By Aman Rizvi
Introduction
The European Commission unveiled its ambitious Digital Services Act (DSA) and Digital Markets Act (DMA) proposals in late 2020.[1] They included a list of dos and don’ts for platforms that act as dominant gatekeepers, and new consumer protection measures like the right to challenge content moderation decisions.[2] But they omitted a controversial and far-reaching idea that the Commission had floated during stakeholder consultations last summer: the New Competition Tool (NCT).
The NCT belongs to a category of competition enforcement powers often called ‘market investigation tools’. It would have allowed the Commission—specifically its Directorate-General for Competition (DG COMP)— to impose structural or behavioural remedies on firms based on concerns about the structure of a market.[3] This contrasts with existing law, under which remedies could only be used to penalise unlawful anticompetitive conduct or as a condition for approving a proposed merger.
Market investigation tools like the NCT have great potential to bridge the gap between antitrust doctrine and concerns about growing concentration of economic power[4] by addressing harms to competition that are hard to define but easy to recognise. They could dislodge stubbornly entrenched monopolies,[5] and could—in digital markets especially—be used to tackle problems like tacit collusion or market tipping.[6] They could also facilitate enforcement against harms that, although covered by existing laws, could otherwise be addressed only through long and complicated processes or would be too hard to detect.[7]
But giving regulators such power is risky. At worst, a market investigation tool could allow competition authorities to dictate the structure of markets by fiat. On the other hand, when such tools operate under legal uncertainty, competition authorities may be timid in using them. This would hobble the tool’s effectiveness and frustrate expectations.
Competition authorities in six countries have market investigation powers: the United Kingdom, Mexico, Iceland, Greece, Italy, and Romania.[8] Some of these countries’ experiences suggest the tool works best with strong procedural safeguards and under a clear legal framework. They show that procedural due process is especially important in mitigating the risk of officials using the tool to micromanage the economy. The UK, whose market investigation tool involves extensive procedural safeguards, shows us what kinds of safeguards work and what can be achieved with them.[9] Mexico’s experience, meanwhile, reveals the importance of legal certainty by illustrating how its absence undermines the tool.
In light of these experiences, it seems that the EC was right to shelve the NCT. DG COMP and the Courts of Justice of the European Union (CJEU) are not equipped to provide the procedural safeguards that make the UK’s tool effective. And the legal basis that the EC had outlined for the NCT in the Treaty on the Functioning of the European Union (TFEU) was shaky, suggesting it could have been undermined like Mexico’s.
On its own, that conclusion might not seem especially interesting: the EU should not do something that it has already chosen not to do. But given the real benefits of market investigation tools, it is still worth exploring why the EU was ill-suited to adopt one. The EU, or another jurisdiction, might decide in the future that those benefits are worth pursuing. It is therefore worthwhile to consider what such a proposal should look like.
The experience of the UK
The UK’s market investigation tool has many procedural safeguards. First, a market investigation can only be initiated by a vote of the board of the Competition and Markets Authority (CMA), the UK competition authority.[10] Half the CMA board consists of “non-executive directors”, often with backgrounds and experience from outside government.[11]
If initiated, the investigation is led by three randomly drawn members of the CMA Panel. Like the Board, the Panel consists of people with a wide variety of experience, including in business, economics, law, finance, academia, within the CMA, or in other regulatory agencies.[12] Panel members will often take on the role alongside other jobs outside the agency, although they must adhere to a code of conduct and declare conflicts of interest.[13] The Panel members decide on remedies (which are binding) at the end of the investigation. These remedies can range from structural remedies as far-reaching as divestment to behavioural ones like information-sharing, to simply making recommendations to other branches of government.[14]
The reliance on people with a breadth of experience reduces the risk of confirmation bias. Panel members are less likely than full-time enforcement officials to have gotten used to seeing the targets of investigations as adversaries. That is not to say that due process means leniency on corporations or that career enforcers are biased. But the Panel’s breadth of experience does make it more likely that the remedies will be designed with a broader range of perspectives in mind, much like the use of juries is supposed to in criminal trials.[15]
