England and Wales Court of Appeal (Civil Division) Decisions

IN THE COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE COMPETITION APPEAL TRIBUNAL
Marcus Smith QC, Brian Landers and Professor Colin Mayer
[2012] CAT 11

		Royal Courts of Justice Strand, London, WC2A 2LL
		06/03/2013

Lord Justice Moses :

1. Mobile and fixed communications providers connect their customers with recipients on different mobile networks. The service provided by the intended recipient's mobile communications provider to the originating communications provider which is necessary for that to take place is called 'wholesale mobile voice call termination'. For this service, operators impose a wholesale charge known as 'mobile termination rates' (MTR). These have been subject to regulatory control since 1999 on the basis that absent control the significant market power exercised by operators would be detrimental. Price control is exercised through "significant market power" (SMP) conditions pursuant to what is now the Communications Act 2003 (the 2003 Act).
2. Ofcom is under a statutory obligation to consider the effects on efficiency and competition of any regulatory control of price and to confer the greatest possible benefits on end-users of public electronic communication services. It had previously, in March 2007, set a price control based on what is known as LRIC plus (long-run incremental cost plus). From May 2009, Ofcom conducted a market review and three consultations. By the time it published its Statement in March 2011, the choice for Ofcom was between pure LRIC and LRIC plus. In that Statement it regulated the charges for mobile call termination of the four national providers on the basis of pure LRIC, a change from the previous basis of LRIC plus. The precise meaning of pure LRIC and LRIC plus is not of relevance to this appeal. Under LRIC plus operators could recover a proportion of fixed costs common to services other than mobile voice call termination. Pure LRIC was regarded by Ofcom as a better approximation of marginal costs and would result in mobile termination rates at less than half of the rates calculated on the basis of LRIC plus. It also had the merit of following a Recommendation of the European Commission on the Regulatory Treatment of Fixed and Mobile Termination Rates in the EU (2009/396/EC) dated 7 May 2009 (issued pursuant to Article 19 of Directive 2002/21/EC).
3. It was predicted that operators would receive about £200m less in revenue from their wholesale termination charges in the final year of the charge control, following a change from LRIC plus to pure LRIC. Pushing down prices in respect of termination charges would lead to price rises elsewhere (economists adopt the vivid image of the waterbed). Operators would seek to make good their losses by charging their retail subscribers higher prices. Ofcom thought that the retail price rises would principally be focussed on post-pay customers.
4. The charge controls are imposed for a limited period. In March 2011, the charge control, like previous controls, was set for four years; in future it will be set for three years (s.84A(3) and (7) of the 2003 Act). Ofcom will start work on the next market review in September 2013.
5. Three of the four national mobile communications providers and British Telecommunications plc (BT) appealed various conclusions in the Statement to the Competition Appeal Tribunal. BT and Hutchison 3G (UK) Limited (Three) supported the change to pure LRIC, but that was challenged by Everything Everywhere Limited (EE) and Vodafone Limited, both of whom sought the maintenance of a control based on LRIC plus.
6. The statutory appeal regime provided by the 2003 Act required price control matters, such as the control of call termination charges, to be referred to the Competition Commission. The Tribunal referred seven questions to the Commission; only the first is relevant:

"Whether the charge controls imposed…[in the Annex to the Decision]…have been set at levels which are inappropriate because Ofcom erred in adopting the pure LRIC cost standard, rather than the LRIC+ cost standard, as the basis for the charge controls (for the reasons set out in…EE's Notice of Appeal (Ground 1) and of Vodafone's Notice of Appeal."

In short, the referred question which the Competition Commission had to determine on the merits and on the basis of the appellants' grounds of appeal was, whether Ofcom was wrong to impose the pure LRIC and not the LRIC plus cost standard.

