1. Introduction
There has always been a practice of competition authorities to accept, on a discretionary basis, commitments by undertakings who promised to mend their ways in return for closing an investigation (vulgo: to “settle a case”). After all, the main goal of competition law enforcement should be to restore competitive conduct in the marketplace, and if this can be achieved at an early stage and without investing many of the competition authority’s resources, then all the better. For a long time, “settling cases” was handled informally by the European Commission and National Competition Authorities (NCAs), on the basis of their discretion to (de)prioritise cases for enforcement. [1] This informality under the old regime had the disadvantage that it was not entirely clear what the consequences would be if undertakings deviated from their commitments, if they had submitted incomplete, incorrect or misleading information, or if there was a material change of circumstances.
Part 2 will outline the formalisation of the commitment procedure on the EU level and the Member State level. Part 3 will describe the EU’s experience with the commitment procedure. Part 4 connects the EU experience (very perfunctorily) to the national experience. Part 5 provides a brief outlook to the future of the commitments and interim measures as substitutes and complements.
2. The formalisation of the commitments procedure
2.1. EU level: Art. 9 Regulation 1/2003
Famously, the commitment procedure was formalised on the EU level with the introduction of Regulation 1/2003 and its Art. 9, which provides that:
“[w]here the Commission intends to adopt a decision requiring that an infringement be brought to an end and the undertakings concerned offer commitments to meet the concerns expressed to them by the Commission in its preliminary assessment, the Commission may by decision make those commitments binding on the undertakings”,
that such a decision “shall conclude that there are no longer grounds for action”, and that it “may be adopted for a specified period”. Art. 9, in combination with Art. 27(4), of Regulation 1/2003 established and made transparent not only the procedure for accepting formal commitments, but also clarified the consequences of a breach of these commitments, or of a finding that the basis of the commitment decision was or becomes unsound. First, the Commission can reopen the proceedings in all these cases (Art. 9(2) Regulation 1/2003), and secondly, where an undertaking intentionally or negligently does not comply with its commitments, the Commission may impose a fine of up to 10 per cent of the global annual turnover under Art. 23(2)(c) Regulation 1/2003 — without any finding that the infringement of the commitments constituted a substantive antitrust infringement.
The Commission quickly embraced the commitment procedure – as has been discussed numerous times, perhaps even a little too enthusiastically.
2.2. Member State level: from Art. 5 Regulation 1/2003 to Directive (EU) 2019/1
Art. 5 of Regulation 1/2003 provides that NCAs also have the power to accept commitments when applying Articles 101 or 102 of the Treaty on the Functioning of the European Union (TFEU). Most, but – as the ECN Recommendation on Commitment Procedures noted [2] – not all, Member States introduced a procedure equivalent to that of Art. 9 Regulation 1/2003 soon after that Regulation came into force. [3]
Even when Member States adopted legislation, there were in some cases substantial variations in the procedure and sanctions for non-compliance. Not all national regimes provide for a market-testing procedure along the lines of Art. 27(4) Regulation 1/2003. [4] Nor is the timeline the same in all Member States: in France, commitments have to be offered before the Statement of Objections is issued. [5] In other jurisdictions, commitments can only be offered after the Statement of Objections has been issued. [6] In some other Member States, commitments can be offered at any stage of the proceedings under the statutory scheme, but there is a strong preference to have the commitments offered before the Statement of Objections. [7] In Italy, the wording of the statute provides that companies may offer commitments “[w]ithin three months from notification of the launch of an investigation”; [8] however, the teeth of this seemingly strict deadline appear to have been pulled out by the Regional Administrative Court of Lazio [9] and more recently by the Supreme Administrative Court (Consiglio di Stato), [10] which interpreted the three-month time limit not as a peremptory deadline but instead as a non-mandatory guideline. With regard to the ability to fine undertakings for non-compliance, the Impact Assessment accompanying the Commission Proposal for the ECN Plus-Directive noted that “11 NCAs either cannot fine failure to comply with such a commitment decision at all or the fines are set at a level which is too low to compel compliance. This means that authorities’ powers lack teeth, e.g. companies easily enter into commitments which cannot be enforced and therefore no market change ensues.” [11]
Market participants and lawyers have deplored these divergences between the different national commitment procedures, in particular with regard to the diverging timelines and the resulting difficulties in coordinating a consensual resolution in cases where several competition authorities were dealing with the same case or identical case scenarios. [12] In the discussion about the harmonization of national competition procedures in the run-up to the ECN+ Directive, there were hopes that the divergences would be addressed and resolved in the harmonization process.
