Skip to main content

Advertisement

SpringerLink
Log in

Pricing Algorithms as Collusive Devices

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

Abstract

This article undertakes a critical review of the prospect that reinforced-learning, or more loosely described machine-based, pricing algorithms will lead to widespread collusion without human involvement. There is no evidence, no antitrust case, and no strong theoretical and practical reasons for this belief. Notwithstanding this dearth of evidence, legal commentators continue to insist that algorithmic collusion poses a significant threat and that there is a “gap” in antitrust laws that needs to be filled. However, a reasonable assessment suggests that EU antitrust law and its enforcement are adaptable enough to deal with the cases that may arise.

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

Similar content being viewed by others

Notes

  1. Margrethe Vestager, Speech at the Bundeskartellamt 18th Conference on Competition, Berlin, 16 March 2017.

  2. Li and Xie (2018), p. 2.

  3. Mehra (2016).

  4. Ezrachi and Stucke (2016, 2017a).

  5. Gal (2019); Gal and Elkin-Koren (2017).

  6. Ezrachi and Stucke (2016), p. 233.

  7. Ezrachi and Stucke (2020).

  8. Ezrachi and Stucke (2017a), p. 1775.

  9. Mehra (2016), p. 1340.

  10. Ezrachi and Stucke (2017b), p. 233.

  11. Ittoo and Petit (2017); Johnson and Sokol (2019); Schwalbe (2018); Tesauro and Kephart (2002); Schrepel (2017).

  12. Cormen et al. (2001).

  13. CMA (2018), p. 2.4.

  14. A supervised learning algorithm predicts outcomes based on the available data and optimizes some function that best explains the relationship between the variables in the dataset. Multiple regression analysis is a supervised learning algorithm.

  15. An unsupervised learning algorithm uses unlabelled data with the objective of finding a descriptive function to structure the dataset based on some undetermined pattern. A search engine uses unsupervised learning algorithms to organize data into different categories.

  16. DLC/BKartA Report (2019), s. II. Also, Schwalbe (2018); CMA (2018); Crandall et al. (2018); and Deng (2017).

  17. Sutton and Barto (2018).

  18. Calvano et al. (2020), p. 6.

  19. Ohlhausen (2017).

  20. ADLC/BKartA Report (2019), p. 15.

  21. OECD (2017a).

  22. Ezrachi and Stucke (2020), p. 226.

  23. Calvano et al. (2019), p. 169.

  24. European Commission (2017).

  25. AdC (2019); ADLC/BKart (2019); CMA (2018); Cremer et al. (2019). Also, OECD (2017a; 2017b).

  26. CMA (2018).

  27. OECD (2017b), p. 52.

  28. Furman Report (2019), p. 15.

  29. ADLC/BKartA (2019), p. 52.

  30. Vestager, op cit.

  31. Decision of the Competition and Markets Authority, Online sales posters and frames, Case 50223, 12 August 2016. See also the similar US “posters” decision often cited United States of America v. David Topkins, Plea Agreement, Department of Justice, Antitrust Division, No. CR 15-00201 WHO.

  32. Case C-74/14, Eturas’ UAB and Others v. Lietuvos Respublikos konkurencijos taryba EU:C:2016:42, Judgment of the Court, 21 January 2016. Another example of the use of algorithms to facilitate collusion is United States v. Airline Tariff Publ’g Co., 836 F. Supp. 9 (D.D.C. 1993).

  33. European Commission Press Release, “Antitrust: Commission fines four consumer electronics manufacturers for fixing online resale prices”, Cases AT.40465 (Asus), AT.40469 (Denon and Marantz), AT.40181 (Philips) and AT.40182 (Pioneer), decisions of 24 July 2018.

  34. Decision of the Competition and Markets Authority, Digital piano, and digital keyboard sector: anti-competitive practices 50565-2 (nonconfidential), 8 October 2019. https://www.gov.uk/cma-cases/musical-instruments-and-equipment-suspected-anti-competitive-agreements-50565-2#non-confidential-infringement-decision.

  35. OFGEM decision, Infringement by Economy Energy, E (Gas and Electricity) and Dyball Associates of Chapter I of the Competition Act 1998 with Respect to an Anticompetitive Agreement, 26 July 2019.

