerce consumers which gather pricing information from multiple online sources, putting online sellers in a position to dynamically adjust their prices in response to real-time changes in the prices set by other online market players.

  15. See OECD (2017), pp. 14–17.

  16. See OECD (2017), at [4.3.2].

  17. OECD (2017), at [4.3.1] and [4.3.3].

  18. Bayer AG v Commission of the European Communities (ADALAT), Case T-41/96 EU:T:2000:242, para. 69; BAI and Commission v Bayer, Case C-2/01 P EU:C:2004:2, para. 97.

  19. These are instances of “any direct or indirect contact between [undertakings] … to influence the conduct on the market of an actual or potential competitor or to disclose to such a competitor the course of action which [an undertaking has] decided to adopt or contemplate adopting on the market”. See Cooperatieve Vereniging Suiker Unie v Commission of the European Communities, Case 40/73 EU:C:1975:174, para. 174.

  20. T-Mobile Netherlands BV, KPN Mobile NV, Orange Nederland NV and Vodafone Libertel NV v Raad van bestuur van de Nederlandse Mededingingsautoriteit, Case C-8/08 EU:C:2009:343, para. 33.

  21. T-Mobile, supra note 20, paras. 51–53.

  22. Eturas UAB and Others v Lietuvos Respublikos konkurencijos taryba, Case C-74/14 EU:C:2016:42, paras. 36 and 37.

  23. Eturas, supra note 22, para. 44. The onus is on the undertakings to rebut these presumptions by showing, for example, that they publicly distanced themselves from the concerted practice by clearly objecting to it, reporting the practice to the relevant authority or pointing to some other objectively verifiable indicator of their non-participation.

  24. It should be borne in mind that this involves an assessment of the evidence, which is a matter governed by national law, with reference to the principles of procedural autonomy, effectiveness and equivalence. What the CJEU decided in Eturas was that the mere dispatch of an electronic message concerning restraints on price competition between travel agencies may constitute sufficient evidence from which a concerted practice may be inferred – so long as other objective and consistent indicia are considered alongside this piece of evidence to support a rebuttable finding that the recipients of the message were aware of its contents. See Rusu (2016), p. 397.

  25. Imperial Chemical Industries Ltd. v Commission of the European Communities, Case 48/69 EU:C:1972:70, para. 66. More recently, the General Court has observed that when the parallel behaviour of firms is solely relied upon as proof of a concerted practice, the Commission has to discharge the legal burden of proving why the alternative explanations given by the parties for the parallel behaviour are implausible. See CISAC v Commission, Case T-442/08 EU:T:2013:188, paras. 96–102, and 137.

  26. ICI (Dyestuffs), supra note 25, para. 64; T-Mobile, supra note 20, para. 26.

  27. European Commission (2011), Guidelines on the applicability of Article 101 of the Treaty of the Functioning of the European Union to horizontal cooperation agreements, [2011] OJ 2011/C 11/1, at Recital 61. Reference was made to Cimenteries CBR SA v Commission, Joined Cases T-25/95 etc EU:T:2000:77, and T-Mobile, supra note 20, para. 26 which illustrate how mere attendance at a meeting where one undertaking discloses confidential price information to its competitors can amount to a concerted practice even in the absence of an explicit price-fixing agreement.

  28. This includes information related to prices, including “actual prices, discounts, increases, reductions or rebates … customer lists, product costs, quantities, turnovers, sales, capacities, qualities, marketing plans, risks, investments, technologies and R&D programmes and their results”. European Commission (2011), supra note 27, at Recital 86.

  29. European Commission (2011), supra note 27, at Recital 61.

  30. European Commission (2011), supra note 27, at Recital 62.

  31. ICI (Dyestuffs), supra note 25, and at paras. 64 and 65.

  32. The competition authority bears the burden of proving that there are no plausible alternative explanations for the parallel behaviour, such as the oligopolistic nature of the market where parallel pricing is expected, to support an inference that concertation between the parties was the real cause for the parallelism. See Cimenteries CBR SA, supra note 27, and T-Mobile, supra note 20.

  33. See Cimenteries CBR SA, supra note 27, and T-Mobile, supra note 20.

  34. In the absence of a specific decision from the European courts, the current state of the law is not entirely certain as to exactly which forms of price-signalling conduct can give rise to a concerted practice, given that it could reduce “strategic uncertainty” between competitors in the sense recognised by the European Commission (2011), supra note 27.

  35. See Grillo (2002), where the author argues for greater convergence between the legal and economic perspectives of anti-competitive conduct in the context of oligopolies, with the former traditionally focusing primarily on the coordinative nature of the conduct of the parties, and the latter paying closer attention to the harmful economic outcomes that result from their actions.

  36. Tesco Stores Ltd v Office of Fair Trading, CAT 31 (2012), para. 56. The UK Competition Appeal Tribunal held, at para. 487, that there was an infringement of the Chapter I prohibition of the UK Competition Act when a concerted practice was established between competing retailers disclosing and transmitting their future pricing intentions to each other via a common supplier.

  37. Such as when an electronic message is sent, via personal mailboxes on a digital platform, to competitors informing them of uniform maximum discounting practices to be applied to all their customers – which was the conduct prosecuted by the Lithuanian Competition Council in Eturas, supra note 22.

  38. Ezrachi and Stucke (2020), p. 222, responding to Schwalbe (2019), p. 600.

  39. See Honoré and Fabre (2019), p. 47, who express concern about the legal uncertainty associated with calls made by other commentators to “reconsider some traditional competition law concepts”.

  40. To the extent that there may be pro-competitive effects associated with the use of algorithmic pricing tools as well, these will of course have to be balanced, within the Art. 101(3) TFEU framework (as explained by the General Court in Van den Bergh Foods v Commission, Case T-65/98 EU:T:2003:281, para. 107), against the harm to competition upon which any infringement of Art. 101 TFEU is premised.

