Market dynamics in the counterfactual: more competitive, not just cheaper

The judgment of Phillips J in Sainsbury’s v Visa [2017] EWHC 3047 (Comm) demonstrates the importance to claimants in competition damages cases of identifying a counterfactual which not only involves lower prices but also involves higher levels of competition.

Sainsbury’s case

Visa’s payment card scheme required ‘acquirers’ (who process card payments on behalf on merchants) to pay an ‘interchange fee’ to the issuer of a payment card whenever a payment was made. All acquirers were required to accept all cards issued in the scheme (the so-called honour all cards rule or ‘HACR’). All issuers were required to remit to the acquirer the whole of the payment made by the customer, less the applicable interchange fee (this was called the ‘settlement at par’ rule). Acquirers passed on all of the interchange fees to merchants, as part of the merchant service charge which also included an element of profit margin for acquirers.

Visa set a default interchange fee (the multilateral interchange fee or ‘MIF’), though acquirers and issuers were free to negotiate different fees bilaterally. However, no acquirer had an incentive to agree to pay more, and no issuer had an incentive to agree to accept less than the MIF. The settlement at par rule prevented issuers from, in effect, forcing a higher interchange fee on acquirers by remitting customer payments at a discount. The HACR prevented acquirers with market power from forcing lower interchange fees on issuers by refusing to accept cards unless bilateral interchange fees were agreed.

The combined effect of these rules was to eliminate any competition as to the level of interchange fees. This was Sainsbury’s case and Visa accepted that these arrangements constituted a restriction of competition ‘in absolute terms’: [103-104].

Sainsbury’s proposed counterfactual, which the court accepted, was one in which there was no MIF set by Visa, but the settlement at par rule and HACR remained in force: [98]. This, in effect, amounted to a MIF of zero and the same dynamics between issuers and acquirers as existed in the factual would have prevented bilateral interchange fees from being agreed in the counterfactual: [126-129]. It followed that on Sainsbury’s counterfactual, the interchange fees paid would have been lower but not because of any re-introduction of competition in the setting of such fees. The suite of rules which had operated to eliminate competition in the factual would operate in the same way and with the same result in the counterfactual, just at different prices: [161]. Sainsbury’s claim therefore failed as it had not established that its loss was caused by a reduction in competition.

What went wrong?

Sainsbury’s counterfactual retained two key elements from the factual scheme: the settlement at par rule and the HACR. All parties agreed that the scheme would be unworkable if it incorporated the HACR but did not require settlement at par: [99]. Such a scheme would be equivalent to allowing issuers to set interchange fees unilaterally, by settling payments at a discount. Assuming acquirers continued to pass interchange fees on to merchants in full, this would result in merchants ceasing to accept Visa, as continued participation in the scheme would mean accepting all card payments regardless of the fee charged.

The parties do not appear, however, to have explored the possibility of a truly bilateral system in which neither the settlement at par rule nor the HACR applied. In a counterfactual from which both of these rules were absent, issuers and acquirers would have been forced to negotiate terms of settlement bilaterally. Issuers would have wished to agree a higher interchange fee (or a larger discount from par) but would have been prevented from demanding too high a fee because of the risk that acquirers (in order to retain merchant business) would cease to accept a certain issuer’s cards.

This counterfactual involves a radical departure from the Visa scheme in the factual, but the market dynamic which would result is familiar: this is how competition works in so-called three-party schemes, such as American Express. The issuer faces competing incentives: higher fees make for greater profits per transaction, but too high fees reduce card acceptance by merchants and reduce transaction volume, ultimately reducing the appeal of the card to customers.

The outcome in this counterfactual could be a patch-work of differing fees charged by different issuers with corresponding variances in merchant acceptance. There are over 50 issuers and around 30 acquirers in the UK[1], which would give rise to a large (but in principle manageable) number of bilateral negotiations assuming the same number of issuers existed in the counterfactual. There may be good reason, however, to think that there would be fewer issuers in the counterfactual: If the setting of MIFs inhibited competition and raised issuer profits, it is likely also to have encouraged more issuers to enter the market than could have been sustained in a competitive scenario. Lower interchange fees, lower profits, the requirement to negotiate deals with all (or almost all) acquirers and the need to reassure prospective customers that the card would be widely accepted would all act to restrain the number of viable issuers in the counterfactual.

Merchants now indicate whether they accept Amex or Diners. Before the emergence of four-party schemes in the UK, merchants indicated whether they accepted Barclaycard, a card issued only by the bank of the same name. In the counterfactual described here, merchants would be required to indicate which of the major banks’ and independent issuers’ cards they accepted.

Of payment cards and dog races: monopoly and monopsony in price setting

Competition within the Visa scheme was inhibited by the fact that fees were set centrally for all participants. The scheme was controlled by participating issuing banks, so central price setting tended to result in higher fees. If merchants (or acquirers who did not also issue cards) had controlled the scheme, the result might have been that low or even negative interchange fees were set, but the result would have been no more competitive. The mischief against competition was the setting of prices centrally; control over the scheme determined which party stood to gain.