7. The Commission agreed with the contention of Vodafone and of EE that Ofcom was wrong to identify post-pay customers as those most likely to bear the burden of a price rise attributable to the change from LRIC plus to LRIC. On the contrary, it predicted that pre-pay customers would bear the burden of price rises. The Commission then considered the likely effects of that burden. It concluded that those effects would not be of such a nature or extent as to cause them to differ from the conclusion of Ofcom that the cost standard should be changed from LRIC plus to LRIC.
8. EE and Vodafone pursued the only statutory means available to challenge the Competition Commission's determination; they contended before the Competition Appeal Tribunal that the determination would fall to be set aside on judicial review principles. The Tribunal rejected those contentions and, accordingly, was bound to uphold the Commission's conclusions. EE now appeals against that rejection, facing opposition from the four respondents identified in the title to these proceedings.
9. Before the Tribunal and before this court, EE claimed to have identified a judicially reviewable error in the Commission's determination. EE submitted that, in the light of the Commission's disagreement with Ofcom, it was necessary to consider the consequences of that disagreement on Ofcom's decision to choose pure LRIC rather than LRIC plus. Ofcom's choice had, after all, been, at least in part, on the basis that price rises would affect post-pay customers. The Commission itself observed that it lacked important survey evidence as to customer responses. Such evidence, so it was submitted, was necessary in order to reach a conclusion as to the choice which would best achieve the statutory objectives of the Communications Act 2003.
10. In the absence of such evidence, EE contended that the Commission should have acknowledged that it was in no position to make an appropriate choice between LRIC and LRIC plus; the matter should have been referred back to Ofcom to obtain the evidence necessary to reach a conclusion consistent with the statutory aims. The Commission's judicially reviewable error, so EE contended, was in misdirecting itself that it was bound to choose between LRIC plus and pure LRIC despite its own recognition that there was an absence of important evidence on a key issue. That misdirection in law should have been corrected by the Tribunal. The Tribunal's failure to remedy that error by remitting the matter to Ofcom, with directions to make good the absence of that evidence, founds EE's main ground of appeal. Two others are of lesser significance: EE alleges unfairness and inconsistency.

The Statutory Scheme

11. A full and unchallenged description of the statutory framework and the statutory appeal system is set out in the Tribunal's judgment, paragraphs 35-59. It would only undermine its comprehensive force were I to repeat it. I merely underline certain features to explain this judgment.
12. The Communications Act 2003 is designed to implement the group of EU Directives known as the Common Regulatory Framework. Article 13 of the Access Directive (2002/19/EC) requires a national regulatory authority such as Ofcom to promote efficiency, sustainable competition and to "maximise consumer benefits" when it imposes price controls. Ofcom is required to promote competition by ensuring that users, including disabled users, "derive maximum benefit in terms of choice, price and quality". It is required to address the needs of specific social groups (Article 8(4) of the Access Directive; Article 8 of the Framework Directive). I have already referred to the power conferred on the European Commission to make Recommendations under Article 19 of the Framework Directive. Member States are required to ensure that their regulatory authorities take "the utmost account" of such a Recommendation and to give a reasoned explanation for not following a Recommendation to the European Commission.
13. When Ofcom sets binding conditions, such as the price control mechanisms in this case, on those with significant market power in a specific market, the conditions must be objectively justifiable and proportionate (s.47(2) of the 2003 Act). Price control set as a significant market power condition must be appropriate for the purposes of :

a) promoting efficiency;

b) promoting sustainable competition; and

c) conferring the greatest possible benefits on the end-users of public electronic communications services (s.88(1)(b) of the 2003 Act).

It is on this section that EE places particular reliance. Absent what the Commission described as important evidence on a key issue, EE contends that its endorsement of Ofcom's choice of pure LRIC ran the real risk of imposing a price control standard which failed to achieve the statutory objective of conferring the "greatest possible benefits" on end-users.

