Giorgio Monti (Tilburg Law and Economics Center)1
1. Introduction
The COVID-19 pandemic has seen the Commission respond readily to state aid measures: protocols established in 2008 were replicated. However, during the financial crisis and the ensuing recession the Commission was not at all sympathetic to private restraints to competition for troubled firms. The then-Commissioner for competition policy, Neelie Kroes, spoke of the need for the Commission to show “tough love” when applying competition law during the crisis.2 COVID-19, however, has brought a softer antitrust response.3 On April 8, 2020 the Commission issued a communication establishing a Temporary Framework for assessing antitrust issues related to business cooperation in response to situations of urgency stemming from the current COVID-19 outbreak (the “Framework”).4 As with other competition agencies, the Commission signals a willingness to tolerate agreements which might harm competition but which may be necessary to deal with the disruption caused by the pandemic.5
The Framework notes that firms may face one of two problems: on the one hand a sharp fall in demand as shops close, businesses buy fewer of their requirements and supply chains are disrupted. On the other, a sharp rise in demand affecting firms supplying medical equipment and other essential goods and placing pressure on them to augment production. The Commission’s Framework addresses the latter market failure, noting that suppliers of essential and scarce products may need to cooperate “in order to overcome or at least to mitigate the effects of the crisis to the ultimate benefit of citizens.”6 The Framework contains two elements: a substantive discussion of what modalities of cooperation may be tolerated and a procedural pathway to communicate such agreements to the Commission. In both respects, the approach chosen may be compared by reference to the legislative framework adopted in the UK and with the approach of the U.S. Federal antitrust agencies. We start with a critical reflection on the Commission framework, followed by a comparative account, and conclude with a discussion of how the Framework might evolve.
2. Tolerating More Cooperation Than Usual
The Framework applies solely to undertakings operating in certain industries (e.g. medicines and medical equipment), as well as other scarce products and services, and to forms of cooperation that ensure the supply and distribution of such goods during this emergency.7 It identifies three scenarios for assessment under Article 101 TFEU.
The first is that firms engage in cooperation that would be largely unproblematic even in normal times (e.g. a trade association that coordinates joint transport, or works on a model to identify supply gaps). Provided no sensitive information is exchanged among its members, these arrangements are unlikely to infringe competition law.8
Second, firms might need to cooperate more intensely, for instance in coordinating production to ensure that demand is met. In these scenarios the Commission is willing to accept such cooperation provided the following criteria are met: (i) the cooperation is objectively necessary to increase output in the most efficient way to ensure supply of essential products; (ii) the measures are temporary; (iii) they are proportionate, such that they do not go beyond what is necessary to address the shortage of supply. Moreover, undertakings should keep records of all agreements so that these are available for the Commission.9
A number of doubts emerge. First, the legal basis for tolerating these forms of cooperation is unclear. The Framework states: “in the current exceptional circumstances, such measures would not be problematic under EU competition law or – in view of the emergency situation and temporary nature – they would not give rise to an enforcement priority for the Commission.”10 Which is it to be? Are these agreements procompetitive such that the positive effects outweigh the negative ones and the agreements are thus not a restriction of competition?11 Is such cooperation seen as in the public interest and excluded?12 Are these ancillary restraints? Is this cooperation likely to benefit from Article 101(3) TFEU? Or is it an infringement of competition law which the Commission does not treat as an enforcement priority? Or is it a combination: the likely procompetitive effects of these agreements are sizeable thus an effects based analysis will be required and this will be costly and therefore unlikely to be in the EU interest to carry out such an investigation as resources are better spent against firms who use the crisis to collude? Granted, the pathway to legality may not matter too much in the current state of emergency, but as with other forms of cooperation which may be in the public interest, we see some reluctance to engage with the legal order and explain the basis upon which agreements may be exempted.13
Having said that, the criteria are largely sensible and seem to reflect some of the requirements in Article 101(3) TFEU,14 although it may be tricky even for the undertakings themselves to work out what constitutes a “temporary” arrangement, not least given the present degree of uncertainty. Tolerating, say, a year-long cooperation agreement might be preferable: firms might then be able to plan production better if given a clear timescale.15 The requirement to retain records appears reasonable, although well-advised undertakings would do so anyway in case an action is brought against such agreements.