More important to due process are the rules on timelines and transparency, as well as the appeals process. Early in the investigation, the CMA draws up an anticipated timeline and sends it to main parties for comment. It updates the timeline as the investigation proceeds, and must release the final report within 18 months of initiating the investigation.[16] After the final report’s release, affected parties may appeal in a specialised court called the Competition Appeal Tribunal (CAT).[17] The appeals process may continue up to the Court of Appeals and then to the Supreme Court.[18] CAT appeals resemble any other adversarial court case: parties can argue their case and present evidence—including witness testimony under oath and witness cross-examination.[19]
These measures reduce the duration of uncertainty for businesses. Appeals in the CAT are relatively quick. Take the Airports investigation, which followed an investigation into the ownership of British airports. In that case, the Competition Commission (CC—the CMA’s predecessor) published its final report of the market investigation in March 2009; the CAT ruled by February 2010; and the Court of Appeals (to which the government appealed the CAT’s judgement) did so by October 2010.[20] Similarly in the Groceries investigation—in which the CMA investigated local competition between grocery stores and their relationships with their suppliers—it took just 11 months from the CC’s final report for the CAT to rule against the CC, and another seven months for the CC to redo its analysis based on the court’s direction and issue a new final report.[21]
The UK’s market investigation tool has achieved significant results. Its greatest moment came in the Airports investigation. After a two-year market investigation into the ownership of British airports ending in March 2009, the CC ordered the British Airports Authority (BAA, a private company) to sell off three large international airports.[22] BAA appealed multiple times, with their final appeal exhausted in 2012.[23] The case showed that the market investigation power could claim large victories despite the cumbersomeness of the due process protections. Indeed, the rules might have enabled the CC/CMA to pursue big targets by providing legitimacy to the process and giving affected parties a clear path for recourse. The Airports investigation has been widely praised.[24] As one CMA affiliate described it: “If you can do one Airports investigation every twenty years, you deserve to take the rest of the century off.”[25]
In other cases, the appeals process has sometimes gone against the CC/CMA. In the Groceries investigation, for instance, the CAT found on appeal that the CC had failed to take certain factors into account when recommending that local authorities include a “competition test” as part of the planning process for new grocery stores.[26] The CC reviewed its analysis based on the CAT decision, and reissued its competition test recommendation with those factors taken into account.[27]
Mexico’s experience
Mexico’s experience shows what can happen when a market investigation tool operates under legal uncertainty. COFECE, the country’s competition authority, has used its market investigation tool seven times.[28] The most prominent cases involve recommendations for regulatory changes, and none so far appear to have directed structural remedies against private actors.[29] COFECE’s recommendations have often been overruled by other branches of government, however, leading to protracted debate and litigation about the scope of COFECE’s market investigation tool and the extent to which the recommendations are binding.
COFECE investigated the freight transport market in the state of Sinaloa in 2017, recommending regulatory changes to the Sinaloa state congress, which adopted the proposals.[30] But the state congress later modified the law again to include provisions contrary to COFECE’s recommendations.[31] Subsequent court decisions upheld those laws, stating that COFECE’s recommendations were not binding.[32] Similarly, COFECE’s recommended regulatory changes following a 2018 investigation into the milk market in the state of Chihuahua were ignored.[33] And after COFECE investigated landing slots at Mexico City airport in 2017, the Ministry of Transport and Communications issued regulations contradicting COFECE’s recommendations, which were again upheld later by the judiciary.[34]
Those afraid of unconstrained regulatory power might see these setbacks as useful restraints on COFECE. But this kind of impediment enhances neither due process nor the tool’s quality. Protracted litigation over the scope and legality of a market investigation tool prolongs uncertainty both for the regulator and affected businesses. An affected business would have little incentive to change its behaviour, in case the recommendations were later deemed invalid, but it would also struggle to attract investment while waiting for the outcome of litigation. Moreover, whether other officials choose to ignore COFECE’s recommendations is more likely to correlate with institutional inertia or the power of vested interests than with the quality or fairness of the recommendations. Ultimately, the tool’s effectiveness is blunted. As COFECE officials put it, “we learned that […] it might not be reasonable to devote too many enforcement resources to this type of case.”[35]