14. The statutory objectives underline the important public interest in setting an appropriate cost standard. That is of no lesser importance when that standard is set following the statutory appeal process. The Framework Directive (2002/21/EC) requires Member States to put in place an independent appeal body with appropriate expertise to take into account the merits which are subject to review by a court or tribunal (Art. 4, implemented in sections 192-196 of the 2003 Act).
15. The Competition Appeal Tribunal is required to decide an appeal on its merits and by reference to the grounds of appeal set out in the notice of appeal (section 195(2)). But the questions relating to the change from LRIC plus to LRIC were price control matters. They had, therefore, to be referred to the Competition Commission for a determination on the merits of the appeal (s.193(2)). The Commission was required to determine the merits within the period of four months specified by the Competition Appeal Tribunal (Amendment and Communications Act Appeals) Rules 2004 (S.I.2004/2068), unless the Tribunal had directed otherwise (Rule 5) and to notify the Tribunal as soon as practicable (section 193(5)). The Tribunal was required to follow the Competition Commission's decision on the appeal unless it concluded that that determination would fall to be set aside on a judicial review application (ss.193(6) and (7) of the 2003 Act).
16. Although the determination of the appeal on the merits of the specified price control issues was for the Commission alone, the disposal of the appeal was a matter for the Tribunal. The Tribunal decides the appeal "on the merits", even where it is bound to follow the determination of the Commission (s.193(5) and s.195(2)). The Tribunal is required to decide what (if any) appropriate action should be taken by Ofcom (s.195(3)). Once notified of the Commission's determination the Tribunal is required to remit the decision under appeal to Ofcom, whether the appeal succeeds or not, and to give directions, if any are appropriate (s.195(4)).

Was the Commission Bound to Choose between LRIC and LRIC plus?

17. Two paragraphs of the Commission's determination encapsulate the origin of primary ground of EE's appeal:

"Overall assessment on customer responses

2.700 We accept that care must be taken when assessing survey results. We do not accept that a well-designed survey provides no relevant information. Since the question of consumer responses to price increases is a key issue in this determination, we would normally expect a robust survey to be important evidence that a regulator would seek to rely on. In this case, there does not appear to be any reliable survey evidence that directly addresses the magnitude of customer loss that would flow from the type of price changes we expect to observe. Vodafone and EE's surveys tell us something about the relative effects of different types of price changes, and about the relative impact on low-income customers compared with other customers, although we have been careful in how much weight to place on them.

2.701 The evidence that Ofcom has relied on, primarily about customers' attitudes to mobile phones, is of limited use. Consumers as a whole may have inelastic demand, but that does not mean that there will not be a significant reduction in number of users, especially if price increases are directed towards those with a lower willingness to pay or those who are more price sensitive." (my emphasis)

18. In the absence of a robust survey on a key issue, contends EE, the Commission was not legally bound to make a choice and should not have done so. The Commission had reached the conclusion that defects in Ofcom's decision had left it without sufficient evidence or material to make a reliable choice consistent with the statutory objectives. In its erroneous belief that it was bound to do so, the Commission risked imposing a price control measure which would be inconsistent with the statutory objectives, on the basis of inadequate and unreliable evidence. This, submits EE, was a judicially reviewable misdirection in law. The respondents challenge the underlying legal premise: the Commission was bound to make a choice.
19. The Competition Commission directed itself, in three separate passages earlier in its determination, as follows:

"1.28 Vodafone cited TalkTalk Telecom Group plc v Ofcom…. as authority for the proposition that the Tribunal (and, hence, the CC) should proceed on the basis that an appeal must succeed if it showed that Ofcom reached the wrong decision or that, in reaching its decision, it applied a methodology which was so unsound as to create a real risk that the decision was wrong. We agree that the proposition largely accords with the matters considered in the Tribunal's judgment in the Wholesale Broadband Access non-price-control judgment. We note, however, that the 'methodology' to which Vodafone refers is in fact the point of process taken in appeal by TalkTalk, rather than the analytical methodology which is subject to scrutiny in the context of these MCT Appeals."