Third, the Commission notes that some forms of coordination may be encouraged or required by Member States. In case of cooperation resulting from an “imperative request” from public authorities then EU competition law is not applicable to agreements between undertakings because these are required by the state.16 However, the state remains at risk for requiring collusion under Article 4(3) TEU read jointly with Article 101 TFEU or under internal market law (e.g. if the national rules are a measure equivalent to a quantitative restriction). The Framework says nothing about how such rules might be applied (largely because if any such actions are brought they will more likely be before national courts) but it is likely that there will be pressure to interpret these in such a way as to excuse state action that harms the internal market in order to pursue the public interest.
3. The Return of Comfort Letters
Ever since Regulation 1/2003, the Commission has refused to provide formal guidance for individual agreements: firms should rely on soft law, precedents, and carry out their own risk assessment. Quiet consultations with the Commission have occurred and are set to continue during the pandemic, but now DG Competition “stands ready, exceptionally and at its own discretion, to provide such guidance by means of an ad hoc “comfort” letter.”17
This is intriguing. First, the Commission has had a procedure to issue informal guidance letters since 2004.18 To my knowledge no such letters have been issued publically.19 Are ad hoc letters under the Framework any different from guidance letters? The Notice regarding guidance letters suggests that these would be sent when new issues arise.20 Is it that under the Framework, the Commission may be willing to write a letter even for agreements where the parameters of legality are tolerably clear?21 Perhaps the procedure is quicker under the Framework, and perhaps (unlike guidance letters) comfort letters under the Framework will not be made publicly available.22 At any rate, the legal effect of comfort letters will not be any different from letters issued under the Framework.23 In sum it is not clear what distinguishes ad hoc comfort letters under the framework and existing guidance letters.
Second, Article 10 of Regulation 1/2003 provides that the Commission may, in the public interest, issue non-infringement decisions. This has never been applied. Given the “exceptional challenges” that the Commission recognizes in this Framework, might this not be an occasion to use Article 10? What greater public interest are we waiting for? Comfort letters hark back to a time when the Commission, overburdened with notifications, found this procedural device to clear its decks quickly. However, at the time, this was not always welcomed because comfort letters are not binding.24 Back then, the Commission had tried to resolve this weakness by instituting a procedure where before issuing a comfort letter the notification would be published, allowing third parties to voice their views. The ensuing comfort letter would then take such views into consideration. This did not make the letter any more binding but the hope was that the procedure would reduce the risk of third parties challenging the practices in question.25 158 such letters were issued in 1990 (similar numbers may be found in other years), indicating that some stakeholders found this an agreeable solution.26 However, the present emergency situation makes it unlikely that a procedure affording third parties the chance to comment (which we now see in commitment decisions) can be deployed. Why not then take a chance with formal decisions? One policy consideration might be that this would re-introduce the notification system. Article 10 is carefully drafted to explain that a non-infringement decision is issued at the Commission’s discretion with it “acting in its own initiative.”27 The Commission might be concerned that use of this provision when responding to a notification is risky in the long term, setting a precedent for how Article 10 decisions are triggered. However, the Regulation also explains that the purpose of Article 10 decisions is for the Commission to offer clarification regarding “new types of agreements or practices.”28 From this perspective, non-infringement decisions in the present context seem appropriate when firms collaborate to meet shortages.
On the day the Framework was released, the first comfort letter was issued to Medicines for Europe, an association whose members supply some 70 percent of medicines in Europe. The agreement is designed to facilitate cooperation in increasing the supply and improving the distribution of medicines. 29 It reveals that there was some steer from the Commission and the DG for Health and Food Safety to secure cooperation in the first place, and the modalities of cooperation are also monitored by these EU bodies, with the Commission assuming a steering role. The agreement contains safeguards on limiting information exchanges and commitments not to engage in price gouging. Even from this brief account we can see that the criteria for toleration set out in the Framework may be necessary but not sufficient to merit the issuance of a comfort letter. Much like with the old exemption decisions under Article 101(3) TFEU one sees authorization combined with the addition of safeguards and conditions.30 It would thus be desirable for the Commission, as it gathers experience, to offer more specific guidance.
