The Office of Fair Trading v Somerfield Stores Ltd & Anor

The issue in this case is whether the Competition Appeal Tribunal (Mr Marcus Smith Q.C.) (the “CAT”) was right to find that there were “exceptional circumstances” justifying an extension of time for appealing a decision of the Appellant, the Office of Fair Trading (the “OFT”).

The Respondents in Case 1197/1/1/12, Somerfield Stores Limited and Co-operative Group Food Limited (which acquired the assets and liabilities of Somerfield Limited) (together the singular “Somerfield”), and the Respondents in Case 1200/1/1/12, Gallaher Group Limited and Gallaher Limited (together the singular “Gallaher”) had each entered into early resolution agreements with the OFT in about July 2008 (respectively the “SERA” and the “GERA”, and together the “ERAs”). The CAT held, in essence, that the Respondents had a legitimate expectation that the OFT would be able to maintain the theory of harm that it had advanced and that had given rise to the ERAs, and that it would be reflected in the OFT’s decision (entitled “Case CE/2596-03: Tobacco” which was issued on 15^th April 2010 – the “Decision”) when it was ultimately made. Accordingly, when other third party addressees of the Decision successfully challenged it, and the OFT had to abandon the theory of harm that underlay it, the Respondents were entitled to initiate their own appeal out of time. It is noteworthy that neither Somerfield nor Gallaher seeks to uphold the reasoning of the CAT as to the finding of a formal “legitimate expectation”.

The term “theory of harm” has been central to this appeal, and it is worth explaining it briefly at the outset. The OFT’s theory of harm is its explanation as to why the agreements that it finds to exist are said to have the anti-competitive objects or effects it alleges are contrary to section 2(1) of the Competition Act 1998 (the “1998 Act”).

The OFT is appealing the CAT’s ruling dated 27^th March 2013 (the “Ruling”) granting an extension of time for the Respondents to appeal the Decision, with permission granted by Lewison LJ on 23^rd July 2013.

The CAT extended the two-month time limit for appealing, which is provided for under rule 8(1) of the Competition Appeal Tribunal Rules 2003 (S.I. No. 1372 of 2003) (the “Rules”), for 28 days from the date of the CAT’s ruling. It did so expressly under rule 8(2) of the Rules, which provides that: “the Tribunal may not extend the time limit provided under paragraph (1) unless it is satisfied that the circumstances are exceptional”. We were referred to the CAT’s case management powers under rule 19 of the Rules, which allow extensions of time whether or not any time limit has expired, but it was not suggested that rule 19 materially affected the operation of rule 8(2) on which this appeal is based.

The OFT contends, in essence, that (a) the finding of “legitimate expectation” was unprincipled and inappropriate, (b) the fact that certain other addressees of the Decision were able to appeal it successfully did not constitute exceptional circumstances, since there must be legal finality and certainty; and (c) the Ruling was irrational, because neither the ERAs nor the subsequent events undermined the Respondents’ ability to bring a timely appeal.

The first and main contention in Gallaher’s Respondent’s Notice is that the OFT misled it. The way in which that allegation has been explained has not been entirely consistent. But Gallaher seems to suggest in particular that it was misled because it was told by the OFT that the Decision included the same theories of harm as had been contained in the GERA, whereas the CAT decided in the appeals against the Decision by third party addressees (“Tobacco I”) that the Decision contained only a different theory of harm from that which Gallaher had been prepared to accept in the ERA. I will explain in a moment the precise way in which the OFT’s theories of harm have varied as time went on. Gallaher also contend that the description of the theory of harm in the ERA was apt to mislead (these arguments are together referred to as the “apt to mislead argument”). Secondly, Gallaher’s Respondent’s Notice alleges that the Ruling should be upheld on the basis that it was entitled to assume, in entering into the GERA, that the OFT would have some proper evidential basis for the theory of harm that would underpin the Decision, whereas it was acknowledged by the OFT in Tobacco I that it did not; in the result Gallaher’s co-operation with the OFT and its abandonment of its legal defence of the infringement claims were procured by false pretences (the “evidential basis argument”).

Somerfield’s Respondent’s Notice raises the further two points based on the contention that the OFT’s conduct during the administrative phase and during Tobacco I amounted to exceptional circumstances. First, it is suggested that Somerfield was entitled to assume that, if the OFT abandoned the theory of harm in the Decision, it would withdraw the Decision and give Somerfield the opportunity to respond to any new theory of harm it pursued (the “withdrawal argument”). Secondly, it is submitted that it is inconsistent with the principle of legal certainty for the OFT publicly to disavow the theory of harm in the Decision, but for Somerfield to remain bound by it in respect of the penalties imposed and for the purpose of civil follow-on proceedings under section 47A of the 1998 Act (the “legal certainty argument”). Gallaher also relies upon this latter legal certainty argument in its Respondent’s Notice.

The relevant factual background is not as complicated as it might at first appear, but it cannot be properly understood without tracking the OFT’s theories of harm as they emerged during the administrative phase of its investigations and during the hearing of Tobacco I.