In Bookmaker’s Afternoon Greyhound Services [2009] LLR 584, cited by Phillips J at [91], the boot was on the other foot. Whereas in the Visa scheme, prices were set by or on behalf of the party receiving payment, in the BAGS case, prices were set by the paying party, which was controlled by the leading bookmakers, and which had a monopsony on buying live television footage from racecourses. When a group of racecourses jointly agreed to sell their footage exclusively through a newly formed distributor, prices for footage rose and BAGS claimed that the racecourses had acted anti-competitively. On the contrary, the court found, the market power of BAGS had been reduced and competition had been increased. As Phillips J pointed out in Sainsbury’s, the BAGS case shows the pitfalls of a facile analysis which equates price decreases with increases in competitive intensity and vice-versa.

It ought to be possible for the payment card market to operate competitively given the large number of retailers, banks and other issuers. Where collective price-setting is interposed between the parties on each side of the transaction, there is an obvious possibility of distortions to competition. It was common ground in Sainsbury’s that the Visa scheme as implemented had eliminated competition in the setting of interchange fees: [103-104]. A claimant seeking damages for the setting of payment card interchange fees should ask the court to consider a simple counterfactual, in which these distortions are eliminated: What would the result have been if the parties on each side had decided whether to transact with one-another and on what terms, without the central setting of prices or a compulsion to transact?

[1] In Arcadia v MasterCard [2017] EWHC 93 (Comm), at [103] Popplewell J found that there were 55 MasterCard issuers in 2015 in the UK; it is assumed that there were a similar number of Visa issuers.

Leave a comment

Filed under Abuse, Agreements, Damages, Economics

Intel Corporation Inc v European Commission

In its recent judgment in Intel, the Grand Chamber shed valuable light on the “qualified effects test” for jurisdiction and on the room for loyalty rebates to be compatible with competition law.


Intel designed computer processors and sold them to original equipment manufacturers (“OEMs”) to use in central processing units (“CPUs”). One of its competitors, Advanced Micro Devices Inc (“AMD”), complained to the Commission that Intel was abusing its dominant position by offering loyalty rebates to its OEMs if they purchased all or most of their processors from Intel.  The Commission agreed and imposed a €1.05 billion fine. The General Court dismissed Intel’s appeal.

On appeal, the Grand Chamber of the CJEU rejected Intel’s complaints about jurisdiction and procedural irregularities but allowed its appeal on the assessment of the rebates as abusive.  That rendered three other grounds unnecessary to consider.

There are two key points of interest arising from the judgment:

  1. Arrangements that are intended to form part of a grander anti-competitive scheme may fall within CJEU jurisdiction, even though they are relatively removed from the EEA, under the “qualified effects” route to jurisdiction.
  2. Loyalty rebates are not automatically anti-competitive; in particular, they can be saved if the undertaking can show that they could not have the effect of foreclosing an as efficient operator from the market.

The qualified effects test for jurisdiction

The Court considered two tests for jurisdiction.

  1. The “implementation test”:  Were the anticompetitive practices implemented in the EEA?
  2. The “qualified effects test”: Would the practices have foreseeable, immediate and substantial effects in the EEA?

The General Court had found that it had jurisdiction over Intel’s agreements with Lenovo (a Chinese OEM) on both tests. Intel unsuccessfully challenged the latter as a valid route to jurisdiction, and the Court’s application of both.

The Court confirmed that the “qualified effects test” is a valid route to jurisdiction.  Although the test had previously been accepted by the General Court in Gencor v Commission (T‑102/96, EU:T:1999:65) at §92, this is the first time it has been recognised by the CJEU.  It explained that it pursues the objective of “preventing conduct which, while not adopted within the EU, has anticompetitive effects liable to have an impact on the EU market”. If EU competition law were confined to the places where agreements were reached or concerted practices engaged in, it would “give undertakings an easy means of evading” Articles 101 and 102 (§§41-45).

How then is that test to be applied?  The Court provided some guidance. The question is whether “it is foreseeable that the conduct in question will have an immediate and substantial effect in the European Union”. In answering that question it is necessary to examine the undertaking’s conduct as a whole.  On the particular facts (§§51-57):

  1. The agreements with Lenovo in China had a “foreseeable” impact on competition, taking account of their “probable effects”.
  2. They had an “immediate” effect, because they formed part of an overall strategy to ensure that no Lenovo notebook equipped with an AMD CPU would be available on the market.
  3. They had a “substantial” effect on the EEA market, having regard to the whole of the conduct.

This last point is the most interesting one. Even though agreements with Lenovo for CPUs for delivery in China would by themselves have had a negligible effect, they formed part of conduct that would have a substantial effect. The Court refused to examine them in isolation on the basis that such an approach would “lead to an artificial fragmentation of comprehensive anticompetitive conduct”.