  36. Ezrachi and Stucke (2017a), p. 71.

  37. For an extensive review of the computer sciences and economic literature, see Schwalbe (2018); and Gata (2019).

  38. An early study by Waltman and Kaymak (2008) found that independent machine learning algorithms generally converge to seemingly collusive outcomes. Zhou et al. (2018) show collusion with one algorithm and human players.

  39. Calvano et al. (2020).

  40. Klein (2021).

  41. Salcedo (2015).

  42. Salcedo (2015), p. 76.

  43. Abada and Lambin (2020).

  44. Harrington (2018).

  45. Brown and MacKay (2020); Hansen et al. (2020). See also Chen et al. (2016).

  46. Assad et al. (2020).

  47. An example is Gal’s claim: “The possibility of collusion is real. This has been proven both theoretically and empirically”, referring to Calvano et al. (2020). Gal MS (2020).

  48. Calvano et al. (2020) and Klein (2021).

  49. Descamps et al. (2021).

  50. All the above papers are based on Bertrand-duopoly models where firms decide on prices.

  51. Tesauro and Kephart (2002).

  52. ADLC/BKartA (2019), p. 21.

  53. ADLC/BKartA (2019), pp. 50–51.

  54. I owe this point to Akos Reger.

  55. Calvano et al. (2020).

  56. Deng (2019a). Also, Deng (2018).

  57. I owe these points to Timo Klein.

  58. CMA (2018), p. 49.

  59. Foster (2018).

  60. Miklós-Thal and Tucker (2019).

  61. Ezrachi and Stucke (2017a, b), Chap. 10.

  62. Their quotations with the quoted text come largely from the European Commission’s (2004) horizontal merger guidelines.

  63. Ezrachi and Stucke (2020).

  64. Ezrachi and Stucke (2020), p. 226.

  65. Ezrachi and Stucke (2020), p. 227.

  66. Ezrachi and Stucke (2020), p. 228.

  67. Ezrachi and Stucke (2020), p. 241.

  68. Posner (1969); Kaplow (2011).

  69. Beneke and Mackenrodt (2019).

  70. Deng (2019b).

  71. Vestager, op cit.

  72. Vestager, op cit.

  73. EU:C:1993:120.

  74. Veljanovski (2020), pp. 7.08–7.09; Castillo de la Torre and Fournier (2017), para. 4.117.

  75. Kone AG v. OBB EU:C:2014:45. Also, AG Kokott’s Opinion of 30 January 2014 in Kone v OBB. See generally Veljanovski (2020), Chap. 8.

  76. CMA (2018), pp. 13–14.

  77. Harrington (2018).

  78. Ezrachi and Stucke (2020), pp. 257–258.

  79. CMA appoints Stefan Hunt to top digital role. Press Release 18 May 2018. https://www.gov.uk/government/news/cma-appoints-stefan-hunt-to-top-digital-role.

  80. Case COMP 38589, Heat Stabilisers (2009); Case C-194/14 P, AC-Treuhand AG v. Commission EU:C:2015:717; Case T-180/15, Icap v Commission EU:T:2017:795.

  81. OFGEM decision, Infringement by Economy Energy, E (Gas and Electricity) and Dyball Associates of Chapter I of the Competition Act 1998 with Respect to an Anticompetitive Agreement, 26 July 2019.

References

  • Abada I, Lambin X (2020) Artificial intelligence: can seemingly collusive outcomes be avoided?

  • AdC (2019) Digital ecosystems, big data and algorithms – issues paper. Autoridade da Concorrência, Lisbon

    Google Scholar 

  • ADLC/BKartA (2019) Algorithms and competition. Joint report of the Autorité de la Concurrence and Bundeskartellamt

  • Assad S, Clark R, Ershov D, Xu L (2020) Algorithmic pricing and competition: empirical evidence from the German retail gasoline market. CESifo Working Paper 8521

  • Beneke F, Mackenrodt M-O (2019) Artificial intelligence and collusion. IIC Int Rev Intell Prop Competence Law 50:109–134. https://doi.org/10.1007/s40319-018-00773-x

    Article  Google Scholar 

  • Brown ZY, MacKay A (2020) Competition in pricing algorithms. Am Econ J Micro (forthcoming)

  • Calvano E, Calzolari G, Denicolò V, Pastorello S (2019) Algorithmic pricing what implications for competition policy? Rev Indus Org 55:155–171. https://doi.org/10.1007/s11151-019-09689-3