There were essentially three important theories of harm that need to be carefully distinguished as follows:-

In the OFT’s Statement of Objections of April 2008 (the “Statement of Objections”), which was addressed to two manufacturers including Gallaher (the “Manufacturers”) and 11 retailers including Somerfield (the “Retailers”), and which was issued before the ERAs were entered into and gave rise to them, the OFT stated its theory of harm as being its intended decision that “each Manufacturer was involved in an agreement and/or concerted practice with each Retailer which restricted each Retailer’s ability to determine its retail prices for the Manufacturer’s products and thereby had the object and/or effect of preventing, restricting or distorting competition in the supply of tobacco products” in the UK in breach of Chapter I of the 1998 Act. Mr Jon Turner Q.C., leading counsel for Gallaher, accepted that this theory of harm was sufficiently broad to encompass those comprised in the next two sub-paragraphs. I shall call this theory of harm the “SO/ERA theory”. It is worth noting at this stage that the SO/ERA theory was reflected in the appendix to the ERAs to which I shall refer in due course.

The theory of harm advanced in the Decision (in April 2010) was litigated at great length in Tobacco I. It was common ground that the CAT in Tobacco I found that the OFT’s theory of harm contained in the Decision was encapsulated in paragraph 40 of the OFT’s skeleton as follows:-

As will appear hereafter, during the Tobacco I hearing, the OFT introduced a new theory of harm which it described in paragraph 2 of the written statement that it provided to the CAT on 9^th November 2011 as follows:-

I can now set out the material background facts. In March 2003, the OFT began its investigation into the tobacco market.

On 24^th April 2008, the OFT issued its Statement of Objections, which ran to some 419 pages and 1466 paragraphs.

In about July 2008, each of Gallaher and Somerfield entered into ERAs with the OFT. The terms of each ERA are substantially similar, so we were only referred to the GERA. The GERA first recited the decision that the OFT proposed to make in substantially similar form to that contained in the Statement of Objections, to which I have referred as the “SO/ERA theory”. The GERA then recorded that Gallaher had indicated its willingness to “admit its involvement in relation to all of the infringements that are applicable to it (see the appendix)”. The appendix to the GERA then set out the alleged infringements describing them again in similar terms to the Statement of Objections as “an agreement and/or concerted practice with each of [the other Retailers] that restricted the Retailer’s ability to determine its retail prices for the Manufacturer’s products”. The substance of the GERA then provided by clause 1 that Gallaher would admit its involvement in the infringements on an object and/or effect basis, by clauses 2 and 3 that Gallaher would co-operate with the OFT, by clause 5 that Gallaher would refrain from seeking access to documents on the OFT’s file, other than those referred to directly in the Statement of Objections, and by clause 6.a.i that the OFT agreed to adopt a decision in respect of the infringements which would, as to substance, set out the OFT’s findings of the facts in materially the same form as in the Statement of Objections, subject to any necessary amendments as a result of representations from Gallaher or other addressees. Clause 7 of the GERA provided that if Gallaher brought appeal proceedings before the CAT in respect of the Decision, the OFT could increase the penalty imposed on Gallaher and require it to pay the OFT’s costs of such an appeal. Clause 10 of the GERA gave Gallaher the right to terminate it and to withdraw its admissions.

It is worthy of note at this point that Asda also entered into an ERA with the OFT, but subsequently appealed the Decision (in time), once it was issued.

The Decision was adopted nearly 2 years after the ERAs on 15^th April 2010, and ran to some 583 pages without annexes. It was addressed to the two Manufacturers including Gallaher and to only 10 retailers including Somerfield (hereafter these 10 retailers are referred to as the “Retailers”). The Decision made clear that it had decided not to make infringement findings in relation, amongst other things, to the allegation that “the Infringing Agreements had the likely effect of preventing, restricting or distorting competition” (as opposed to having had that object). Accordingly, the statement of the proposed OFT decision and of infringements in the copies of the GERA and the SERA actually attached to the Decision were redacted so as to remove (amongst other things) the suggestion that the infringements had the likely effect of preventing, restricting or distorting competition.

The Decision explained (particularly in paragraphs 6.213, 6.214 and 6.216) that the OFT had found that the Infringing Agreements restricted competition because, when a parity or fixed differential requirement was implemented, an increase or reduction in the retail price of one brand leads to a corresponding increase or reduction in the retail price of the competing linked brand by an equivalent amount, thus giving rise to a significant degree of certainty that the retail prices of the competing brands will move in parallel and decreasing competition. This was an exposition of the paragraph 40 theory.

Between September and November 2011, the CAT heard appeals in Tobacco I against the Decision by 6 of the addressees of the Decision (to use abbreviated names for the parties, they were Imperial, Asda, the Co-op, Morrisons, Safeway and Shell).