Loyalty rebates

The main substantive implications of this case arise from the findings that loyalty rebates are not always be abusive: it will depend on their scope and effect.

The purpose of Article 102 is to promote, not inhibit, competition.  So it does not protect undertakings which are not as efficient as the dominant undertaking.  Rather, it prevents illegitimate competition that pushes equally efficient undertakings out of the market.  That includes forcing purchasers to meet their requirements from the dominant undertaking.  It also includes inviting purchasers to undertake a contractual obligation to do so.  By extension, it might include incentivising purchasers to do so through loyalty rebates.  But, the Court has now made clear, that latter category is not inherently abusive.

In order to determine whether it is, it is necessary:

  1. First, to consider all the circumstances, including the level and duration of the rebates, the market shares concerned, and the needs of customers. Most importantly, the Commission must consider the capability of the rebates to foreclose an “as efficient competitor” (the “AEC test”).  That is, could the rebates force such a competitor to sell below cost price?
  2. Second, even if the rebates do have an exclusionary effect, they might still be redeemed if that effect is counterbalanced by efficiency advantages (§§139-140).

There are three key points of interest.

First, it appears that the general rule remains that loyalty rebates are abusive, unless the undertaking can produce evidence to the contrary.  The Court recounted that loyalty rebates have an anti-competitive effect, and “clarified” its case-law to say that undertakings can displace that presumption by showing that they could not have that effect in the particular case.  That brings Article 102 in line with the position under Article 101.

Second, the Court made clear that the AEC test applies generally to assessing whether conduct is an abuse of dominant position. Article 102 is not calculated to come to the aid of less efficient undertakings. Accordingly, to determine whether the practice is illegitimate, it is necessary to determine the effect it would have on a competitor who is as efficient as the dominant undertaking. That principle had been applied to attracting purchasers and excluding competitors by predatory pricing in Post Danmark v Konkurrencerådet (C‑209/10) and AKZO Chemie BV v Commission (C-62/86). In a victory for consistency, it is now clear that it applies more generally, including to attracting purchasers and excluding competitors by loyalty pricing schemes.

Third, in the context of that AEC test the Court said the inquiry was as to the “capability” of the rebates to foreclose an as efficient competitor (§§138, 141), even though Intel’s objection was that the General Court had failed to consider the “likelihood” of the rebates having that effect (§§113-114).  That is a harder task for an undertaking seeking to avoid breaching the Article 102 prohibition.  However, it is also consistent with other cases of actions with an anti-competitive object (which are less easily excused, only if they could not have that effect) rather than those with an anti-competitive effect (which are, for obvious reasons, excused if they are not likely to have that effect).

Leave a comment

Filed under Abuse, Conflicts, Economics

Collective (in)action? The CAT’s recent judgments on collective proceedings orders

At first glance, two recent judgments from the CAT may give the impression that the new UK class action regime is dead in the water. However, on closer inspection there is much in these judgments that prospective claimants will welcome.

The first decision was in the Pride mobility scooters case (see Tom Coates’ blog here). The CAT made clear that it might have been prepared to grant a collective proceedings order (“CPO”), but on a basis so narrow that the claimants chose not to proceed. In the second decision, Merricks v Mastercard Inc & Ors [2017] CAT 16, the CAT rejected the CPO application, bringing an end to what would have been an extraordinarily ambitious claim—on behalf of 46.2 million people, seeking aggregate damages of approximately £14 billion, for Mastercard’s unlawful setting of fallback multilateral interchange fees in breach of Article 101 TFEU.

Under the new provisions in s.47B of the Competition Act 1998, a CPO application must satisfy the CAT of two criteria. They are, in brief, that (i) the person bringing the proceedings is an appropriate representative of the class of claimants, and (ii) the claims are eligible for inclusion in collective proceedings.

In Merricks, as in Pride, the applicants succeeded on the first criterion but failed on the second. The CAT adopted a relatively liberal approach to certifying the class representative in both cases: a former ombudsman and consumer protection advocate in Merricks (§§93-94), and an advocate for pensioners’ rights in Pride (§§125-139).

The CAT was also satisfied with the litigation funding arrangements in both cases (Pride, §§140-145; Merricks, §§95-140); although it strongly criticised the “impenetrable” drafting of the American-style funding agreement in Merricks, and was only prepared to approve it in light of amendments proposed at the hearing: §§121-127. Prospective claimants will welcome the fact that, in neither Pride nor in Merricks was the CAT unduly concerned by the prospect of a shortfall between the applicants’ costs cover and respondents’ likely costs.

Where both claims failed, however, was on the eligibility criterion. This second criterion is further broken down in rule 80 of the CAT Rules 2015, which provides that claims will be eligible for inclusion in collective proceedings where they (a) are brought on behalf of an identifiable class of persons; (b) raise common issues; and (c) are suitable to be brought in collective proceedings.