"1.33 We have carried out our examination, in respect of Reference Questions 1 to 6, with the purpose of determining whether Ofcom erred for any of the specific reasons put forward by the parties. In determining whether it did so err, we have not held Ofcom to be wrong simply because we considered there to be some mistake in its reasoning on a particular point—the error in reasoning must have been of sufficient importance to vitiate Ofcom's decision on the point in whole or in part. This is the standard set out in paragraph 1.32 of the MCT Determination and we believe it to be the appropriate approach to the matters at issue in these MCT Appeals."

"2.59 The CC agrees that it must determine whether Ofcom made the 'right' choice and that the appeal should succeed if the appellant can demonstrate that Ofcom applied a methodology which was so unsound as to create a real risk that the decision was wrong (as is the case in our answer to Reference Question 6). In determining whether Ofcom's judgement was wrong in relation to the choice of cost standard the CC takes into account all the factors that Ofcom had to weigh. We do not think that it automatically follows that mistakes in findings of fact that are designed to inform this judgement can render the judgement 'wrong' in the round. The question is whether any mistake is of sufficient importance to vitiate Ofcom's decision as to the appropriate cost standard."

20. I read those passages as containing an acknowledgment that there may be cases where the basis of Ofcom's decision has been so undermined on appeal that the decision cannot stand. If the Commission is then left in a position where it cannot make a determination as to which measure is most likely to fulfil the statutory objectives, then it appears to accept that it is free to say so.
21. EE did not suggest that the Commission misdirected itself in those passages. But the Tribunal doubted whether the Commission's formulation in [1.28] was correct. It endorsed its previous dicta in TalkTalk Telecom Group plc v Office of Communications [2012] CAT 1 (over which the Chairman had also presided):

"Where a decision can be challenged by way of a merits appeal, it is incumbent upon an appellant to show – if necessary by way of new evidence – that the original decision was wrong "on the merits". It is not enough to suggest that, were more known, the Tribunal's decision might be different." ([134]of TalkTalk, cited [220])

22. It is beyond question that this is correct, so far as it goes. If the appellant can do no more than show that there is a "real risk that the decision was wrong" then it has not shown that Ofcom's decision was wrong and the appeal should be dismissed. But there remains scope for dispute as to what is meant by showing that an original decision is wrong "on the merits".
23. It is for an appellant to establish that Ofcom's decision was wrong on one or more of the grounds specified in s.192(6) of the 2003 Act: that the decision was based on an error of fact, or law, or both, or an erroneous exercise of discretion. It is for the appellant to marshal and adduce all the evidence and material on which it relies to show that Ofcom's original decision was wrong. Where, as in this case, the appellant contends that Ofcom ought to have adopted an alternative price control measure, then it is for that appellant to deploy all the evidence and material it considers will support that alternative.
24. The appeal is against the decision, not the reasons for the decision. It is not enough to identify some error in reasoning; the appeal can only succeed if the decision cannot stand in the light of that error. If it is to succeed, the appellant must vault two hurdles: first, it must demonstrate that the facts, reasoning or value judgments on which the ultimate decision is based are wrong, and second, it must show that its proposed alternative price control measure should be adopted by the Commission. If the Commission (or Tribunal in a matter unrelated to price control) concludes that the original decision can be supported on a basis other than that on which Ofcom relied, then the appellant will not have shown that the original decision is wrong and will fail.
25. Usually an appellant will succeed by demonstrating the flaws in the original decision and the merits of an alternative solution. But that is not necessarily so. I would not rule out the possibility that there could be a case where an appellant succeeds in so undermining the foundations of a decision that it cannot stand, without establishing what the alternative should be. In such a case, if there is no other basis for maintaining the decision, the Commission or Tribunal would be at liberty to conclude that the original decision was wrong but that it could not say what decision should be substituted. The Tribunal would then be required to allow the appeal under s.195(2) and direct Ofcom to make a fresh decision with such directions as the Tribunal thinks are necessary to reach a properly informed conclusion. The Tribunal may wish to specify the steps to be taken by Ofcom to make good any deficit in evidence and material so as to reach a fresh decision, or leave it to Ofcom to act as it sees fit in the light of the Commission's conclusion.
26. I would expect such an outcome to be rare, all the more rare since the Commission's own Guidance (CC 13 April 2011) envisages that an error may be identified but:

"Establishing that Ofcom erred might not of itself establish what Ofcom should have done." [3.27]

In such cases the Commission's Guidance provides for separate views to be invited on key issues at a remedies conference [3.28-3.34]. In the instant case the Commission saw no necessity for such a process in relation to EE's and Vodafone's appeals against the choice of pure LRIC.