On 3^rd November 2011, which was day 26 of the Tobacco I appeal hearing, matters came to a head. Counsel for Imperial had been repeatedly complaining that counsel for the OFT had not been putting its theory of harm to the witnesses. On day 26, the OFT’s counsel told the CAT that the OFT thought that “each and every one of the specific circumstances relied on in the Decision to support the finding of object infringement “may or may not be established to the appropriate legal standard”, and that the Decision might have been “cast too narrowly”. He suggested that (a) it could invite the CAT to deal with the appeals under paragraph 3(2)(d) and (e) of schedule 8 to the 1998 Act so as to allow the OFT to expand its case as contained in the Decision, or (b) the OFT could amend the Decision by removing the existing infringing agreements and issuing a new Statement of Objections. The OFT did not suggest that the proceedings could continue on the basis of its existing theory of harm (i.e. the paragraph 40 theory). As a result, the CAT adjourned the proceedings directing the OFT to provide a written statement of its revised position.

On 9^th November 2011, the OFT provided the written statement to the CAT in Tobacco I, to which I have already referred, indicating that its revised theory of harm was the paragraph 2 theory. The OFT also invited the CAT to deal with the appeals under paragraph 3(2)(d) and (e) of schedule 8 to the 1998 Act, and suggested (for the first time) that the paragraph 2 theory reflected a part of the original Decision.

On 12^th December 2011, the CAT in Tobacco I quashed the Decision in relation to the 6 appealing addressees. The Decision did, however, remain in force against Somerfield and Gallaher. It held that:-

The Decision had not included findings by the OFT that the paragraph 2 theory was an infringement of the prohibition in Chapter I of the 1998 Act;

Given that the OFT had abandoned its defence of the Decision beyond arguing that the paragraph 2 theory was contained as part of it, that must mean that the Decision should be set aside against the 6 appellant addressees; and

The CAT did not have jurisdiction under schedule 8 to the 1998 Act to continue the appeal hearing on the basis of the paragraph 2 theory, and, even if it had such jurisdiction, it would exercise its discretion against doing so.

It is worth noting also that the Tobacco I decision:-

Did not address any of the substantive issues as to the infringements alleged in the Decision (paragraph 3);

Found that the OFT had conceded that (i) if it wanted to put forward a case outside the paragraph 40 theory, that would require the Decision to be set aside (paragraph 50); and (ii) a restriction on retailer-led changes to retail prices in the absence of a change in wholesale prices was not part of the paragraph 40 theory (paragraph 59);

Did not accept the OFT’s submission that the change to the OFT’s case followed on from cross-examination of the witnesses (paragraphs 80 and 81);

Found that there was no sworn evidence before the CAT in either written or oral form in which any witness said that he or she had entered into or operated any agreement of the kind condemned in the Decision (paragraph 86).

On 13^th July 2012, Somerfield applied for an extension of time to lodge an appeal against the Decision. On 25^th July 2012, Gallaher applied for an extension of time to lodge an appeal against the Decision.

On 27^th March 2013, the CAT delivered the Ruling permitting Somerfield and Gallaher to appeal out of time on the grounds that there were exceptional circumstances. The CAT decided as follows:-

It analysed the law as to extensions of time on the ground of “exceptional circumstances” in paragraphs 38 to 54 of the Ruling in a manner that has not been criticised by any of the parties to this appeal.

It held, following RG Carter Limited v Office of Fair Trading [2011] CAT 25 (“RG Carter”), that the fact that the Decision was later overturned and the Respondents’ subjective reasons for not appealing were not exceptional circumstances justifying an extension of time for the Respondents to appeal. The Decision was not apt to mislead.

It held that the principle of legal certainty did not make it an exceptional circumstance for the Decision to stand against some addressees and be overturned as against others.

It held that the OFT’s unprecedented conduct and eventual defeat in the Decision and alleged breach of its public duties did not constitute exceptional circumstances.

It held that the fact that the OFT had indicated that it did not intend to proceed with any further investigation, and the suggestion that the only proper course was for the OFT to abandon its Decision against all addressees and start again, were also not exceptional circumstances.

It held that it was not possible to say that it was necessary to imply a term into the ERAs to the effect that the OFT would defend the Decision.

It held that the disjunction between the admissions made by the Respondents in the ERAs, and the position ultimately reached whereby neither the OFT nor the CAT has reached any concluded view as to whether or not the paragraph 2 theory forms any credible basis for a finding of infringement against the Respondents meant that the Respondents had a legitimate expectation that the OFT would (i) adopt a Decision which reflected the admissions in the ERAs, (ii) have the wherewithal to make good the factual basis on which the Decision rested, (iii) be able to defend the theory of harm in its Decision (even if not necessarily successfully) on the merits; (iv) not concede that its theory of harm articulated in the Decision could not be sustained generally. The basis on which the Respondents entered into the ERAs had been fundamentally undermined and the Respondents’ decisions not to appeal had been subverted by these matters so that there were exceptional circumstances justifying an extension of time.

On 23^rd July 2013, Lewison LJ granted permission to appeal on the grounds that the CAT correctly set out the legal principles but may have failed to apply them correctly to the facts.