Evaluating the EU’s NCT proposal
Based on the structure of DG COMP, the EU’s competition authority, it would likely struggle to provide the procedural safeguards offered in the UK. DG COMP has no history of using the kinds of internal checks that the UK’s CMA has, like internal separation of decision-making or panels with outside experience. Some believe that DG COMP also offers less transparency and clarity on its decision-making during investigations.[36] Ian Forrester and Nicholas Forwood, both former judges of the Courts of Justice of the European Union (CJEU), argue that the European Commission’s “structures are unique in the world of competition law enforcement” for its lack of separation of roles within the agency.[37] They add:
“The same case team is charged with several functions of which some might seem to be inconsistent. Thus officials may receive complaints, investigate the facts, reach a preliminary view that there has been an infringement, make an accusation, receive the defence, consider whether it is persuasive, organise the hearing, consider afresh the defence and then frame the condemnation and help decide the penalty. The decision is formally taken by a political and not a judicial or quasi judicial body.”[38]
This institutional design is especially problematic for the New Competition Tool, since the penalties wouldn’t be based on wrongdoing (like anticompetitive conduct cases) nor invited by the affected parties (like a merger review). With the NCT, the decision rests on a judgement of the extent to which a party should be penalised for the greater good of the market. Such inquiries, to which there can be many answers, lend themselves to a more consultative, deliberative and transparent decision-making process.[39]
The EU’s bigger problem is judicial review. Some parts of the process are relatively fast but less rigorous, and others are more rigorous but far too slow. DG COMP holds administrative hearings before deciding on penalties. These are ‘trials’ of sorts—except they lack important components like the ability to call or cross-examine witnesses.[40] The nature and timing (before the decision) of the hearings mean they do not qualify as judicial review.[41]
Parties can appeal to the CJEU after DG COMP’s decision, but this appeal is a very slow process. For example, after DG COMP imposed a fine on Intel in 2009, the CJEU’s General Court ruled on Intel’s appeal only in 2014, and the European Court of Justice ruled in 2017.[42] But while a fine can be paid back, many of the NCT’s potential remedies would be harder to undo. If remedies were stayed till the end of the appeals process, the NCT would be glacially slow—which would be especially unworkable in fast-moving digital markets. If they were implemented immediately, they would likely be hard to reverse, thus eliminating appeal to the CJEU as a meaningful form of recourse.
Like Mexico’s tool, the NCT would operate under a cloud of legal uncertainty. The European Commission’s powers require a legal basis in the Treaty on the Functioning of the European Union. The tool could not rest solely on Article 103 (which covers competition), because it gives the Commission powers beyond the competition enforcement powers currently specified in the TFEU.[43] Officials had tentatively suggested a dual basis of Articles 103 and 114 (which covers distortions of the single market).[44] But that seemed shaky under the “centre of gravity” test the Courts of Justice of the European Union have used to evaluate dual legal bases. The law’s centre of gravity cannot rest heavily on one basis with the other being “merely incidental.”[45] And in this case, it is uncertain whether the use of Article 114 would pass that test. The question would likely be the subject of prolonged litigation, with DG COMP unlikely to make much use of the NCT until the CJEU had decided on the legality of its existence.
Like the Mexican state congresses, the EU’s member states may push back on penalties against some of their larger firms. As shown by the reaction to the blocked Siemens-Alstom merger—in which DG COMP, to the dismay of France and Germany, blocked a merger aimed at creating a European corporate giant—some member states already disagree with DG COMP.[46] They may perceive the NCT as an unacceptable deviation from established competition doctrines. Any lack of procedural due process within the NCT would compound this risk. If firms feel unable to seek adequate redress through the NCT’s formal processes, they would be more likely to try and lobby their host governments into resisting the NCT.