In both cases, the CAT was prepared to accept that the claims were brought on behalf of an identifiable class of persons. In Pride that conclusion was uncontroversial, given that the class was defined as “any person who purchased a new Pride mobility scooter other than in the course of a business in the UK between 1 February 2010 and 29 February 2012” (§§5, 85). In Merricks, however, the CAT’s apparent acceptance of the class was no small matter. The class included all individuals who were over 16 years old at the time of the transaction, resident in the UK, and who purchased goods or services from UK businesses which accepted MasterCard cards, at any time over a 16 year period (§1). This included more than 46 million potential claimants; and yet, the CAT was untroubled by the “identifiable class” criterion.

As to the requirement that the claims raise common issues, in both cases the CAT emphasised that the appropriate approach was that followed in Canada, rather than the much stricter approach in the United States (Merricks, §58; Pride, §105). Although only three of the six issues in Merricks could properly be regarded as common, the CAT considered that to be sufficient.

In Pride, the applicant faced the difficulty of proving causation in circumstances where the regulator had focused on a small sample of infringing agreements (“the low-hanging evidential fruit”: §109), and the claimants were time-barred from pursuing anything other than a follow-on claim for the infringement (§110). The CAT’s decision on this issue may well create difficulties for other follow-on vertical infringement claims, but that category of claims is likely to be quite narrow.

In Merricks, the CAT was concerned about the methodology by which the applicant proposed to assess individual losses. The methodology needed to distinguish between three sets of issues: “individuals’ levels of expenditure; the merchants from whom they purchased; and the mix of products which they purchased” (§88). Regrettably, there had been “no attempt to approximate for any of those in the way damages would be paid out” (§88). The CAT observed that the experts’ oral evidence in response to questions from the Tribunal was “considerably more sophisticated and nuanced than that set out, rather briefly, in their Experts’ Report” (§76), but it still could not be satisfied that the damages sought would broadly reflect “the governing principle of damages for breach of competition law”, that is, “restoration of the claimants to the position they would have been in but for the breach” (§88). The judgment sounds a valuable warning to future claimants of the necessity for a detailed and precise methodology for calculating both individual and aggregate losses.

The CAT showed little sympathy for the applicant’s argument that refusing the CPO would result in a vast number of individuals who suffered loss going uncompensated, since there was no realistic prospect of claimants pursuing Mastercard individually. The CAT observed shortly that this was “effectively the position in most cases of widespread consumer loss resulting from competition law infringements” (§91).

The judgments in Pride and Merricks provide important guidance on the CAT’s likely approach to CPOs in future. In spite of the outcomes in both cases, the CAT’s ready acceptance of the proposed class representatives, its flexibility in regard to litigation funding, and its affirmation of the Canadian approach to collective action, are all likely to give heart to prospective claimants. Further, the judgment in Merricks leaves the door open to mass claims in the future, while signalling the heightened importance which expert evidence on calculating losses is likely to assume in such cases.

Leave a comment

Filed under Abuse, Agreements, Damages, Economics, Policy, Procedure

Brexit and implications for UK Merger Control – Part 3/3: Managing and prioritising the CMA’s mergers workload

The Competition Bulletin is pleased to welcome the third in a three-part series of blogs on Brexit and merger control by Ben Forbes and Mat Hughes of AlixPartners.  Ben and Mat are (with others) co-authors of the new Sweet & Maxwell book, “UK Merger Control: Law and Practice”.

Part one focused on the issues associated with the voluntary nature of UK merger control (and can be found here), and part two considered options for change that in our view should not be adopted (and can be found here).


Prioritising the CMA’s work load is clearly important due to Brexit.  This is because the European Commission will cease to have exclusive jurisdiction over large UK mergers that currently fall for consideration under the EU Merger Regulation. In March 2017, the Competition and Markets Authority (CMA) indicated that this could lead to increase in its mergers caseload by 40-50% since 1 April 2014, potentially amounting to an additional 30-50 phase 1 cases and six phase 2 cases per year.

This final blog in our series considers how the CMA could prioritise and manage its mergers workload, with particular focus on the CMA’s consultation of 23 January proposing amendments to its guidance on the application of the exception to its duty to refer a merger in markets of insufficient importance (i.e. the de minimis exception).[1]

The CMA is proposing increasing the upper threshold for markets considered to be sufficiently important to justify a merger reference from £10 million to £15 million, and would raise the lower threshold for markets not considered to be sufficiently important from below £3 million to below £5 million.  Where the size of the market falls between these two thresholds, the CMA would continue to evaluate, on a case-by-case basis, the potential harm caused by the merger against the cost of an investigation.

The importance of the de minimis exception in UK merger control

The de minimis exception is designed to save the CMA[2], and therefore the public purse, money by not referring insignificant mergers to phase 2. These costs are material as the National Audit Office’s 2016 report on the UK Competition Regime estimated that the average cost of a phase 2 investigation to the CMA is £275,000.[3] Moreover, this figure does not include any costs to the merging parties (or third parties), and for a large complex phase 2 investigation, these costs are high and often substantially more than the CMA’s costs.