Both sides have referred us to the well-known line of authority to the effect that appeal courts should approach appeals from expert tribunals with an appropriate degree of caution, because it is probable that in understanding and applying the law in their specialist field, the tribunal will have got it right (see Cooke v. Secretary of State for Social Security [2002] 3 All ER 279 per Hale LJ at paragraph 16, and National Grid plc v. Gas and Electricity Markets Authority & others [2010] EWCA Civ 114 per Richards LJ at paragraphs 22-26). I have certainly borne these cautions well in mind in considering the appeal in this case.

The second general factor concerns the relevance of the analogy with applications for extensions of time in the Court of Justice of the European Union (the “CJEU”) under article 45 of Protocol (No. 3) on the Statute of the Court of Justice of the European Union, which requires the party concerned to prove the existence of unforeseeable circumstances or of force majeure. Whilst the domestic jurisdiction of the CAT under Chapter 1 of the 1998 Act largely tracks the jurisdiction of the CJEU in relation to competition within the European Union (see, for example, section 60 of the 1998 Act), the wording of the two relevant procedural provisions are very different. The CAT’s own Guide to Proceedings of October 2005 only says that the “possibilities of obtaining an extension of the time limit for appealing are … extremely limited” before referring to article 45 supra (and incidentally misquoting that provision as referring to “unforeseen” rather than “unforeseeable” circumstances). Ultimately, Mr Daniel Beard QC, leading counsel for the OFT, did not wish to put the matter higher than that the European jurisprudence indicated that the domestic court should interpret its own procedural provision strictly. That seems to me to be a well-founded submission. In my judgment, however, it should be noted that article 45 employs two phrases that specifically direct the court’s attention to the circumstances prevailing at the time when an appeal ought to have been lodged; the term “unforeseeable circumstances” seems to direct primary attention to circumstances that were unforeseeable when the appeal ought to have been lodged, and the term “force majeure” seems to be directed at circumstances that prevented an appeal being lodged at that time. Rule 8(2) of the Rules is framed more generally in terms of “exceptional circumstances”. Those words do not inevitably have any temporal consequence, but the first question that must be answered in seeking to establish exceptional circumstances will, as it seems to me, always be “why did the appellant not lodge an appeal in time?” My conclusion on this second general factor is, therefore, that whilst the court should certainly have regard to CJEU jurisprudence, and whilst it is clear that the words “exceptional circumstances” should be strictly construed, they are apt to apply to any circumstances that are shown to be truly exceptional.

The third general factor relates to the Respondents’ submission that this appeal can only be on a point of law, and the OFT has not challenged any of the legal principles set out in the CAT’s Ruling. This is perfectly true, but, as Lewison LJ said in granting permission to appeal, the OFT’s argument is that the CAT failed properly to apply those legal principles to the facts of this case. If it had made such an error, that too would be an error of law. It is true that the CAT was exercising a discretion, but if the CAT exercised that discretion on the wrong basis, that would also be an error of law.

Somerfield raised a related fourth point, based upon the decision of this court in The London Borough of Newham v. Khatun [2004] EWCA Civ 55, to the effect that, since rule 8(2) of the Rules does not specify the meaning of “exceptional circumstances”, the hurdle facing the OFT is effectively a double irrationality test, since it has to show that both (a) the criteria that the CAT determined as being relevant to the question, and (b) the weighting of those criteria, were irrational. It seems to me that irrationality is not the only ground of appeal in this case. This is not a claim for judicial review as it was in Khatun. As I have said, an appeal may be brought from the CAT on a point of law. Those points are not limited to irrationality. In these circumstances, I do not accept the relevance of the analogy that Somerfield seeks to draw.

With that introduction, I shall now deal with each of the main points raised on the appeal and the Respondents’ Notices in turn.

This point is now, as I have said, not much contested. Neither Respondent suggested that the CAT was justified in referring to the legal doctrine of legitimate expectation. No reliance had been placed by them on that doctrine in argument before the CAT. The doctrine, of course, requires, amongst other things, that the body concerned made a representation which was clear, unambiguous and devoid of relevant qualification, that it was reasonable for the person concerned to rely upon it and that he did indeed rely upon it to his detriment (see, for example, R v. Inland Revenue Commissioners ex parte Unilever plc [1996] STC 681, and R (Bhatt Murphy) v. Independent Assessor [2008] EWCA Civ 755). The only representations relied upon in this case were said impliedly to arise from the ERAs. I shall return to those points in due course, but it seems to me that reliance on the doctrine of legitimate expectation was inappropriate in this case. That does not necessarily mean that the CAT was wrong to refer to the disjunction between the ERAs and the events that followed, but that point did not, in my judgment, entitle the CAT to found itself on any legitimate expectation, properly so called, available to the Respondents.

The disjunction argument is raised again under the next ground of appeal. Accordingly, I propose to deal with it at that stage.