However, one wonders again whether a set of non-infringement decisions might not be preferable to more soft law: not only would the recipients benefit from enhanced legal security, but we would also see an official legal text explaining the basis for a finding of non-infringement which could serve as a template for other collaborations.
4. A Comparative Account
The British Government has taken a different approach: it has issued three Orders in the form of Statutory Instruments to exclude the application of the Competition Act 1998 to certain industries for specific activities.31 A further Order is expected for the dairy industry, and more may follow. The legal basis for these is found in the national competition legislation: if the Secretary of State finds that there are “exceptional and compelling reasons of public policy” it may order that restrictive practices are excluded from the Chapter 1 and/or Chapter 2 prohibitions of the Competition Act 1988 (the domestic equivalent of Articles 101 and 102 TFEU).32 This legislative option is not open to the EU legislator (although it could opt for issuing a Block Exemption Regulation which has similar effects) but is it a smart strategy to begin with?
On the plus side, these agreements confer much more legal certainty than the Framework: undertakings who cooperate are immune from competition law challenge under UK Law.33 On the negative side, however, the Orders are narrowly cast: they apply to a defined set of undertakings and to discrete agreements. For instance, one of the Orders applies to five maritime operators that offer transport services between the Isle of Wight and the UK mainland and it allows them to coordinate timetables, routes and labor facilities to ensure continuity of services. Why select just this one route when similar problems might affect any transport route? The other two Orders are slightly more widely cast. For example, the Order about grocery supplies is applicable to all grocery-chain suppliers and logistics service providers and excludes a wider range of collaborative agreements. Moreover some flexibility is built into it: supposing a removals firm decides that rather than leave its trucks idle it will enter the market for food delivery, it seems like it will qualify as a logistics service provider under the Groceries Order because it encompasses “persons seeking to enter” the relevant markets, and not just established actors.34 However, logistics service providers are only allowed to exchange information: what if sharing facilities or more intense forms of cooperation become necessary to mitigate disruption? Will Orders need to be continually updated? Will new Orders arrive regularly as other industry sectors lobby for support, with all the risks that are associated with this?
The need for flexibility is recognized by the Competition and Markets Authority (“CMA”) in explaining its approach to business cooperation, indicating that as the pandemic evolves, “the issues faced by businesses as they participate in efforts to mitigate the effects of the pandemic” will also change requiring ongoing monitoring and responsive guidance.35 Thus, the CMA will plug what gaps remain in the Orders issued by the Government. Coordination between Government and the CMA does not seem to have been optimal: the authority’s approach is to guide business conduct to facilitate beneficial cooperation and avoid firms taking advantage of the crisis to harm consumers, while the Government is more interested in bright line exclusions. Orders may offer greater legal security but the risk of under-inclusion through lack of foresight (understandable) or over-inclusion through skillful lobbying (less excusable) weigh in favor of leaving these issues to be solved by an independent competition authority.
Similarly, it does not appear wise for the Commission to use this approach as a guide to design a Block Exemption Regulation. Normally these emerge after some experience is gathered so that they are well-calibrated and anyhow modern block exemptions only apply to agreements where there is no significant market power, while some of the sectors concerned here are highly concentrated. Moreover, identifying the precise scope of such exemptions seems tricky in circumstances of uncertainty. Regular review of the Framework seems preferable to a legislative route.36
Procedurally, the UK Orders last for as long as the Government considers “that there is no longer a significant disruption or a threat of significant disruption” in the relevant markets, much like the Commission Framework. One wonders whether a fixed duration might have afforded more opportunities for smart planning by the beneficiaries. Agreements may only benefit from the Order if they are notified and these notifications are to be recorded in a publicly available register. But what is the added value of notification? It might help the Government to understand what sort of agreements firms find useful and thus shape further Orders. But it does not seem that such notifications can have much of a deterrent effect.37
In the United States, the two Federal agencies redeployed a playbook already developed to deal with natural disasters.38 They issued a Joint Antitrust Statement Regarding COVID-19.39 Like the Commission, the agencies recall that many agreements may be assessed with the existing toolbox but provide for some additional flexibility given the pandemic. Of interest for this comparative overview is that the agencies already operate an ex ante notification system. The joint statement accelerates this procedure: for agreements linked to the pandemic the agencies promise to deliver a business review letter (DOJ) or an advisory opinion (FTC) to notified agreements in seven days. Parties have a relatively light notification requirement, and the information requested is not dissimilar to what the Commission will seek when considering a request for a comfort letter. The Commission requires information about the firms and products concerned, an outline of the cooperation, its benefits and competition risks.40 The agencies instead ask for a description of the nature and rationale of the proposal, the participants, the scope of the agreement, copies of the contract and the names of major customers and potential competitors.41 The difference on paper is that the U.S. does not ask for a preliminary self-assessment, but in practice this may not be such a significant difference: parties who notify will likely make a case for approval.