    Article  Google Scholar 

  • Calvano E, Calzolari G, Denicolò V, Pastorello S (2020) Artificial intelligence, algorithmic pricing and collusion. Am Econ Rev 110:3267–3297. https://doi.org/10.1257/aer.20190623

    Article  Google Scholar 

  • Chen L, Mislove A, Wilson C (2016) An empirical analysis of algorithmic pricing on Amazon Marketplace. In: WWW ’16: Proceedings of the 25th International Conference on World Wide Web, April 2016, pp 1339–49. https://doi.org/10.1145/2872427.2883089

  • CMA (2018) Pricing algorithms—economic working paper on the use of algorithms to facilitate collusion and personalised pricing. Competition and Market Authority, CMA94, 8 October 2018

  • Cormen TH, Leiserson CE, Rivest RL, Stein C (2001) Introduction to algorithms, 2nd edn. The MIT Press

    Google Scholar 

  • Crandall JW et al (2018) Cooperating with machines. Nat Commun 9:233. https://doi.org/10.1038/s41467-017-02597-8

    Article  Google Scholar 

  • Crémer J, de Montjoye Y-A, Schweitzer H (2019) Competition policy for the digital era: final report. European Commission

  • Deng A (2017) When machines learn to collude: lessons from a recent research study on artificial intelligence. Available at: SSRN https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3029662

  • Deng A (2018) What do we know about algorithmic tacit collusion? Antitrust 33(Fall):88–95

    Google Scholar 

  • Deng A (2019a) Algorithmic tacit collusion is a limited threat to competition. Law 360

  • Deng A (2019b) From the dark side to the bright side: exploring algorithmic antitrust compliance. Available at: SSRN https://ssrn.com/abstract=3334164

  • Descamps A, Klein T, Shier G (2021) Algorithms and competition: the latest theory and evidence. Comp LJ 20:32–39

    Google Scholar 

  • European Commission (2004) Guidelines on the assessment of horizontal mergers under the Council Regulation on the control of concentrations between undertakings. 2004 O.J. (C 31)

  • European Commission (2017) Final report on the e-commerce sector inquiry. COM(2017) 229 final, 10 May 2017

  • Ezrachi A, Stucke ME (2016) Virtual competition: the promise and perils of the algorithm-driven economy. Harvard University Press

    Book  Google Scholar 

  • Ezrachi A, Stucke ME (2017a) Artificial intelligence and collusion: when computers inhibit competition. Univ Illinois LR:1776–1810

  • Ezrachi A, Stucke ME (2017b) Algorithmic collusion: problems and counter-measures—note, OECD Roundtable on algorithms and collusion. 21–23 June 2017. DAF/COMP/WD(2017)25. Accessed 14 Mar 2022

  • Ezrachi A, Stucke ME (2020) Sustainable and unchallenged algorithmic tacit collusion. Nw J Tech. Intell. Prop. 17:217–259

    Google Scholar 

  • Foster D (2018) Algorithms and price collusion—learning to love artificial intelligence. Competition Law Insights, May 2018

  • Furman Report (2019) Unlocking digital competition—report of the UK Digital Competition Expert Panel, March 2019

  • Gal MS (2019) Algorithms as facilitating practices. Berkeley Tech LJ 34:67–118. https://doi.org/10.15779/Z38VM42X86

    Article  Google Scholar 

  • Gal MS (2020) Algorithms and competition law: interview of Michal Gal by Thibault Schrepel. E-Competitions Interviews Series

  • Gal MS, Elkin-Koren N (2017) Algorithmic consumers. Harv J Law Tech 30:309–352

    Google Scholar 

  • Gata JE (2019) Controlling algorithmic collusion: short review of the literature, undecidability, and alternative approaches. Revista de Concorrência e Regulação, No. 39, July 2019

  • Hansen K, Misra K, Pai M (2020) Algorithmic collusion: supra-competitive prices via independent algorithms. CEPR Discussion Paper No. DP14372

  • Harrington JE (2018) Developing competition law for collusion by autonomous artificial agents. J Comp Law Econ 14:331–363. https://doi.org/10.1093/joclec/nhy016