Directive (EU) 2019/1 has, unfortunately, not addressed these divergences. [13] Practically nothing in Art. 12 of the Directive (with recital 39), which deals with commitment decisions, requires the Member States to introduce new measures or to modify their existing practice. The Directive states explicitly that the decision is to be taken “after formally or informally seeking the views of market participants”, allowing those Member States that do not have a formal market-testing procedure to continue their flexible practice. Nor are the timelines harmonized. The only – if not unimportant – changes in the Directive are the obligations on the Member States to provide the NCAs with means to monitor compliance effectively (Art. 12(2) of the Directive), to reopen proceedings in the case of non-compliance or where the basis for the decision was or becomes unsound (Art. 12(3)), and to impose “effective, proportionate and dissuasive fines” on the basis of the undertaking’s turnover where they intentionally or negligently fail to comply with binding commitments (Art. 14(2)(f) of the Directive).
3. The commitment practice on the EU level
3.1. Commission
A lot has been written about the eagerness with which the Commission and undertakings embraced the commitment procedure, and the advantages and disadvantages that are associated with the high prevalence of commitment decisions. [14]
3.1.1. Reasons for and against the commitment procedure
On the plus side, the commitment procedure has been perceived as being quicker and less resource intensive than the infringement procedure, which brings direct benefits not only for the enforcer and the undertakings concerned, but also for the quick reinstatement of competition as an undistorted process on the market concerned. Also, there is more flexibility in the choice of remedies, which can be better tailored to addressing the effects of the restriction of competition. From the undertakings’ perspective, it also avoids the binding effect of a final infringement decision in private enforcement proceedings – even though, as discussed below (3.2.4.), it does not insulate them against private actions.
On the minus side, the fact that commitments are the result of a negotiation rather than the application of legal principles has not only advantages, but also drawbacks. While the adversarial position of the parties should ensure some balance, one negotiator may gain an advantage over the other by leveraging outside options — e.g., the Commission imposing disproportionate commitments by threatening to impose a large fine or regulation —, or by taking advantage of information asymmetries — e.g., an undertaking promising inadequate commitments to resolve a particular concern where the Commission has only uncovered a small part of a much larger, undetected restriction of competition. Similarly, the negotiation dynamics, fatigue, and public choice consideration (“getting that case closed before the summer break at all costs”) may lead to a partial alignment of interests between the negotiating parties with external effects on the public interest in a system of undistorted competition. The commitment procedure may also either neglect third parties’ interests—e.g., where these interests are diffuse, such as end consumers’ interests —, or conversely inflate the impact of these interests, where well-organised lobby groups take advantage of the market-test procedure, or where political actors seek to influence the outcome of the competition procedure. The commitment procedure can lead to “regulatory antitrust”, which may on occasion be a good thing, but may also prevent the otherwise permissible use of competitive parameters (i.e., be restrictive of competition), or bypass democratic controls of regulation. A particular problem of commitment decisions is that they do not provide clarity on the borderline between legal and illegal conduct — the undertakings do not admit an infringement of the law while the Commission raises concerns. Undertakings may decide to enter into (costly) commitments because the undertakings fear that they would lose before court; but alternatively, they may make this decision because the commitments avoid an unfounded infringement decision, and the costs of the commitments are lower than the nuisance value of appealing such a decision. The temptation to choose the commitment procedure is particularly great for both the Commission and the undertakings when the issues are novel and the outcome of a court decision uncertain. While the commitment procedure resolves the individual case, this means that the legal issue remains novel and unresolved – potentially leading to future proceedings which could have been avoided if the issue had been resolved once and for all.