27. The Commission's reference to unsound methodology (in [1.33] and [2.59]) harks back to an argument which failed in TalkTalk that the procedural deficiencies in Ofcom's decision were so great that it was unsafe for the Tribunal to conclude that Ofcom's decision was correct. The Tribunal commented:

"It may be that there are cases where Ofcom's approach in reaching its decision was so defective as to preclude the tribunal from reaching an "on the merits" conclusion" (TalkTalk [131] cited by the Tribunal at [219]).

28. It seems to me that, in that acknowledgment that a tribunal might not be able to identify the correct solution, there is a false dichotomy. A conclusion that an approach is so defective, whether as to substance or procedure, that the original decision cannot stand, where there is no alternative basis on which it can be sustained is a conclusion "on the merits". In such circumstances there is no obligation on the Commission or the Tribunal to substitute another decision.
29. From time to time in its comprehensive judgment of 139 pages the Tribunal seems to suggest that the Commission was obliged to determine which of the two price control standards was correct in order to comply with its obligation to answer the referred question (quoted in [6], see Tribunal [208] and [209]). It may be that it was this point which prompted it to grant leave to appeal.
30. The Tribunal relied upon Re B [2008] UKHL [2009] 1 AC 11 for the proposition that a judge, and as the Tribunal put it "an appeal body", must make up its mind on the evidence before it. But, the Tribunal acknowledged, Re B was concerned with a different area of law. Baroness Hale said that a judge should make his mind up on the evidence, if necessary by reliance on inherent probabilities and, in a rare case, the burden of proof ([30]-[32] of Re B, cited Tribunal [210]). But this was a response to the failure of the judge in that case to decide, on the balance of probabilities, whether an allegation of sexual abuse was true or not. In my view, the issues in that case were so far removed from that which Ofcom and the Commission had to resolve that it affords little, if any, assistance. Certainly, it is not authority for the proposition that in every case the Commission is bound to decide the appropriate price control measure, even if it concludes that there is insufficient evidence to determine what that measure should be.
31. The statutory scheme does not impose an obligation on the Commission to reach a conclusion as to the appropriate measure of price control, in circumstances where it thinks the original decision is in error but is unable to fix an alternative solution. Section 193(2) requires the Competition Commission to determine the referred "price control matter". "Price control matter" is defined in broad terms in s.195(10):

"(It)… means a matter relating to the imposition of any form of price control by an SMP condition the setting of which is authorised by ---

(a) section 87(9)."

These words are sufficiently wide to embrace a determination that Ofcom's decision cannot be maintained. It does not in every case require the Commission to reach an alternative solution.