The de minimis exception has become an important part of UK merger control.  It is particularly important since it only applies where, in principle, clear-cut undertakings in lieu of reference could not be offered, and the parties would thus otherwise face the costs and risks of a phase 2 investigation. In particular, over the last seven years, 28 mergers have been cleared on de minimis grounds, and absent this exception the number of merger references would have increased by 46 per cent (28/61). (Over this period, a further 40 mergers were cleared conditionally at phase 1 on the basis of undertakings in lieu of reference).

As many cases affect markets worth under £10 million per annum, any assessment of merger control risks needs to consider even overlaps in relation to even small parts of the merging parties’ businesses.

The Office of Fair Trading’s (OFT) guidance on exceptions to its duty to make merger references, which has been adopted by the CMA, also indicates that, for now, it draws a distinction between markets with an annual value of below £3 million and those with values of between £3m and £10m.[4] In particular, their guidance notes that it would “expect to refer a merger where the value of the market(s) concerned was less than £3 million only exceptionally, and where the direct impact of the merger in terms of customer harm was particularly significant”. For mergers between £3m and £10m, the OFT/CMA will weigh up the size of the market and the likely harm to customers, as well as considering the wider implications of the decision.

Our analysis of 400+ UK merger decisions since 1 April 2010 also included many cases where the OFT/CMA considered applying the de minimis exception. As set out in the chart below, there were 45 cases between 1 April 2010 and 31 March 2017 where the OFT/CMA considered applying the de minimis exception and how the decision reached varies according to the size of the relevant markets affected.[5]


Consistent with the OFT’s guidance, there were very few cases where the relevant market was valued at under £3 million per annum where the CMA/OFT nevertheless decided to refer the merger for a phase 2 investigation. An analysis of all the various cases is set out in the next chart, and the two exceptions relate to mergers between local bus operators, where the OFT and CMA respectively decided that these mergers warranted phase 2 investigation.  This reflected a recommendation from the Competition Commission for a cautious approach following its market investigation into local bus services.

Where the relevant market size is greater than £3 million per annum, a merger reference is more likely. This is set out in detail in the following chart, which covers all 45 merger decisions that considered the de minimis exception from 1 April 2010 through to 31 March 2017.  This chart also covers the phase 2 outcomes of those cases where the OFT/CMA decided not to apply the de minimis exception and referred the merger.


Note: The OFT/CMA decisions often only included a range for the relevant market size. Therefore, the bars represent the expected market size (i.e. the middle of the range) where applicable. The top of the range is represented by the black triangular dotes.

This review highlights the rather sharp distinction in outcomes between mergers that affect markets with an annual aggregate value of around £3 million per annum and those valued between £3 million and £10 million.

Looking more closely at the 17 cases that were referred at Phase 1 despite the small sizes of the markets affected, seven of these mergers were abandoned. This is entirely unsurprising given the high costs to the parties of phase 2 investigations, which will often exceed merger synergies in small markets.  For non-abandoned mergers, three cases were cleared unconditionally, six were cleared with remedies, and only one prohibited. Accordingly, it seems reasonable to speculate that some of the seven abandoned mergers would have been cleared either unconditionally or with remedies.

The CMA’s proposed changes in market size thresholds

At first sight, the proposed changes in the thresholds at which the de minimis exception will apply suggests that it may be applied substantially more often.

However, some words of caution are warranted.  First, as a matter of policy, the CMA will not apply the de minimis exception if, in principle at least, clear cut undertakings in lieu of reference could be offered.

Second, considering the CMA’s proposed £5 million threshold for mergers where the de minimis exception will generally apply, since 1 April 2010 only another two referred mergers would fall into this category.

Third, turning to those mergers with turnover between £5 million and the proposed £15 million threshold, the de minimis exception may be considered in many more cases. However, historically the OFT/CMA has referred many of the mergers below the £10 million upper bound, and it remains to be seen whether this will be the case in relation to the higher £15 million threshold.


As discussed in part one and part two of this blog series, productively improving the UK merger control regime is not simple. The CMA has had to address the changes associated with becoming a single authority covering phase 1 and phase 2 decisions, and having a full prenotification regime and statutory phase 1 deadlines.

In our view, the CMA’s proposed changes to the de minimis thresholds is sensible. However, it remains to be seen whether this will substantially reduce the CMA’s workload, and the burden of UK merger control on small mergers.


[2]   As of the 1 April 2014, the OFT transferred its merger related functions to the newly created CMA. CMA has been used throughout this document, but if any examples occur before this date, read as OFT.

[3] See paragraph 2.2:

[4] OFT, “Mergers, Exceptions to the duty to refer and undertakings in lieu of reference guidance”, December 2010.