3.1.2. A shift from the commitment procedure to infringement decisions?
While Commissioner Almunia was very much in favour of the commitment procedure, the experience with the Google Shopping case in particular has resulted in a more skeptical approach by the Commission since 2014. This change in tone is recognizable in particular in speeches, which begin to acknowledge that the commitment procedure is not always quicker and less resource-intensive than the infringement procedure, and that it may be preferable instead to use interim decisions when the persistence of a restriction of competition threatens to result in irreparable harm. The first indication for a change in attitude was the “double act” of the infringement decision without imposing a fine for a novel issue in Motorola and the commitment decision in Samsung, and the switch from the commitment procedure to a Statement of Objections and an infringement decision in Google Shopping. In terms of statistics, however, the shift is not discernible at all: Since 1 January 2014, commitment decisions were used in nine cases, and infringement decisions in seven non-cartel cases. This is exactly the same proportion as in the time period between 1 May 2004 and 31 December 2013, during which commitments were made binding in 27 cases and 21 non-cartel cases ended in infringement decisions.
3.1.3. Fines for non-compliance with commitments
In 2013, the Commission imposed the first fine on an undertaking for non-compliance with a commitment decision. The fine of €561m on Microsoft (corresponding to 1.02% of the total annual turnover) was imposed for non-compliance with the commitment to make available to users a browser choice screen which had been made binding on Microsoft in the browser tying case. [15] In setting the amount of the fine, the Commission considered that (1) “regardless of the specific circumstances of the case, a failure to comply with a commitment decision is, in principle, a serious breach of Union law”, [16] because “it undermines the effectiveness of the mechanism provided for in Article 9 of Regulation (EC) No 1/2003”, [17] (2) the browser choice screen was an essential part of the commitments in the Microsoft-tying decision, (3) the failure to comply with the commitment had been negligent, “due to inadvertent human and technical error” when OEMs had pre-installed Internet Explorer as the default on computers with Windows 7 SP 1, (4) some 15.3 million PCs had erroneously not displayed the browser choice screen because of the omission, (5) the period during which the browser choice screen was not displayed had a duration of 14 months, which was a “significant part of the overall duration of Section 2 of the Commitments (4 years and 39 weeks)”, [18] but (6) Microsoft had immediately admitted the error, invested resources into an independent enquiry, and cooperated in a timely manner with the Commission. [19]
3.2. Commitment decisions before the Court of Justice
The Court has had relatively little contact with commitment decisions. This is not particularly astonishing: part of the attraction of the commitment procedure is the avoidance of appeals. It is this avoidance of appeals that makes the commitment procedure quicker than the infringement procedure; at the administrative stage, there is no good evidence that the commitment procedure is any quicker. [20]
The main reason why there are few appeals against commitment decisions is that the consensual nature of the procedure generally means that the undertaking offering the commitments made binding on it and the European Commission do not have an interest in undoing the compromise, unless the factual basis underlying the compromise is called into question or the underlying circumstances change. It is not even clear whether the undertaking offering the commitments would have standing to appeal the commitment decision if it changed its mind. This means that it will be third parties who appeal against commitment decisions. And there have been a number of such appeals by third parties (see below). However, these appeals have ultimately proved unsuccessful, and the underlying reasoning makes it clear that future appeals will remain unsuccessful except in very exceptional circumstances.