The OFT placed emphasis on the need for finality and legal certainty. It referred us to a series of authorities, many of which were referred to by the CAT, making this point. I do not propose to recite them again here (but see, for example, Prater Limited v. Office of Fair Trading [2006] CAT 11 at paragraph 30, and Fish Holdings Ltd v Office of Fair Trading [2009] CAT 34 at paragraph 21).

The CAT, in this case, drew a distinction between this case and the circumstances of RG Carter. It will be recalled that in that case the OFT’s decision had found that 103 undertakings had entered into arrangements that infringed Chapter I of the 1998 Act. 25 of those 103 undertakings brought appeals on penalty, and 6 of them also appealed on liability. In considering the penalty appeals, the CAT looked at the methodology that the OFT had applied in calculating the penalties it had imposed, and determined that they were excessive because the OFT had incorrectly applied its own guidance. Another addressee first applied to the OFT for a refund of the penalty it had paid, based on the CAT’s decision, and when that was refused, it applied for permission to appeal out of time. The CAT referred at paragraphs 21 and 23 to the fact that the applicant, having made an informed decision not to appeal, regretted that decision in the light of the outcomes of appeals by other addressees. The CAT said that decisions which penalise breaches of the Chapter I prohibition will often involve the consideration of the activities of multiple undertakings over the course of many years, and held that no exceptional circumstances were made out in that case.

The CAT drew attention in the Ruling to a distinction between this case and RG Carter. In this case, the OFT was relying on a series of bilateral agreements that formed a part of a wider network of agreements aimed at controlling the whole sector, whilst in RG Carter, the cases were all individual ones featuring similar legal issues, but were structurally unconnected. The Ruling says that the Respondents could, therefore, have had a legitimate expectation that there was some sustainable factual underpinning to the case against all the industry participants.

The OFT placed considerable reliance in this connection on the CJEU’s decision in AssiDoman Kraft Products v. Commission (Case C-310/97 P) (the “Wood Pulp II” case), where there were 43 parties to a cartel, but only 30 appealed the Commission’s decision. After the 30 had succeeded in obtaining annulment, the other addressees sought to appeal out of time. The CJEU rejected their contention that they had grounds for doing so, saying that it was settled case law that a decision that is unchallenged becomes definitive. It said at paragraph 61 that such a rule is based on “the consideration that the purpose of having time-limits for bringing legal proceedings is to ensure legal certainty by preventing Community measures which produce legal effects from being called in question indefinitely …”.

Of course, we are not bound by RG Carter, and the Wood Pulp II decision is no more than analogous to our situation. But it seems to me that both decisions point the way to the need for finality in competition cases. I do not regard the CAT’s supposed point of distinction from RG Carter as compelling, since in that case too, it could be said that the appellants had been fined on a false basis, just as the Respondents here can say that the basis of the Decision has been undermined.

This last issue, however, requires some further analysis, since the CAT was substantially reliant on its “disjunction” point. The disjunction, it will be recalled, was between the admissions made by the Respondents in the ERAs, and the position ultimately reached whereby neither the OFT nor the CAT has reached any concluded view as to there being any credible basis for a finding of infringement against the Respondents.

The admissions in the ERAs were as to the SO/ERA theory, which, anyway according to Gallaher, was always understood as possibly encompassing both the paragraph 40 theory and what later became the paragraph 2 theory. Gallaher actually filed evidence from its company secretary, Mr Andrew Bingham, to this effect, and also explaining how Gallaher had always told the OFT “why the lockstep theory [the paragraph 40 theory] was not sustainable”. Accordingly, so far as Gallaher was concerned, when it agreed to the ERA, it understood that the SO/ERA theory being advanced by the OFT was very broad, and it signed up on the basis that it was accepting (no doubt for sound commercial reasons including reduced penalties and management disruption etc.) the SO/ERA theory only on the basis that it incorporated something along the lines of what later became the paragraph 2 theory.

Somerfield adduced no evidence of its position at the time. But it seems to me that that circumstance cannot put it in a better position. It accepted the SO/ERA theory for whatever reason, and it must be taken to have understood that that SO/ERA theory was a very broad one. The OFT, of course, later narrowed the SO/ERA theory, when it came to issue the Decision nearly 2 years after the ERAs. It had not promised not to change the theory of harm that it relied upon in the eventual Decision. It had simply agreed, by clause 6.a.i of the ERAs to adopt a decision in respect of the infringements which would, as to substance, set out the OFT’s findings of the facts in materially the same form as in the Statement of Objections, “subject to any amendments deemed necessary as a result of the representations from other recipients of the [Statement of Objections]” (emphasis added). In the result, the OFT received many such representations (to which the Respondents accept they had access), and it decided to narrow its theory of harm in the Decision to the paragraph 40 theory to which I have referred. The OFT was entitled to do this under the terms of the ERAs. The Respondents then had the opportunity to consider the theory of harm advanced in the Decision, to evaluate the redactions to the theory of harm in the copies of the ERAs appended to the Decision, and to decide, no doubt with the benefit of legal advice, whether they wished to appeal.