[5]    Note that this excludes certain mergers where the OFT/CMA briefly considered applying the de minimis but where the parties could offer clear-cut undertaking in lieu of reference to address the competition concerns identified. In these cases, the OFT/CMA did not consider the size of the relevant market (see for, example, Reed Elsevier / Jordon Publishing (2015)).

Leave a comment

Filed under Mergers

Illegal counterfactuals: the Court of Appeal shuts the back door

Suppose a defendant to a competition claim runs a defence that, in the counterfactual world in which no anticompetitive conduct occurred, pricing would have been no different; and that the claimant replies, “maybe so, but only because you were at the same time operating some independent anti-competitive scheme, which must also be purged from the counter-factual”. Can the claimant amend his claim to plead the independent anti-competitive scheme raised in his Reply as the basis for a new substantive claim even where it would ordinarily be time-barred?

In February last year, Barling J appeared to answer, “Yes”, in a judgment given in the MasterCard litigation. On one view, the curious result of that judgment was that a claimant could apparently circumvent limitation rules by introducing a time-barred allegation of unlawfulness in his Reply, then using that as a basis to apply to amend his original claim. In other words, when a limitation point blocked the front door, claimants could still bring in new claims through the back.

The Court of Appeal, however, has now shut this back door, by overturning the High Court’s judgment. For the background to the judgments, and the details of Barling J’s decision, see my previous post here.

The issue before the Court was whether or not the new claim, premised on MasterCard’s Central Acquiring Rule (CAR) arose out of the same or substantially the same facts as the existing claim, premised on MasterCard’s Multilateral Interchange Fees (MIFs) (see CPR 17.4 and section 35(5) of the Limitation Act 1980). If it did, the Court could permit an amendment notwithstanding that it was time-barred. Barling J had held that it did on the following two grounds: first, the existing claim would already require an investigation into and evidence on the CAR; and, secondly, the claimants’ reply had pleaded that the CAR was unlawful and had to be excised from MasterCard’s counterfactual – so the new claim arose out of facts already in issue with respect to the existing claim.

The Court of Appeal disagreed with Barling J on both scores. Sales LJ said that the facts underlying each claim could not be said to be the same because the counterfactual inquiry required by each claim was so different (§46). On the existing claim, the counterfactual world was one in which the MasterCard rules in dispute (principally the MIFs) were excised but the CAR remained in place. On the new claim, however, the Court would have to investigate both the counterfactual world in which the MasterCard rules were excised as well as the CAR and the counterfactual world in which all the MasterCard rules remained in place but the CAR was excised.

Sales LJ, doubting the obiter comments of Waller LJ in Coudert Brothers v Normans Bay Ltd [2004] EWCA Civ 215, further said that the claimants could not introduce the new claim by pointing to their reply and saying that the CAR’s lawfulness was already in issue. The proper rule was that, where the defendant had pleaded facts by way of defence to the original claim, the claimant could introduce a new claim premised on those facts: Goode v Martin [2002] 1 WLR 1828. However, that was not the case here because MasterCard did not specifically rely on the CAR in its defence.

The Court of Appeal was further clearly motivated by a concern about the avoidance of limitation rules. Sales LJ said at §64:

“…it would be unfair to a defendant and would improperly subvert the intended effect of limitation defences set out in the Limitation Act if a claimant were to be able to introduce new factual averments in its reply (which are not the same as or substantially the same as what is already pleaded in the claim), after the expiry of a relevant limitation period, and then rely on that as a reason why it should be able to amend its claim with the benefit of the “relation back” rule to circumvent that limitation period.”

The curious result of Barling J’s judgment has therefore been reversed by the Court of Appeal. A claimant can no longer pull himself up by his own bootstraps; limitation now guards the back door as jealously as the front.


Leave a comment

Filed under Abuse, Agreements, Damages, Procedure

Collective Proceedings in the CAT: mobility scooters roll on for now

Last Friday the CAT handed down a judgment on the first ever-application for a collective proceedings order under the new regime introduced by the Consumer Rights Act 2015. The judgment will generally be welcomed by potential claimants, but it has a sting in the tail which may cause serious difficulties for class actions in other vertical infringement cases.

The new collective proceedings regime, contained in section 47B CA98 and CAT Rules 75-98, was one of a suite of claimant-friendly measures aimed at improving the remedies for individual victims of competition infringements whose losses were low (other measures included, for example, the new fast-track procedure). Consistently with the regime’s objective, the CAT, although stopping short of reaching a final decision, said much about the scheme which will encourage claimants.

The proposed collective action is a follow-on claim against Pride, formerly the UK’s largest supplier of mobility scooters. The OFT (the national competition regulator) had found that Pride infringed the Chapter I prohibition by object by entering into 8 vertical agreements with retailers forbidding them from advertising mobility scooters online at prices below RRP. Those 8 agreements were in fact the result of a market-wide policy that Pride had been operating and which it had communicated to all of its retailers.