3.2.1. Judicial review of commitments for disproportionality? - Alrosa
In Alrosa, Alrosa as a third party affected by the commitments that had been offered by and made binding on De Beers appealed the decision claiming that the commitments were disproportionate, because they limited the extent to which Alrosa and De Beers could trade with each other. The Court of First Instance had considered commitment decisions to be subject to the general principle of proportionality, requiring the Commission to assess whether intermediate solutions would be sufficient to remove the Commission’s concerns. [21] However, the Court of Justice decided that the Commission’s proportionality analysis in the commitment procedure was “confined to verifying that the commitments in question address the concerns it expressed to the undertakings concerned and that they have not offered less onerous commitments that also address those concerns adequately”, taking interests of third parties into consideration. [22] Moreover, the Court decided that judicial review, in so far as complex economic assessments are in question, is limited to determining whether the Commission’s assessment was “manifestly incorrect”. [23]
3.2.2. Judicial review of commitments as insufficient to address concerns? – Morningstar
In Morningstar, the reverse situation was in question: Morningstar alleged that the commitments imposed on Thomson Reuters were insufficient in so far as they did not address the competitive concerns the Commission had identified in the preliminary assessment. The General Court repeated the statement from Alrosa that the commitment procedure is meant to enable the more rapid resolution of cases, that accordingly the Commission enjoys “wide discretion” in whether to accept or reject commitments, [24] and that the Commission also has a margin of discretion when it came to complex economic assessments; [25] however, the Court added the standard disclaimer—which arguably swallows the rule—that this “does not mean that the EU Courts must refrain from reviewing the EU institutions’ interpretation of information of an economic nature” and that “the EU Courts must, in particular, not only establish whether the evidence relied on is factually accurate, reliable and consistent, but also review whether that evidence contains all the information which must be taken into account in order to assess a complex situation and whether it is capable of substantiating the conclusions drawn from it”. [26]
Nevertheless, combined with the “wide discretion” which the Commission has in deciding whether to accept or reject commitments, this means that judicial review of commitment decisions—regardless whether the claim is that the commitments are disproportionate (as in Alrosa) or that they are insufficient to address the competitive concerns (as in Morningstar)—is largely without teeth, and is at best reduced to a comparison to other sets of commitments offered by the undertakings.
3.2.3. Judicial review of a failure to enforce commitments? – CEEES
The wide discretion of the Commission was also affirmed by the General Court in CEEES. [27]
In 2006, the Commission had made commitments binding on Repsol, in which Repsol committed to refrain, inter alia, from entering into anticompetitive exclusive long-term distribution agreements and from entering into resale price maintentance (RPM) agreements (except for price recommendations or maximum RPM). [28]
Shortly thereafter, in 2007, Repsol nevertheless allegedly entered into prohibited RPM arrangements, and the applicants in CEEES complained to the European Commission. The Spanish NCA investigated against Repsol, and, in 2009, found an infringement of Article 101 TFEU and its Spanish equivalent, ordered termination of the practice, and imposed a fine of €5m on Repsol. The Commission decided not to pursue the matter further. While the applicants withdrew their complaint to the Commission insofar as the substantive infringements of Article 101 TFEU were concerned, because they had been dealt with by the Spanish NCA, they continued to pursue their complaint insofar as they asked the Commission to reopen the proceedings under Article 9(2) of Regulation 1/2003, impose periodic penalty payments under Article 24(1)(c) of this Regulation, and impose a fine under Article 23(2)(c) of this Regulation for non-compliance with the 2006 commitment decision. The Commission rejected this complaint in 2011. The applicants appealed this decision to the General Court. The Court decided that the Commission had wide discretion in prioritising cases and that the resulting standard for judicial review did not permit the judiciary “to substitute their assessment of the European Union interest for that of the Commission, but focuses on whether the contested decision is based on materially incorrect facts, or is vitiated by an error of law, manifest error of appraisal or misuse of powers”. Applying this standard to the case, the Court found no manifest error or misuse in the Commission’s decision not to pursue the matter further in the light of the Spanish NCA’s action against Repsol. The applicants had reasoned that while the Commission generally had wide discretion, the fact that Repsol had infringed its commitments both increased the public interest in an additional fine for this procedural infringement and reduced the resources the Commission would have to dedicate to finding the infringement. The Court held that this was not enough to render the Commission’s assessment manifestly erroneous.