The remainder of the history does not need to be recited again at this point. Suffice it to say that the CAT relied upon the fact that, in the result, after the vicissitudes of litigation to which the Respondents were not party, the OFT had to concede that it could not make good the paragraph 40 theory that it had relied on in the Decision, and it sought without success to reintroduce the paragraph 2 theory (which had been at least an element of the SO/ERA theory).

I turn then to consider whether the CAT was right to consider that those subsequent events created a disjunction that amounted to exceptional circumstances. What seems to me to have weighed heavily on the mind of the CAT was the fact that the ultimate position was that the OFT had established no theory of harm, so that the infringements accepted at an early stage by the Respondents were effectively left in limbo. I can see why that situation was different from RG Carter, and I understand how, standing back with the benefit of hindsight, one might think such an overall outcome to be unsatisfactory. But, for my part, I do not think that the later events can properly be said to create exceptional circumstances justifying an appeal out of time. The principle of finality and legal certainty is important. It is relatively commonplace for the factual basis of guilty pleas in criminal cases to be later undermined, and yet they are not overturned on that basis (and see DPP v. Shannon [1975] A.C. 717 per Lord Morris at pages 753-4, where the conviction of a defendant following a guilty plea at a first trial was not overturned because his co-conspirator was acquitted of the same charge at a subsequent trial). I do not regard the criminal analogy as conclusive even though penalties imposed under Chapter I of the 1998 Act are quasi-criminal in nature. What seems to me to be most important here is that the Respondents had the fullest opportunity to consider the SO/ERA theory advanced by the OFT at the stage that they signed up to the ERAs, and the same opportunity to consider the paragraph 40 theory advanced in the Decision, before deciding whether to appeal. They could have appealed the Decision, notwithstanding they had signed up to the ERAs. Indeed Asda did so. Of course, there would have been adverse financial consequences as a result of the terms of the ERAs had they done that, but they assumed those obligations voluntarily. It is true that they had agreed not to seek further documentation from the OFT and so to limit their rights of defence, but they did that voluntarily as well. They knew in advance the limitations that they would face in reaching a decision as to whether to appeal the Decision that was ultimately made. And, moreover, they were never promised that the OFT would be able to establish the ERA/SO theory or the theory of harm that would later appear in the Decision. They chose not to appeal with their eyes open.

I will consider whether the GERA and/or the Decision were “apt to mislead” later, but, in my judgment, the CAT was wrong to think that the disjunction between what had been agreed in the ERAs and the later outcome of the Decision and its appeal process automatically provided exceptional circumstances justifying the Respondents in claiming an extension of time. The need for finality in legal process and legal certainty does sometimes lead to unsatisfactory overall outcomes as it did in this case. But I do not think that future events arising from third party appeals against the same Decision will normally form a proper basis for a plea of exceptional circumstances justifying an extension of time for those that sat back and decided not to appeal in due time.

I have considered particularly whether it is appropriate to overturn the CAT’s reasoning in paragraphs 92 and 93 of its Ruling, bearing in mind its specialist knowledge and understanding. But it seems to me that it fell into error for the reasons I have tried to explain. It started by employing incorrectly the concept of legitimate expectation, and then placed too much emphasis on the ultimate unfortunate, even messy, outcome, rather than concentrating on whether there was any justification, bearing in mind the need for legal certainty, for the Respondents having failed to appeal in time, knowing what they knew at the time.

I have already largely dealt with this point. I am far from sure that the Ruling can properly be described as irrational, but I do think, as I have said, that the CAT failed adequately to focus on the reasons why an appeal was not brought within time. I asked counsel for Gallaher this question and he gave me four answers as follows: (1) that in a quasi-criminal case, it was for the OFT to prove its case; (2) that Gallaher had not admitted in the GERA all the matters in the Statement of Objections; (3) that Gallaher agreed in the GERA not to seek further information, so was not in as good a position as other addressees of the Decision; and (4) that Gallaher was in an invidious situation, having disabled itself from exercising rights of defence that the other parties enjoyed. He relied on the comments made by the CAT chaired by Lord Carlile QC in Tesco Stores Limited v. Office of Fair Trading [2012] CAT 31 at paragraph 110 as to the brief and formulaic nature of ERAs and the commercial factors that might lead a party to admit liability for infringement in respect of an allegation that might turn out to be either untrue or not properly investigated.

As it seems to me, however, these answers do not address the real point. It is true that the OFT has the role of a prosecutor and has wide powers to impose penalties, and that those powers must be exercised on a proper basis, but that does not stop commercial parties from taking a commercial view as to whether or not to sign up to an ERA after a long investigatory process and the publication of a lengthy Statement of Objections. The addressee knows precisely the terms that are being offered. It knows what it has done in relation to the alleged infringements, and what it is being asked to admit, and the terms requiring its co-operation and the fetters on its rights of defence to which it is being asked to agree. It can take it or leave it. Having taken it, the only thing that the OFT had to do in this case was to promulgate a decision with the amendments that it regarded as necessary having considered the submissions it received from all the relevant parties. The OFT never guaranteed or promised, as I have said, that it would establish or prove the SO/ERA theory that the Respondents admitted to in the ERAs. The Respondents acted voluntarily in making those admissions. There was no coercion or obligation on them to do so.