The key issues before the CAT were (broadly) the authorisation of Dorothy Gibson as the class representative (under section 47B(5) and CAT Rule 78), and certification of the claims for inclusion in collective proceedings on the basis that they raised common issues. On both issues, the CAT’s approach to the claimants was benevolent.

The CAT first dismissed the defendant’s preliminary objection. Pride pointed out that both the OFT decision and the underlying infringement pre-dated the introduction of the collective proceedings regime. On that basis, it fired a salvo of arguments based on Article 1 Protocol 1 of the ECHR, the EU Charter, and EU principles of legal certainty/legitimate expectations, the thrust of which was that the CAT should interpret the regime so as to disallow its “retrospective” application. The CAT shot down all these arguments in a comprehensive discussion that should see the end of any similar threshold points about collective proceedings applications.

The CAT also had little difficulty in authorising Ms Gibson as the class representative. Although not a mobility scooter consumer, her status as an advocate of pensioners’ rights (she is the chair of a representative body, the National Pensioner Convention), who had experienced lawyers, satisfied the CAT that she would act fairly and adequately in the interests of the class (§139; see CAT Rule 78(2)(a)). Moreover, the CAT was not concerned about her ability to pay Pride’s costs (see CAT Rule 78(2)(d)). Even though Ms Gibson’s ATE insurance cover level was less than Pride’s anticipated costs, the CAT stated shortly that those costs might not be reasonable or proportionate, so it would not be appropriate to disallow collective proceedings at that stage (§145).

The CAT’s approach to certification and commonality was also – in principle – liberal. Although it said that it could not “simply take at face value” (§102) the applicant’s expert evidence, it nonetheless rejected Pride’s submission that it should take a hyper-rigorous US-style attitude. Rather, the CAT endorsed the Canadian approach, approving at §105 the following comment of Rothstein J in Pro-Sys Consultants Ltd v Microsoft Corp [2013] SCC 57:

“In my view, the expert methodology must be sufficiently credible or plausible to establish some basis in fact for the commonality requirement. This means that the methodology must offer a realistic prospect of establishing loss on a class-wide basis so that, if the overcharge is eventually established at the trial of the common issues, there is a means by which to demonstrate that it is common to the class (i.e. that passing on has occurred). The methodology cannot be purely theoretical or hypothetical, but must be grounded in the facts of the particular case is question.”

The stumbling block for the claimants, however, was the CAT’s reasoning on the proper counterfactual. The claimants posited a world in which not only the 8 infringing vertical agreements were absent but also where Pride had operated no policy of prohibiting below-RRP advertising at all. The CAT, however, endorsed a narrower counterfactual from which only the specifically unlawful agreements (i.e. the 8 vertical agreements about which the OFT had made findings) were assumed to be absent (§112). With this narrower counterfactual, the CAT held that it was not clear whether there was sufficient commonality in the issues of loss, or whether the likely damages would justify the costs of collective proceedings. However, the CAT (again perhaps generously) did not dismiss the application altogether but rather adjourned it to enable the claimants’ expert to formulate a case on common loss on the basis of the revised counterfactual.

Notwithstanding the generally claimant-friendly approach, the CAT’s reasoning on the counterfactual could render other collective cases premised on vertical restraints very difficult in practice. In a large number of vertical restraint decisions, the infringer has adopted a market-wide policy but the regulator focuses, for practical reasons, on a small number of ‘implementations’ of the policy as the basis for its infringement findings. If one only excludes from the counterfactual the particular instances of unlawful implementation, rather than the more general policy which underlay them, the issues between class members may diverge: some will have been direct victims of the anti-competitive agreements, while others will have suffered only from what would have to be characterised as an “umbrella effect”. In addition, if claimants cannot proceed on the basis that the entire policy should be excluded from the counterfactual, the likely quantum may fall to a level where collective proceedings are not worth the candle. It remains to be seen whether the mobility scooter claimants will overcome these difficulties.

1 Comment

Filed under Agreements, Damages, Policy, Procedure

When is an antitrust/competition claim caught by an arbitration clause? The Microsoft Mobile decision

The decision of the High Court in Microsoft Mobile Oy (Ltd) v Sony offers some helpful guidance as to when a competition law tort claim will be caught by an arbitration clause in a sale or supply agreement.

Competition law claims frequently complain about prices, on ground of collusion or abuse.  Those prices may already have been charged, or they may yet to be charged.  If the price in issue has already been charged then it will almost invariably be contained in a written sale or supply agreement.  It may be the product of a specific contractual mechanism to settle a price.  If the relevant agreement contains an arbitration clause, does it catch a competition law claim complaining about the price charged?