U.S. agencies publish the resulting comfort letters. It is not clear why the U.S. agencies are willing to release fuller documents than the Commission. The first business review letter issued by the DOJ under this joint statement is an 11-page document explaining how U.S. healthcare distributors propose to cooperate to expedite the manufacture and distribution of personal protective equipment: it explains how Government encouraged such cooperation, what aspects of cooperation fall outside of antitrust laws, and how the parts of the agreement that fall to be scrutinized under Section 1 of the Sherman Act are likely to benefit society, as well as explaining the safeguards put in place to prevent benevolent cooperation becoming malignant collusion.42 This is the kind of legal detail and transparency other agencies should aspire to.
5. Beyond the Framework
The Commission Framework is cast narrowly; it will be interesting to see if it is revised to also consider markets where firms wish for lenient treatment in cases where there is a fall in demand and coordinated closures are planned. These are trickier to handle because they carry a greater risk of anticompetitive effects.43 The UK Government, for instance, has promised an exclusion Order for dairy farmers to allow them to coordinate reductions in supply.44 It will also be worth observing whether, as the Commission sees more cases, the Framework becomes more concrete not only in explaining what forms of cooperation are desirable but also what safeguards should be put in place to minimize anticompetitive effects.
As we have seen, the Framework does not make it clear what the basis for tolerating agreements is, hedging between a suggestion that there is no infringement or that even if it is an infringement, the agreement is not an enforcement priority. One additional option that has yet to be considered so far is the use of Article 106(2) TFEU. In brief, this applies to undertakings which a Member State entrusts with the provision of services of general economic interest. It provides that the Treaty rules (and in particular the competition rules) do not apply when they would “obstruct the performance, in law or in fact, of the particular tasks assigned to them.” This provision has received much political interest because it serves to shield the rules of the market from certain services where states may legitimately take the view that other considerations should prevail. Successive Treaty revisions have introduced provisions to strengthen the role of services of general interest, but these have largely been cosmetic.45 At the same time, what little case law there is suggests that the Court is fairly accommodating to the choices Member States make, both when it comes to the designation of services and the regulatory framework. It remains to be seen if there are certain economic sectors (e.g. the manufacture of personal protective equipment, medicines, or even supermarkets) which might be designated under this provision on a temporary basis. It would afford operators the maximum degree of flexibility to cooperate in the public interest. An even bolder approach might be for the EU to draft secondary legislation to exclude certain sectors from the application of EU competition law as they furnish a service of general interest, not least since coordination among firms across Member States may be desirable. However, it may be tricky to find a legal basis for this.46 Moreover, as with exclusion Orders, this might be too blunt an instrument to consider.
Moreover, as noted above when it comes to medical equipment, EU and U.S. institutions have both played a role in encouraging coordination by undertakings. There may well be further instances where Member States support similar schemes. In such settings it may be useful to create a process whereby the agreement is regularly monitored not only to ensure that the parties deliver what they promise, but that they avoid collusion – such monitoring should form a condition for tolerating an otherwise potentially anticompetitive practice. Embedding competition law consideration in state policy is what advocacy is all about.
A final reflection is how far the experience with agreements reviewed during the pandemic will shape the Commission’s review of the guidelines on horizontal agreements and the Block Exemptions. There is some political pressure to facilitate forms of cooperation to enhance sustainability initiatives, for example.47 This topic is closely linked to the one at play here: how far should the EU facilitate cooperation among competitors if this may achieve certain vital public interest goals? It may be that the experience generated during the pandemic can help inform the Commission’s review. For this to occur, however, it is imperative that the Commission and NCAs operate more openly. Ideally, Article 10 decisions would be the preferred route, for they would allow the public to see what is exempted and why, they would avoid a perceived risk of favoritism, they would force the Commission to take a stance on the legal basis for exemption, and they would reveal what restrictions on conduct the Commission fixes in exchange for granting exemption. Short of that publishing comfort letters that give an inkling on the competition assessment may serve as a second-best. The values of accountability, equal treatment and good administration converge to press for a more open form of decision making than which is provided for under the current Framework.