    Article  Google Scholar 

  • Ittoo A, Petit N (2017) Algorithmic pricing agents and tacit collusion: a technological perspective. Working paper. Available at: SSRN https://ssrn.com/abstract=3046405

  • Johnson J, Sokol D (2019) Understanding AI collusion and compliance. In: Sokol D, van Rooij B (eds) Cambridge handbook of compliance. Cambridge University Press, Cambridge

    Google Scholar 

  • Kaplow L (2011) An economic approach to price fixing. Antitrust LJ 77:343–449

    Google Scholar 

  • Klein T (2021) Autonomous algorithmic collusion: q-learning under sequential pricing. RAND J Econ 52:538–558. https://doi.org/10.1111/1756-2171.12383

    Article  Google Scholar 

  • Li S, Xie CC (2018) Automated pricing algorithms and collusion: a brave new world or old wine in new bottles? The Antitrust Source Dec: 1–9. https://www.nera.com/content/dam/nera/publications/2019/Algorithmic%20Collusion%20-%20Li%20and%20Xie.pdf. Accessed 14 Mar 2022

  • Mehra SK (2016) Antitrust and the robo-seller: competition in the time of algorithms. Minn LR 100:1323–1335

    Google Scholar 

  • Miklós-Thal J, Tucker C (2019) Collusion by algorithm: does better demand prediction facilitate coordination between sellers? Manag Sci 65:1552–1561. https://doi.org/10.1287/mnsc.2019.3287

    Article  Google Scholar 

  • OECD (2017a) Algorithms and collusion—background note by the Secretariat. DAF/COMP(2017)4. Accessed 14 Mar 2022

  • OECD (2017b) Algorithms and collusion: competition policy in the digital age. 14 Sept 2017. Available at: www.oecd.org/competition/algorithms-collusion-competition-policy-in-the-digital-age.htm

  • Ohlhausen MK (2017) Should we fear the things that go beep in the night? Some initial thoughts on the intersection of antitrust law and algorithmic pricing. Remarks from the Concurrences Antitrust in the Financial Sector Conference, New York, 23 May 2017

  • Posner RA (1969) Oligopoly and the antitrust laws: a suggested approach. Stanf LR 21:1562–1606

    Article  Google Scholar 

  • Salcedo B (2015) Pricing algorithms and tacit collusion. Working paper. Penn Stat Univ

  • Schrepel T (2017) Here’s why algorithms are not (really) a thing. Revue Concurrentialiste 15 May 2017. Available at: https://leconcurrentialiste.co

  • Schwalbe U (2018) Algorithms, machine learning, and collusion. J Comp Law Econ 14:568–607. https://doi.org/10.1093/joclec/nhz004

    Article  Google Scholar 

  • Sutton RS, Barto AG (2018) Reinforcement learning—an introduction, 2nd edn. The MIT Press, Cambridge

    Google Scholar 

  • Tesauro G, Kephart JO (2002) Pricing in agent economies using multi-agent q-learning. Auton Agent Multi-Agent Syst 5:289–304. https://doi.org/10.1023/A:1015504423309

    Article  Google Scholar 

  • Veljanovski C (2020) Cartel damages—principles, measurement and economics. Oxford University Press, Oxford

    Google Scholar 

  • Waltman L, Kaymak U (2008) Q-learning agents in a Cournot oligopoly model. J Econ Dyn Control 32:3275–3293. https://doi.org/10.1016/j.jedc.2008.01.003

    Article  Google Scholar 

  • Zhou N, Zhang L, Li S, Wang Z (2018) Algorithmic collusion in Cournot duopoly market: evidence from experimental economics. arXiv:1802.08061v1 [econ.EM], 21 February 2018

Download references

Author information

Authors and Affiliations

  1. Managing Partner, Case Associates; Fellow in Law and Economics, Institute of Economic Affairs, London, UK

    Cento Veljanovski

Authors
  1. Cento Veljanovski
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Cento Veljanovski.

Additional information

Publisher's Note

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

The author is grateful to Timo Klein, Aurelien Portuese, Thibault Schrepel, Akos Reger and anonymous referees for comments on an earlier draft. The views expressed here are solely those of the author.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Veljanovski, C. Pricing Algorithms as Collusive Devices. IIC 53, 604–622 (2022). https://doi.org/10.1007/s40319-022-01177-8

Download citation

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s40319-022-01177-8

Keywords

Search

Navigation