Conclusion
As the UK shows, market investigation tools can potentially do good by filling some of the gaps in existing competition law. In particular, by focussing on the structure of markets, they can revive competition in rigidly monopolistic markets and address anticompetitive harms that evade easy doctrinal definition. But adequate procedural due process and a certain legal basis are crucial in balancing—and indeed enabling—the tool’s substantive power. Mexico’s experience shows how uncertainty can open the door to informal political pressure on the competition authority and hamstring the tool. Considering that the EU was ill-equipped to provide the safeguards the UK offers and the NCT’s vulnerability to legal challenge and political pressure, the decision to abandon the NCT was justified.
The EU may later decide it does want a market investigation tool, as it grapples with the many new challenges facing competition law. But any such proposal must reckon with the need to create institutions that the EU does not yet have, including independent decision-making bodies within its bureaucracy and either a new set of specialised courts or a radical overhaul of its administrative hearings, and must find a firmer legal basis than the one suggested so far. Sidestepping these issues would only result in a tool that is both less just and less effective.
[1] European Commission, “The Digital Services Act Package,” available at https://ec.europa.eu/digital-single-market/en/digital-services-act-package.
[2] Id.
[3] European Commission, “Single Market – new complementary tool to strengthen competition enforcement,”; available at https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12416-New-competition-tool.
[4] Patrick Foulis, “Across the West powerful firms are becoming even more powerful,” The Economist, 15 November 2018, available at https://www.economist.com/special-report/2018/11/15/across-the-west-powerful-firms-are-becoming-even-more-powerful.
[5] Tim Wu. (2018). The Curse of Bigness. La Vergne: Columbia Global Reports, at 133.
[6] Nicholas Hirst and Lewis Crofts, “New EU antitrust power to tackle ‘structural’ problems eyes algorithms, tacit collusion,” MLex, 1 June 2020; available at https://mlexmarketinsight.com/insights-center/editors-picks/area-of-expertise/antitrust/new-eu-antitrust-power-to-tackle-structural-problems-eyes-algorithms-tacit-collusion.
[6] American Bar Association, “Comments to European Commission on New Competition Tool,” 8 September 2020, at 1–2; available at https://www.americanbar.org/content/dam/aba/administrative/antitrust_law/antitrust-comments-eu-new-competition-tool.pdf.
[7] For an example of this, see Iceland’s fossil fuels market investigation, 15 December 2015, at 3; available at https://en.samkeppni.is/media/skyrslur-2015/Summary_enska.pdf
[8] Margrethe Vestager, “Competition in a Digital Age,” ASCOLA Annual Conference, 26 June 2020; https://ec.europa.eu/commission/commissioners/2019-2024/vestager/announcements/competition-digital-age-changing-enforcement-changing-times_en.
[9] Iceland’s market investigation tool has some similar safeguards and results, but I lack space to address them here.
[10] CMA, “Market Studies and Market Investigations: Supplemental guidance on the CMA’s approach,” at 7, 2017; available at https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/624706/cma3-markets-supplemental-guidance-updated-june-2017.pdf.
[11] CMA, “Our Governance”; available at https://www.gov.uk/government/organisations/competition-and-markets-authority/about/our-governance.
[12] CMA, Supplemental Guidance, supra note 10, at 7 and 47.
[13] CMA, “Our Governance,” supra note 11.
[14] Amelia Fletcher, Market Investigations for Digital Platforms: Panacea or Complement?, Centre for Competition Policy University of East Anglia, at 8; available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3668289.
[15] Id. at 12.
[16] CMA, Supplemental Guidance, supra note 10, at 26–27.
[17] Id. at 35.
[18] Id.
[19] Competition Appeal Tribunal (CAT), Guide to Proceedings (2015), at 107; available at https://www.catribunal.org.uk/sites/default/files/2017-12/guide_to_proceedings_2015.pdf.
[20] Competition Commission v. BAA Limited [2010] EWCA Civ 1097, C3/2010/0589 [1]–[4], available at
https://www.catribunal.org.uk/sites/default/files/1097_BAA_CofAJudg_131010.pdf . BAA filed additional appeals later, but on the basis of changed circumstances.