32. The referred Question 1 assumed that the Commission would be able to say whether LRIC or LRIC plus was the correct cost standard. That was the consequence of the way the issue was framed in the grounds advanced by the appellants, EE and Vodafone. But non constat that the statutory scheme requires the Commission to do so. Disposal of the appeal without substituting an alternative decision is not unknown, although, I repeat, I would expect it to be rare. In Vodafone Limited & Others v Office of Communications [2008] CAT 22, Ofcom's decision as to a change to the General Conditions in relation to customers' ability to keep their numbers when changing network operators was appealed to the Tribunal. The Tribunal agreed that Vodafone had demonstrated that Ofcom's cost benefit analysis was not sufficiently reliable or sound [47] and [126]. The Tribunal was not in a position to form a view on the basis of a more soundly based cost benefit analysis (see e.g. [127] and [149]). In those circumstances it allowed the appeal and remitted the whole matter to Ofcom for reconsideration [156] and [159].
33. For those reasons, I accept the legal premise on which EE's main ground of appeal is based. There was no obligation to choose between LRIC plus and LRIC, if the Commission concluded that there was no basis for Ofcom's choice of LRIC and it concluded that there was no or no sufficient basis for choosing the alternative LRIC plus.
34. The real question, however, on this appeal is whether that was the conclusion of the Commission. The appellant faces an almost insuperable task if it can do no more than show that the Commission ought to have taken that view. It could only succeed before the Tribunal and it can only succeed before this court if it can identify a judicially reviewable error on the basis of which the Commission's decision should be set aside. EE cannot successfully impugn the Commission's assessment of the adequacies of the evidence and material before it, unless that assessment was outwith the range of reasonable conclusions. It is no exaggeration to describe that task as almost insuperable where that assessment is made by an expert tribunal in relation to matters of prediction and economic judgment. Recognising this difficulty, EE sought to argue that the Commission had itself assessed the evidence as insufficient.

The Need for Caution

35. Certain features should infuse consideration of the instant appeal. The subject matter of the appeal is a complex question of economic judgment. It involves questions of policy in a highly technical field. The regulator, Ofcom, and the Competition Commission are required to make educated predictions for the future as to the effect of any price control measure to be imposed. Although decisions relating to the control of charges are of great importance to communication providers and to the general public, the exercise of seeking an appropriate solution is necessarily imprecise; when looking to the future, there is unlikely to be any one right answer.
36. EE, like the other mobile communications providers, has had every opportunity to present its case for LRIC plus and to oppose pure LRIC. Its predecessor companies, Orange and T-Mobile, made submissions to the European Commission between June and September 2008 before the Recommendation was made. They were engaged in the two rounds of consultation conducted by Ofcom in May 2009 and April 2010 and in two further consultations on specific issues in November 2010.
37. This attempt by EE to identify a judicially reviewable error comes after independent consideration by two expert bodies of the merits in this complex area of price control and after a comprehensive judgment by the Competition Appeal Tribunal, itself another expert tribunal, concerned only with the question whether such an error can be identified.
38. If that context and all those factors are given their due weight, the caution which Baroness Hale and others have urged should be taken in any appeal from or judicial review of the decisions of expert tribunals, or in granting permission to appeal, needs to be underlined and underlined in red. Where a tribunal has particular expertise in a highly specialised area of law it is "quite probable" that that tribunal will have got it right. A reviewing court, amongst which can be counted the expert Competition Appeal Tribunal, should not be astute to find an error of law (Cooke v Secretary of State for Social Security [2002] 3 All ER 279, paras 15-17, AH (Sudan) v Secretary of State for the Home Department [2008] 1 AC 678 para 30 and R (Cart) v Upper Tribunal(SC(E)) [2012] 1 AC 663 paras 49-50).
39. The Competition Commission is the second expert tribunal to consider the merits but it is an appeal body and not a duplicate regulatory authority. Article 4 of the Framework Directive calls for:

"…an appeal body and no more, a body which can look into whether the regulator has got something materially wrong. That may be very difficult if all that is impugned is an overall value judgment based upon competing commercial considerations in the context of a public policy decision." (T-Mobile (UK) Limited v Office of Communications [2008] EWCA Civ 1371 [31] per Jacob LJ)