While the General Court’s CEEES decision again demonstrates the wide discretion of the Commission and the concomitant (much reduced) scope for judicial review, the facts before the court made this arguably the worst case possible to test the limits of the Commission’s discretion: First, the Spanish NCA had actually intervened, taken measures to end the infringement for the future, and imposed a fine for the infringement; in this regard, it had essentially done nearly everything the Commission could have done by reopening the proceedings, and in a decentralised system of enforcement it should (ideally) not matter whether a case is pursued by the national NCA or the Commission. [29] Secondly, in most other cases, it is substantially easier to prove the non-compliance with a commitment decision than an infringement of the substantive competition rules. [30] In CEEES, however, the commitment in question was not to engage in the blacklisted restriction of RPM (in what is today Article 4(a) of the vertical Block Exemption Regulation 330/2010); accordingly, proving non-compliance with the commitments entailed substantially everything that proving a substantive competition law infringement would have entailed. Hence, in the CEEES case, action by the Commission would neither have been less resource intensive than prosecuting any other RPM case without a preceding commitment decision, nor would the benefits of the Commission’s intervention have been such as to make the expenditure of these resources worthwhile. Even though the facts of the CEEES case made it particularly unlikely that the Commission’s discretion was fettered to such an extent as to make non-intervention manifestly erroneous, the General Court indicated quite clearly that it would defer to the Commission’s assessment even in less extreme cases. It remains to be seen where the line is which the Commission must not overstep; but it seems that the Court is so deferential to the Commission that it may be hard to find applicants willing to expend resources on a judicial challenge—and two challenges of commitment decisions have been withdrawn, [31] while a further challenge (as it happens, also concerning the Repsol commitment decision) was deemed inadmissible, because the applicant (being unaware of its inclusion in the list of those protected under the commitment decision) had missed the deadline for the appeal. [32]
The General Court in CEEES left open whether it is permissible that an NCA imposes a fine for the substantive infringement of the competition rules and the Commission imposes an additional fine for the breach of a commitment for the same conduct; the Court held only that the Commission is not obliged to impose a second fine for the breach of the commitment. [33]
3.2.4. Commitment decisions and private actions — Gasorba
One issue whose details are still not entirely resolved is to what extent commitment decisions interact with public enforcement by NCAs and national courts and private enforcement before national courts. Recitals 13 and 22 of Regulation 1/2003 seem to indicate that there is no connection whatsoever: recital 13 contains the sentence “[c]ommitment decisions are without prejudice to the powers of competition authorities and courts of the Member States to make such a finding and decide upon the case”, and recital 22 states, inter alia, that “[c]ommitment decisions adopted by the Commission do not affect the power of the courts and the competition authorities of the Member States to apply Articles [101 TFEU] and [102 TFEU].”
Nevertheless, there is a consensus that commitment decisions are not completely irrelevant for national public or private enforcement actions. Where the Commission makes binding on the undertaking commitments that require the addressee to do (or refrain from doing) something, then Art. 16 of Regulation 1/2003 precludes NCAs and national courts from finding that this act (or omission) constitutes an infringement of competition law. [34]
Some have argued that, beyond the limitation outlined in the previous paragraph, the binding effect of the Commission decision exceeds the operative part of the commitment decision, at least if the wording of the commitment decision suggests that certain conduct does not constitute an infringement, or that, where the Commission has considered a commitment decision to be sufficient to address its concerns, NCAs or national courts should not be allowed to find an infringement, at least not without additional factual findings beyond those already made by the Commission. [35] The better view, based on the wording of recitals 13 and 22 of Regulation 1/2003, was always that the commitment decision neither finds nor rejects a finding of an infringement, and that therefore NCAs and national courts are free to investigate the same conduct and reach the decision that it infringes substantive competition law. [36]