Click here for a PDF version of the article
1 Professor of Competition Law, Tilburg Law and Economics Center. TILEC has received funding by Qualcomm Inc. This research was conducted in accordance with the Royal Dutch Academy of Sciences (“KNAW”) Declaration of Scientific Independence. I am grateful to Or Brook and Francisco Costa-Cabral for stimulating comments; any errors are mine.
2 N. Kroes, “Competition, the crisis and the road to recovery,” March 30, 2009 (SPEECH/09/152).
3 For an early assessment, see F. Costa-Cabral, L. Hancher, G. Monti & A. Ruiz Feases, “EU Competition Law and COVID-19,” TILEC Discussion Paper No. DP2020-007. Available at https://ssrn.com/abstract=3561438.
4 [2020] OJ CI 116/7.
5 See e.g. ECN Antitrust: Joint statement by the European Competition Network (“ECN”) on application of competition law during the Corona crisis. (undated), ICN Steering Group Statement: Competition during and after the COVID-19 Pandemic (April 2020).
6 Framework (supra note 4), paragraph 3.
7 Supra note 4, paragraph 4. But no exhaustive list of industries is provided.
8 Supra note 4, paragraphs 12 and 13.
9 Supra note 4, paragraph 15.
10 Supra note 4, paragraph 15.
11 This draws on Gazdasági Versenyhivatal v. Budapest Bank Nyrt. and Others, Case C-228/18, in particular AG Bobek’s Opinion at paragraph 78 and ECJ at paragraphs 74-75. Similarly, in Generics UK Ltd. And others v. CMA, Case C-307/18, the Court held that a finding of “significant pro-competitive effects” means that the agreement may not be restrictive of competition by object (paragraph 107) and its overall impact on competition must be assessed. We leave for another day a discussion of what “pro-competitive” might mean, but it may follow that an agreement under the Framework which increases the supply of essential goods is on balance likely to stimulate competition.
12 E.g. following one interpretation of Wouters, C-309/99, EU:C:2002:98.
13 It should be noted that the precise role of public interest considerations in competition law assessment is contested. For a selection of the debates, see G. Monti and J. Mulder “Escaping the Clutches of EU Competition Law: Pathways to Assess Private Sustainability Initiatives” (2017) 42 ELRev 635; F. Costa-Cabral, “Future-Mapping the Three Dimensions of EU Competition Law: Modernisation Now and After COVID-19,” TILEC Discussion Paper No. DP 2020-011 pp.17-23 Available at https://ssrn.com/abstract=3581535; O. Brook, “Priority-Setting as a Double-Edged Sword: How Modernisation Strengthened the Role of public policy,” Journal of Competition Law and Economics (forthcoming). Available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3580837.
14 Of the four conditions in Article 101(3) TFEU, two are present: efficiencies and proportionality, two are missing: consumer fair share and the requirement that the agreement does not eliminate competition (although these seem to be implied).
15 The U.S. agencies (see below n 37) indicate that similar agreements would receive a business review letter valid for one year.
16 Supra note 4, paragraph 16.
17 Supra note 4, paragraph 18.
18 Commission Notice on informal guidance relating to novel questions concerning Articles 81 and 82 of the EC Treaty that arise in individual cases (guidance letters) [2004] C 101/78.
19 According to the Commission this is because “only a few approaches have been made to the Commission and none of them fulfilled the conditions for a request for a guidance letter set out in the Notice.” Ten Years of Antitrust Enforcement under Regulation 1/2003 SWD(2014) 230/2 p.58, Note 274.
20 Supra note 18, paragraphs 5 and 7.
21 Supra note 18, paragraph 8 provides a set of fairly restrictive conditions before the Commission will consider writing a letter.
22 Supra note 18 paragraph 21 clarifying the letters would be public.
23 Supra note 18 paragraphs 22-25 explaining these are not binding and the Commission may take up a complaint if new facts emerge. There is no legal basis under Regulation 1/2003 to issue such letters.