Whilst I do not, as I say, think the CAT’s decision was irrational, in my judgment neither the ERAs nor the subsequent events undermined the Respondents’ ability to bring a timely appeal. This ought, I think, to have been a central feature of the CAT’s decision as to exceptional circumstances.

Before turning to deal with the arguments raised by the Respondent’s Notices, I should state my conclusions on the OFT’s appeal. In my judgment, for the reasons I have given, the CAT failed properly to apply the law that it correctly stated as to exceptional circumstances under rule 8(2) of the Rules to the facts of this case. The CAT’s reasoning in paragraphs 92 and 93 was flawed in 4 main respects: (1) the doctrine of legitimate expectation was not applicable to the facts of this case; (2) there was no disjunction between the admissions made by the Respondents in the ERAs and the ultimate position, since the principle of finality and legal certainty required the Respondents to be allowed to admit the alleged infringements on one basis, even if that basis changed or was ultimately not established in litigation with other parties in the same case; (3) the fact that third parties successfully appeal a decision is no reason for holding that there are exceptional circumstances justifying a non-appealing addressee being granted an extension of time; and (4) there were in any event no good reasons for the Respondents having failed to appeal the Decision in time.

Subject then to the arguments in the Respondent’s Notices, I would allow the OFT’s appeal.

This argument is, as I have said, to the effect that Gallaher was misled by the theory of harm in the GERA and because it was told by the OFT that the Decision included the same theories of harm as had been contained in the GERA, whereas the CAT decided in Tobacco I that the Decision contained only a different theory of harm from that which Gallaher had been prepared to accept in the ERA. The CAT rejected the argument as to the Decision itself being misleading relying on the decision in RG Carter and the fact that an addressee can decide whether to appeal in its own way for its own reasons; the lack of clarity of the Decision and its potential defences to the theories of harm advanced are (amongst others) all factors it will and can weigh in the balance. I agree. Indeed, this is very much the same reasoning as I have advanced above.

In short, I do not accept that Gallaher was misled in any operative sense, whether by the GERA or by the Decision. Whilst I echo the CAT’s caution about taking into account privileged material as to a party’s decision-making, Gallaher quite clearly understood that the SO/ERA theory encompassed both what became the paragraph 2 theory and the paragraph 40 theory. It decided to sign up to the ERA on the basis that it accepted an infringement based on the paragraph 2 theory. It was in no sense misled at that stage in so doing. When it saw the Decision, it had the same opportunity as anyone else to evaluate it. The OFT did not mislead it by continuing to attach to the Decision the redacted GERA referring to object infringements alone, but retaining the SO/ERA theory. It was up to Gallaher to decide with its legal advisers if it thought the SO/ERA theory had disappeared from the Decision, and then, if it wished, to appeal. As it was, it took many lawyers many days finally to establish that only the paragraph 40 theory survived into the Decision. But that is just how complex litigation sometimes turns out. It was not the result of the OFT misleading Gallaher.

This argument suggests that Gallaher was entitled to assume, in entering into the GERA, that the OFT would have some proper evidential basis for the theory of harm that would underpin the Decision. This point formed a component of the CAT’s reasoning in paragraph 93 of the Ruling (see paragraph 93(2) and 93(6) in particular). It is true, of course, that OFT is the prosecutor and must, if no admissions are made, prove its case. It is true also that the Chapter I process requires a decision to be reached and the GERA envisages that this will happen. But all that is, as it seems to me, a long way away from implying a term into the GERA to the effect that the OFT will have “some proper evidential basis” for the theory of harm in the Decision. In my judgment, such a term is too uncertain to be implied. It is neither necessary nor appropriate to do so. To put the matter bluntly, the Respondents are grown-up commercial parties. They knew what evidence was relied upon in the Statement of Objections. They knew what evidence was available to them as to their own infringements. They could evaluate both when they concluded the ERAs. It would be quite impossible for the OFT to conduct such an investigation and bring it to a timely conclusion if it were to be taken as representing at the time of an early resolution agreement that it would in the future have a “proper evidential basis” for its decision. Of course, it would be expected to have such a basis, but litigation sometimes proves otherwise. In any event, it may be asked rhetorically: how could it be decided whether the OFT had such an evidential basis? No doubt it thought it did, even though it turned out it did not.

In this latter regard, of course, Gallaher relied strongly on the CAT’s finding in Tobacco I that there was no sworn evidence before it in either written or oral form in which any witness said that he or she had entered into or operated any agreement of the kind condemned in the Decision. But this was the outcome of hard fought litigation. The key point here is that it was up to Gallaher to evaluate at the time of the ERA whether there were grounds for it to admit the infringements alleged. It was not obliged to do so. I cannot see that the OFT was making any representation or agreement as to its future conduct beyond the terms of the GERA itself.