The Commission's Determination

40. EE's crucial task is to establish that the Commission did regard the evidence as insufficient to sustain a choice consistent with the statutory objectives but made a choice because it erroneously thought that it was bound to do so. This attempt has an unpromising start. It would be surprising if the Commission failed to follow the approach it set out in the passages I have already quoted, in which there is no suggestion that it regarded itself as compelled to choose LRIC or LRIC plus. EE does not quarrel with those directions. EE is forced to rely upon an inference to be drawn from the lengthy and detailed determination. This poses a forensic difficulty for even as skilled an advocate as Mr Turner QC, a difficulty which is common in judicial review. If he picks out passages which he describes as key, the court may accuse him of failing to read the determination as a whole. If he cites passages at greater length he risks blunting his point with wearisome quotation.
41. The Commission, like Ofcom, assessed the choice of cost standard against four criteria: economic efficiency (allocative and dynamic), competitive impacts, distributional effects on vulnerable consumers and commercial and regulatory consequences. The passages on which EE focussed relate to allocative efficiency, (i.e. the allocation of existing resources given current technology and consumer preferences Commission [2.12]). That was an important element, but not the only important element, in relation to complex economic issues. It was by no means dispositive.
42. Ofcom had concluded that allocative efficiency did not provide a clear answer on whether LRIC plus or LRIC was to be preferred; the move from LRIC plus to LRIC was "highly unlikely to trigger a substantial reduction in ownership and was likely to generate a limited increase in usage" [Commission 2.13]. The Commission examined ten different aspects of allocative efficiency (there is an invaluable table of topics given by the Tribunal at [147]). They were considered in 298 separate paragraphs (2.525-2.823) over 69 pages.
43. The Commission agreed with Ofcom that allocative efficiency grounds alone did not provide a clear answer, although it did not agree with all aspects of Ofcom's reasoning [2.929(b)]. The Commission's starting point was to accept EE's view that Ofcom was wrong to predict that retail price rises would primarily be focussed on post-pay customers. On the contrary, the Commission agreed with EE that such price rises would primarily be focussed on pre-pay customers and especially on low usage customers [2.624]. The Commission then considered the effect of lower mobile termination rates on ownership and on subscriptions and on mobile and fixed-line usage in light of its conclusion as to where the impact would primarily be felt. There was a substantial dispute as to the magnitude of those effects; the appellants EE and Vodafone were arguing that they would be substantially greater than was suggested by those who supported the shift to LRIC. The Commission considered the effects on ownership and subscriptions on the basis of three types of evidence which it regarded as "directly relevant": traditional elasticities of demand, survey evidence and, third, a study commissioned by Ofcom as to the relationship between termination rates and subscriptions [2.673].
44. The Commission did not place any significant weight on traditional elasticities of demand or on the research. It then turned to what it felt able to draw from surveys proffered by EE to Ofcom and by Vodafone on appeal to the Commission. It is in that limited context that it subsequently made the comments on the need for robust surveys as to customer responses on which EE so heavily relies at [2.700]. I must follow the Commission's route to that passage.
45. Mr Turner QC, on behalf of EE, seemed to suggest that the Commission had eschewed reliance on the surveys advanced by EE and Vodafone. Indeed, that was the foundation of his argument that in criticising those surveys, and in noting that Ofcom relied on little relevant evidence, the Commission acknowledged that it had no reliable evidence on which to base its own conclusion.
46. On the contrary, it is plain that whilst it had reservations as to those surveys, the Commission did deploy them in order to reach a view as to the extent of the impact on pre-pay customers. The Commission disagreed with Ofcom's dismissal of EE's survey; despite Ofcom's criticisms, the Commission thought that that survey might provide some useful information on the relative effects of different types of price increase and the relative response of low-income households [2.691].
47. The Commission also considered the second of two ICM surveys commissioned by Vodafone. It found ICM's approach reasonable and sensible but subject to the unavoidable problem that when customers are asked what they would do when faced with a price rise, they may not behave exactly as they say they would. The Commission commented that a well-designed survey can provide useful and reliable evidence on consumer behaviour [2.692].
48. The Commission then observed that Vodafone's survey failed to assess the difference between customers' predicted reaction to price increases flowing from LRIC plus as opposed to those which would be expected to follow the introduction of LRIC. Despite that difficulty the Commission chose to rely on the survey's prediction that a number of pre-pay customers would stop using mobile services. It commented that the survey did not tell the Commission 'how bad' a move from LRIC plus to LRIC would be [2.694].
49. The Commission then made what it described as a strong assumption in favour of the appellants. Based on the strong assumption that price difference would be directly reflected in reduction in ownership (which the Commission did not think entirely reliable or defensible) it predicted loss of use of 230,000-275,000, about 0.5% of mobile users, or 0.3% of subscriptions [2.695].
50. The high watermark of EE's case, the Commission's overall assessment on customer responses [2.700], must be considered in the context of the Commission's refusal to reject the surveys on which the Appellants relied and its reliance on them to make strong assumptions in their favour on the issue of the magnitude of the predicted loss of mobile ownership. The Commission did not in any way conclude that absent a robust survey as to consumer responses to price increases it could not evaluate the magnitude of customer loss. On the contrary it used Vodafone's survey to make predictions based on strong assumptions as to reduction in ownership in favour of the appellants.
51. Moreover, the Commission continued by demonstrating its view that although the customer response to price increases was a key issue, it was by no means determinative of the question of allocative efficiency. It regarded three further categories of evidence as relevant to that issue: handset subsidies, calling patterns and the ongoing costs of retaining customers. None of those considerations depended upon survey evidence of customer responses.
52. The Commission then turned to issues relating to the scale and form of price changes. It introduced those topics by making the comment:-