24 SA Lancôme and Cosparfrance Nederland BV v. Etos BV and Albert Heyn Supermart BV, Case 99/79, EU:C:1980:193.
25 Notice from the Commission on procedures concerning notifications pursuant to Article 4 of Council Regulation No 17/62 [1983] OJ C 295/6.
26 Twentieth Report on Competition Policy (1990) p.111.
27 Regulation 1/2003 on the implementation of the rules on competition laid down in Articles 81 and 82 of the Treaty [2003] OJ L1/1, Article 10.
28 Regulation 1/2003, Recital 14.
29 https://ec.europa.eu/competition/antitrust/medicines_for_europe_comfort_letter.pdf. It was released on the website some weeks after the date on the letter.
30 G. Monti, EC Competition Law (Cambridge University Press, 2007) pp.115-116 where I called this a “conditional legality” approach to exemption. We now see this replicated in commitment decisions.
31 The Competition Act 1998 (Health Services for Patients in England) (Coronavirus) (Public Policy Exclusion) Order 2020 (SI 2020/368), The Competition Act 1998 (Groceries) (Coronavirus) (Public Policy Exclusion) Order 2020 (SI 2020/369), The Competition Act 1998 (Solent Maritime Crossings) (Coronavirus) (Public Policy Exclusion) Order 2020 (SI 2020/370).
32 Competition Act 1988, Schedule 3, paragraph 7.
33 But there is no exclusion for abuse of dominance and EU competition law may applies until 31 December 2020.
34 Groceries Order, supra note 31, Article 2.
35 Guidance, “CMA approach to business cooperation in response to COVID-19,” (March 25, 2020). Like the Commission the CMA indicates a bifurcated approach: indicating that cooperation necessary to face the pandemic are not enforcement priorities because unlikely to infringe competition law and suggesting that some agreements may benefit from exemption under section 9 Competition Act 1998 (the equivalent to Article 101(3) TFEU).
36 Supra note 4, paragraph 21 indicates the Commission may review the Framework.
37 Suppose undertakings notify an excluded agreement and then collude to fix prices. In these cases, the CMA will be entitled to proceed for price fixing but the Order does not have any impact on the investigation or the penalty to be imposed.
38 DOJ Business Reviews and FTC Staff Advisory Opinion letters related to hurricane Katrina and hurricane Rita aftermath https://www.justice.gov/atr/doj-business-reviews-and-ftc-staffadvisory-opinion-letters-related-hurricane-katrina-and.
39 Issued in March 2020 (available at: https://www.justice.gov/atr/joint-antitrust-statement-regarding-covid-19).
40 These are not found in the Framework but on the Commission website, at https://ec.europa.eu/competition/antitrust/coronavirus.html.
41 Supra note 38.
42 US Department of Justice, McKesson Corporation, Owens & Minor, Inc., Cardinal Health, Inc., Medline Industries, Inc., and Henry Schein, Inc. Business Review Request Pursuant to COVID-19 Expedited Procedure (April 4, 2020).
43 See the discussion in Competition Authority v. Beef Industry Development Society Ltd, Case C-209/07, ECLI:EU:C:2008:643. In particular see this excellent analysis of the position taken by various actors in this saga: C. Talbot, “Finding a Baseline for Competition Law Enforcement During Crises: Case Study of the “Irish Beef” Proceedings,” (2015) 18(2) Irish Journal of European Law, Vol 18, Issue 2, 2015. Available at https://ssrn.com/abstract=2719506.
44 Dairy industry to join together to manage milk supply (April 17, 2020) Press Release at: https://www.gov.uk/government/news/dairy-industry-to-join-together-to-manage-milk-supply
45 W. Sauter, Public Services in EU Law (Cambridge University Press, 2014) ch.2.
46 Article 14 TFEU, introduced to recall the importance of services of general economic interest to the Union as a whole, provides for the possibility for the EU legislator to issue regulations. Nothing has ever resulted and it is not clear whether such regulation should be limited to coordinating national initiatives or if there is power to design EU-wide services of general economic interest.
47 G. Monti, “Four Options for a Greener Competition Law,” (2020) Journal of European Competition Law & Practice (forthcoming).