This is Somerfield’s primary argument on its Respondent’s Notice. It contends, as I have said, that Somerfield was entitled to assume that, if the OFT abandoned the theory of harm in the Decision, it would withdraw the Decision and give Somerfield the opportunity to respond to any new theory of harm it pursued.

Mr Rhodri Thompson Q.C., leading counsel for Somerfield, placed reliance on two important authorities to support his argument. In Napp Pharmaceutical Holding Limited v. Director General of Fair Trading (2002) (Case No. 1001/1/1/01), the CAT (chaired by Sir Christopher Bellamy) said at paragraph 133 that the Director should not be permitted to advance a wholly new case at the judicial stage, and that if he wished to do so, his proper course was to withdraw the decision and adopt a new decision. Likewise in Mastercard UK Members Forum Limited v. Office of Fair Trading [2006] CAT 14, the CAT (again chaired by Sir Christopher Bellamy) at paragraph 20 endorsed the approach in Napp.

Mr Thompson argued that the failure of the OFT to withdraw the Decision left Somerfield in a legal limbo that it could never have anticipated. There is, he submits, therefore an important policy reason why a later appeal should be permitted.

The first answer to this submission is that Somerfield was not left in any limbo, legal or otherwise. The Tobacco I decision only quashed the Decision as regards the appellants before the CAT, not generally. Thus the Decision stands as against Somerfield. That, of course, does not answer the question as to whether the OFT ought to have withdrawn the Decision having accepted that the paragraph 40 theory that underlay it could not be made good. The Napp and Mastercard decisions seem to me only to have gone as far as saying that, if the OFT wants to introduce a new case, it should withdraw the earlier decision. In this case, the OFT has decided, for whatever reasons, not to do so. It does not intend now to pursue the successful appellants in Tobacco I. In other words, those appellants have been successful in their appeal and in resisting the OFT’s attempts to prove infringements against them. But Somerfield has not been similarly successful because it decided not to appeal the Decision in time. Instead, it decided at the time of the ERA, (I say again) no doubt for good commercial and other reasons, to admit the infringements alleged and pay a (reduced) fine. At the time of the Decision, it decided not to question the paragraph 40 theory alleged in the Decision, nor to contend that it differed from the SO/ERA theory. The result, in short, is no different from what it would have been if the Decision had been challenged by some, but not all of the addressees, and set aside by the CAT after deciding the appeals on the merits. I cannot see why either situation gives the non-appealing addressee exceptional circumstances justifying it appealing out of time, maybe years later.

In conclusion on this point, therefore, in the circumstances of this case, the OFT was under no obligation to withdraw the Decision that underlay the ERAs, unless it wished to advance a wholly new case, which it did not. Legal certainty demands that the somewhat uncomfortable final outcome in this case will sometimes be inevitable, because these are multi-party investigations. Each party can decide for itself whether to settle or appeal or (in this case) both. They will be bound by their informed decisions, but the outcome of the litigation will sometimes show that they made an unwise call.

Finally, Somerfield (this time supported by Gallaher) sought to invoke the principle of legal certainty to argue that it was contravened if the OFT publicly disavowed the theory of harm in the Decision, whilst Somerfield remained bound by it in respect of the penalties imposed and for the purpose of civil follow-on proceedings under section 47A of the 1998 Act.

Section 47A(9) of the 1998 Act provides that, in determining a civil claim for damages arising from an OFT decision under Chapter I of the 1998 Act, the CAT is bound by any such decision. Somerfield, therefore, submits that it will, in effect, be bound in law by the paragraph 40 theory, which has been abandoned by the OFT. It is in the public interest for the OFT to be held to its election and for the Respondents to be allowed to appeal.

In my judgment, this argument misses the point. The SO and the ERA were based on the SO/ERA theory. It was that theory which Somerfield accepted in the SERA and Gallaher accepted in the GERA. The SO/ERA theory included what later became the paragraph 2 theory as well as the paragraph 40 theory. The OFT would have liked to continue to advance the paragraph 2 theory, but has decided, no doubt for pragmatic reasons, not to do so. But it cannot be said that there is any inconsistency in the effect of the ERA admissions on follow-on claims. Somerfield and Gallaher admitted a broader theory of harm. Their penalties and any follow-on claims will be based on one constituent of that theory. Somerfield adduced no evidence as to what element of the SO/ERA theory it was accepting, but even if it had (as Gallaher did) that cannot affect the legal consequences of it having made the broader admissions it did.

In my judgment, the theory of legal certainty does not advance this point. Somerfield and Gallaher made broad admissions as to the ERA/SO theory and are bound by them. They each had a proper opportunity to appeal and chose not to. The fact that litigation resulted in a bad outcome for the OFT’s paragraph 40 theory that underlay the Decision cannot provide exceptional circumstances justifying an appeal long out of time.

For the reasons I have sought to give, I would allow OFT’s appeal, refuse an extension of time for appealing to both Gallaher and Somerfield, and dismiss both Respondent’s Notices.

I agree.

I also agree.