This view has also been taken by the Court of Justice in Gasorba, which unsurprisingly used recitals 13 and 22 to justify why a national court could find an infringement based on the same facts on which the Commission had, after a preliminary assessment, made commitments binding on the undertaking. [37] Gasorba concerned—as did the CEEES decision discussed above—the Repsol commitment decision. The plaintiffs in the underlying private action had granted Repsol the usufruct of a plot of land and the petrol station on it, and then entered into a 25-year lease from Repsol, under which Repsol was the plaintiffs’ sole supplier for the entire period. As described above, the Commission had made commitments binding on Repsol in 2006, which, inter alia, obliged Repsol to provide incentives for distributors to terminate early their long-term exclusive supply contracts. [38] In 2008, plaintiffs initiated the underlying private action before the Madrid Commercial Court, with which they sought the annulment of the lease under Article 101(2) TFEU and compensation for harm. Their action was dismissed, a decision that was affirmed on appeal. Plaintiffs pursued a further appeal to the Spanish Supreme Court, which made the preliminary reference to the Court of Justice of the European Union that led to the Gasorba decision. The Court of Justice answered, as mentioned above, that the commitment decision did not preclude a finding of an infringement by national courts. This was unsurprising (although not completely uncontroversial). Much more surprising was the following paragraph: [39]
“Nonetheless, national courts cannot overlook that type of decision [scil.: commitment decision]. Such acts are, in any event, in the nature of a decision. In addition, both the principle of sincere cooperation laid down in Article 4(3) TEU and the objective of applying EU competition law effectively and uniformly require the national court to take into account the preliminary assessment carried out by the Commission and regard it as an indication, if not prima facie evidence, of the anticompetitive nature of the agreement at issue in the light of Article 101(1) TFEU.”
This paragraph has resulted in a wide-ranging discussion about the extent of any such “indication” or “prima facie evidence” of commitment decisions. [40]
4. The commitment procedure in practice in the Member States
As indicated in part 2.2, the commitment procedures and their practice in the Member States are varied, and this special issue provides a panorama of this variety. Space restrictions allow only a couple of remarks connecting the EU practice to the practice in the Member States. First, the eventual outcome of the Gasorba case was that the Spanish Supreme Court declared the exclusive supply contracts to be void, taking the Commission’s commitment decision into consideration. [41] Secondly, with regard to the standard for judicial review, the UK Competition Appeal Tribunal (CAT) established a much more intrusive standard than the European Court of Justice. In Skyscanner, the CAT held that the commitment decision is tested for illegality, irrationality (or ‘Wednesbury unreasonableness’), and procedural impropriety; while this standard does not substitute the Court’s or Tribunal’s assessment for that of the authority, it nevertheless allows the court to enquire into the question whether the depth of the fact-finding exercise of the competition authority was reasonable. [42] My own view is that stricter judicial review (either similar to the Skyscanner standard or to the standard applied by the General Court in Alrosa) would be helpful to rein in too enthusiastic an embrace of the commitment procedure by the competition authorities. [43] However, one also has to acknowledge that stricter judicial review standards are costly for the competition authorities, because they have to ensure in every commitment decision its fact-finding is “appeal proof”. The need to invest substantial resources into appeal-proofing even ultimately uncontroversial decisions may be a factor contributing to the extremely low number of final decisions taken by UK competition authorities. An empirical analysis whether the benefits from a more intrusive judicial review standard in terms of substantive accuracy outweigh the costs would be interesting, if difficult to conceive.
5. Commitment and interim procedure
The commitment procedure has been hailed as providing a rapid response to distortions of competitive markets. More recently, doubts have arisen whether there really are (any or substantial) time and resource savings from the commitment procedure as compared to the infringement procedure at the administrative level. It seems that the commitment procedure takes (nearly) as long and takes up (nearly) as many resources as it takes to get to an infringement decision. Time and resource savings eventuate not at the administrative level but when it comes to appeals, because of the laxer judicial review standard.
If swift intervention in the market is desired, for example in fast-moving digital markets, a move from the currently prevalent commitment decisions to an improved interim measures procedure would be desirable; and best results would arguably be achieved by combining the interim measures procedure with the commitment procedure, resulting in “negotiated interim measures”. [44] But this is a matter for another day.