  41. Groupement des Cartes Bancaires v European Commission, Case C-67/13P EU:C:2014:2204, para. 58.

  42. Even though it must be acknowledged that there is both case law and commentary to support the view that engaging in anti-competitive “facilitating practices” can be challenged under European competition law as by-object infringements of Art. 101 TFEU. See UK Agricultural Tractor Registration Exchange (1992), infra note 48; and Grillo (2002), pp. 160–161.

  43. O2 (Germany) GmbH & Co, OHG v Commission, Case T-328/03 EU:T:2006:116, para. 68.

  44. See supra note 20 and accompanying text.

  45. Posner (1969).

  46. Whish and Bailey (2018), p. 572.

  47. OECD (2017), p. 35.

  48. UK Agricultural Tractor Registration Exchange, OJ L68/19 (1992), paras. 37–38, which found that there was a “prevention of hidden competition in a highly concentrated market”. This was subsequently upheld on appeal to the General Court in Fiatagri and New Holland Ford v Commission, Case T-34/92 EU:T:1994:258, para. 91, which found that “exchanges of precise information at short intervals … on a highly concentrated oligopolistic market … and on which competition is as a result already greatly reduced and exchange of information facilitated, likely to impair considerably the competition which exists between traders. In such circumstances, the sharing, on a regular and frequent basis, of information concerning the operation of the market has the effect of periodically revealing to all the competitors the market positions and strategies of the various individual competitors”. On further appeal, the Court of Justice in John Deere v Commission, Case C-7/95 P EU:C:1998:256, paras. 88–90, found that the Court of First Instance had taken into account the nature of the information exchanged, the frequency with which it was disseminated and of the persons to whom it was disclosed, and as such the Court of First Instance “must be considered to have concluded correctly that the information exchange system reduces or removes the degree of uncertainty as to the operation of the market and that the system is therefore liable to have an adverse influence on competition between manufacturers”.

  49. OECD (2017), at [4.2.1]. See OECD (2017), p. 24, where the authors argue that “[d]espite the apparently ambiguous effects, algorithms appear to have changed more substantially the structural characteristics that raise competition concerns, namely market transparency and frequency of interaction” and construct an economic model to support their conclusion that “if markets are sufficiently transparent and firms can adjust their decisions very fast, for instance by changing prices in real time, collusion is always sustainable … as the combination of perfect market transparency with very frequent interaction entirely eliminates the profitability of deviations, which can be easily identified and immediately retaliated”.

  50. European Commission (2011), supra note 27, at Recital 77.

  51. European Commission (2004), Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings, [2004] OJ 2004/C 31/03, at Recital 39.

  52. Supra note 51, at Recital 41.

  53. See supra note 48, John Deere v Commission, paras. 86–87. Honoré and Fabre make a similar observation in relation to the settled case law in the Wood Pulp case, arguing that “the immediate effect of a pricing algorithm may be to invert the paradigm for which the Wood Pulp case was the exception” since the level of price transparency which “can … be readily exploited by pricing algorithms” is so much higher in contemporary online markets than at the time when that case was decided. See Honoré and Fabre (2019), p. 46.

  54. See ICI (Dyestuffs), supra note 25, and at paras. 66–67, where anti-competitive conduct is established through “strong evidence of [infringing conduct] if it leads to conditions of competition which do not correspond to the normal conditions of the market, having regard to the nature of the products, the size and number of the undertakings” or by demonstrating that the conduct of the parties enabled them to “stabilize prices at a level different from that to which competition would have led”.

  55. See, for example, the market studies conducted by the Competition Commission of India in 2018 (Airline ticket pricing) and 2019 (E-commerce industry in India), available online at <https://www.cci.gov.in/sites/default/files/whats_newdocument/Market-study-on-e-Commerce-in-India.pdf>. Similarly, the Competition Commission of Singapore has conducted a market study of the online travel booking industry in Singapore between 2018 and 2019. See <https://www.cccs.gov.sg/media-and-consultation/newsroom/media-releases/launch-event-of-cccs-9-apr-18> (all websites last accessed 9 October 2020).

  56. See, for example, the cooperative efforts of the French and German competition authorities in developing a project to study the “typology of algorithms” and their impact on competition. See <https://www.bundeskartellamt.de/SharedDocs/Meldung/EN/Pressemitteilungen/2018/19_06_2018_Algorithmen.html> (last accessed 9 October 2020).

  57. See, for example, the European Union’s Final Report on the E-commerce Sector Inquiry (2017), which recognised the potential price-coordination problems arising from e-commerce market players using software programmes to engage in automated price-tracking and price-adjustment practices. See <https://ec.europa.eu/competition/antitrust/sector_inquiry_final_report_en.pdf> (last accessed 9 October 2020).

References

Download references

Author information

Authors and Affiliations

  1. Associate Professor, Faculty of Law, National University of Singapore, Singapore, Singapore

    Burton Ong

Authors
  1. Burton Ong
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Burton Ong.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

I am grateful for the research assistance of Kenji Lee and Priscilla Seah in the preparation of this article, as well as the feedback I received from participants at the “Multidisciplinary Perspectives on Algorithms Regulation, Governance, Markets” conference in Kyushu University organised by Steven van Uytsel. I would also like to acknowledge the helpful comments of an anonymous reviewer. Any remaining errors are mine alone.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Ong, B. The Applicability of Art. 101 TFEU to Horizontal Algorithmic Pricing Practices: Two Conceptual Frontiers. IIC 52, 189–211 (2021). https://doi.org/10.1007/s40319-021-01016-2

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s40319-021-01016-2

Keywords

Search

Navigation