[21] Competition Commission (CC), “CC Renews Competition Test Recommendation” (press release), 2 October 2009; available at https://assets.publishing.service.gov.uk/media/55194b9ded915d14240002fe/46-09.pdf.
[22] CC, BAA airports market investigation (final report), 19 March 2009, at 14–16; available at https://www.gov.uk/cma-cases/baa-airports-market-investigation-cc.
[23] BBC News, “Stansted Airport to be sold by BAA,” 20 August 2012 ; available at https://www.bbc.com/news/business-19318094.
[24] See Wu, supra note 5, at 133.
[25] Interview with a CMA affiliate.
[26] CC, The supply of groceries in the UK market investigation (final report), 30 April 2008, at 10; available at https://webarchive.nationalarchives.gov.uk/20140402235418/http://www.competition-commission.org.uk/assets/competitioncommission/docs/pdf/non-inquiry/rep_pub/reports/2008/fulltext/538.pdf.
[27] Competition Commission (CC), “CC Renews Competition Test Recommendation” (press release), 2 October 2009; available at https://assets.publishing.service.gov.uk/media/55194b9ded915d14240002fe/46-09.pdf.
[28] OECD, OECD Peer Reviews of Competition Law and Policy: Mexico (2020), at 67; available at http://www.oecd.org/daf/competition/Mexico-Peer-Reviews-of-Competition-Law-and-Policy-en.pdf.
[29] COFECE, “Using Market Studies to Tackle Emerging Competition Issues – Contribution from Mexico (21 November 2020), at 7. Available at https://www.cofece.mx/wp-content/uploads/2020/12/Using-Market-Studies-Mexico.pdf.
[30] OECD, Mexico peer review, supra note 28, at 69.
[31] Id.
[32] Id.
[33] Id.
[34] OECD, Mexico peer review, supra note 28, at 69 and OECD, Contribution from Mexico, supra note 30, at 7.
[35] Id.
[36] International Chamber of Commerce, Due process in EU antitrust proceedings, 8 March 2010, at 3–4; available at https://www.icc-portugal.com/images/documentos/comissao_concorrencia/2010_Due-process-in-EU-antitrust-proceedings.pdf.
[37] Ian Forrester and Nicholas Forwood, “Due process in European law: a post-judicial view,” Concurrences, 3 September 2020, at 2; available at https://www.concurrences.com/en/bulletin/special-issues/due-process-research-program-en/due-process-in-european-law-a-post-judicial-view-en.
[38] Id.
[39] See also Forrester and Forwood (who make a version of this argument), supra note 37, at 3.
[40] ICC, supra note 36, at 3–4.
[41] Id.
[42] Court of Justice of the European Union (CJEU), Press Release No 90/17, 6 September 2017; available at https://curia.europa.eu/jcms/upload/docs/application/pdf/2017-09/cp170090en.pdf.
[43] Alfonso Lamadrid and Pablo Ibanez Colomo, “Can This Be the New Normal? 10 Questions on the Proposed New Competition Tool,” Chillin’Competition, 11 June 2020; available at https://chillingcompetition.com/2020/06/11/can-this-be-the-new-normal-10-questions-on-the-proposed-new-competition-tool/.
[44] European Commission, “Annexes to the Adjusted Commission Work Programme 2020,” 27 May 2020, at 2 (¶ 9); available at https://ec.europa.eu/info/sites/info/files/cwp-2020-adjusted-annexes_en.pdf.
[45] C-491/01, The Queen v. Secretary of State for Health ex parte: British American Tobacco Ltd. (ECJ 2001), at ¶ 168; available at http://curia.europa.eu/juris/document/document.jsf?text=&docid=47635&pageIndex=0&doclang=EN&mode=lst&dir=&occ=first&part=1&cid=4514638#Footref77.
[46] Elisa Braun, Thibault Larger, and Simon Van Dorpe, “EU Big Four Press Vestager to Clear Path for Champions,” Politico, 6 February 2020; available at https://www.politico.eu/article/eu-big-four-france-germany-italy-poland-press-executive-vice-president-margrethe-vestager-to-clear-path-for-champions/