The starting point in this regard is that a detailed semantic analysis of the particular arbitration clause in question is unlikely to provide the answer.  As Lord Hoffmann observed in Fiona Trust v Privalov [2007] 4 All ER 951 at para 13:

“…the construction of an arbitration clause should start from the assumption that the parties, as rational businessmen, are likely to have intended any dispute arising out of the relationship into which they have entered or purported to have entered to be decided by the same tribunal. The clause should be construed in accordance with this presumption unless the language makes it clear that certain questions were intended to be excluded from the arbitrator’s jurisdiction.”

This “one-stop shop” presumption has helped greatly in reducing disputes about the ambit of arbitration clauses.  But its application in the competition law context has been less straightforward, because of debates about precisely what the rational businessman would intend in respect of such claims.

The net effect of the law to date is that competition law claims will be regarded as coming within an arbitration clause only if they are closely related factually to a viable contractual claim which has already been, or could be, made.  Thus, for example, in ET Plus SA v Welter [2006] I.L.Pr. 18, the key consideration for Gross J in finding that competition law claims were within an arbitration clause was that they were “simply a variant on the familiar factual theme” which could be discerned from contractual claims already made (see para 51).

Conversely, in Ryanair Ltd v Esso Italiana Srl [2015] 1 All ER (Comm), the Court of Appeal held that the absence of any viable form of contractual complaint about an allegedly cartelised price rendered it impossible to claim that a competition law complaint about the same price was within an exclusive jurisdiction clause.  Having referred to the one-stop shop presumption in Fiona Trust, Rix LJ then held at para 53 that:

“Such reasoning, however, does not carry over into a situation where there is no contractual dispute (by which I intend to include disputes about contracts), but all that has happened is that a buyer has bought goods from a seller who has participated in a cartel. I think that rational businessmen would be surprised to be told that a non-exclusive jurisdiction clause bound or entitled the parties to that sale to litigate in a contractually agreed forum an entirely non-contractual claim for breach of statutory duty pursuant to article 101, the essence of which depended on proof of unlawful arrangements between the seller and third parties with whom the buyer had no relationship whatsoever, and the gravamen of which was a matter which probably affected many other potential claimants, with whom such a buyer might very well wish to link itself.”

The English Courts therefore treat competition law claims essentially as falling outside the one-stop shop presumption unless they are, factually, simply a variant on the theme of an arguable contractual claim.  This was, furthermore, the approach of the CJEU in Case C-352/13 CDC v Akzo [2015] QB 906, in which it was held that a clause “…which abstractly refers to all disputes arising from contractual relationships” would not cover tortious liability as a result of a cartel, because “…the undertaking which suffered the loss could not reasonably foresee such litigation at the time that it agreed to the jurisdiction clause” (paras 69-70).

The latest word on this topic is the Microsoft decision.  Microsoft brought a claim in the English Courts for damages for the allegedly anti-competitive tortious conduct of Sony, LG and Samsung in relation to the pricing of Li-Ion batteries.  All the allegedly cartelised supplies by Sony had been made pursuant to an agreement with an arbitration clause requiring “any disputes related to this Agreement or its enforcement” to be settled by ICC arbitration.  Sony applied to stay the proceedings under section 9 of the Arbitration Act 1996, arguing that the arbitration clause covered the tort claims made against it.

Mr Justice Marcus Smith held that, on an orthodox application of the principles identified in Ryanair, the question of whether the tortious claims were within the arbitration clause depended on whether the conduct giving rise to the tortious claims also gave rise to an arguable contractual claim.  As he observed, “…it is difficult to see how a tortious claim can arise out of a contractual relationship when the only claim in contract that can be said to be related is unarguable” (para 54).

The unusual incentives created by this (correct) understanding of Ryanair can be seen from the fact that Sony was accordingly required, in order to succeed in its application, to formulate a contractual claim against itself which Microsoft had not advanced.  Sony argued that because the relevant prices had been subject to an express obligation that they be negotiated in good faith, and because Sony was subject to a further obligation to disclose events that reasonably may affect its ability “to meet any of its obligations” under the agreement, the operation of a cartel would have been a clear breach of contract as well as tortious.

The Court accepted this submission, holding that it was “very difficult” to see how Sony could have engaged in the conduct complained of in the tort claims, without also breaching the contract.  On that basis, the competition law claims fell within the arbitration clause.  It did not matter in this regard that Microsoft had not advanced the contractual claim which Sony successfully contended it could have done, because otherwise “…it would be easy for a claimant to circumvent the scope of an arbitration or jurisdiction clause by selectively pleading or not pleading certain causes of action” (para 72(ii)).

The upshot for practitioners is that a decisive consideration when assessing whether a competition law claim falls within a jurisdiction clause is likely to be whether there are any viable contractual claims which “…would be sufficiently closely related to the tortious claims actually advanced by the Claimant so as to render rational businessmen likely to have intended such a dispute to be decided (like a contractual dispute) by arbitration” (Microsoft at para 72).  The existence and extent of any express or implied contractual obligations to observe competition law therefore looks set for detailed examination in competition law claims in the future.

Leave a comment

Filed under Conflicts, Policy, Procedure