"2.731 The appellants argued that Ofcom had underestimated the effects on mobile ownership and subscriptions. Part of that claim is based on their view that Ofcom's reasoning on the pattern of price changes is incorrect. As we discussed above, we find force in that view. However, that also means that some of the arguments made a start from Ofcom's conclusions on price changes rather than the position we have taken. Therefore we apply the parties' logic and evidence as best we can."

This was not an admission that the Commission was relying on evidence which it regarded as inadequate and merely making the best of it. It was doing no more than recognising that some of the arguments were comments on the scale, target and form of price changes predicted by Ofcom. The Commission went on to make its own predictions as to the scale, target and form of such changes.

53. The Commission rejected the appellants' contention that there would be "relatively large" price increases for some users. Even adopting, again in favour of the appellants, conservative increases of £5 or £8 per subscription per year [2.737] (and footnote 637), it concluded that there would only be modest increases in the level of charges and usage charges [2.742]. This significant conclusion did not depend in any way upon evidence obtained from surveys.
54. The Commission could reach no definitive conclusion as to the form those price changes would take, but made judgments so as to predict the effect different forms might have on customers, [2.745], a useful example of the need to deploy its expertise and skill in making predictions.
55. After concluding that it was difficult to draw a firm conclusion as to the effects on mobile usage and considering the effects on fixed-line users, the Commission reached its conclusions as to the implications of its views on allocative efficiency. It did not accept that any reduction in ownership or usage would damage allocative efficiency [2.799], on the contrary, it thought that the loss of some customers could increase it [2.808]. It concluded that its discussion as to the implications of changes for allocative efficiency "illustrates the difficulty of drawing strong and robust conclusions on allocative efficiency in a complex market" [2.812]. That comment is a powerful demonstration of the Commission's realisation that its assessment of the impact of price changes was a prediction which was itself based on other predictions (as to, for example, scale, target, form).
56. The Commission then made an overall assessment on allocative efficiency. That overall assessment merely summarised its previous detailed discussion. EE relied on two paragraphs in that section. Neither purported to add anything to what the Commission had said before. At 2.819 it again commented on the survey evidence :

"2.819 We have also considered evidence on the responsiveness of consumers to price increases. We would normally expect this question to be addressed using empirical evidence, but Ofcom relied on little relevant evidence in its decision and we found that the evidence of the appellants did not allow us to a make a reliable assessment on the scale of reactions to price increases. This is further complicated by the fact that most evidence refers to the number of subscriptions rather than the number of subscribers.719 We did not think that any of the evidence demonstrated that moving from LRIC+ to LRIC would lead to significant reductions in subscriber numbers, relative to the level of subscribers in the UK today."

This was no more than a repeat of its view as to the inadequacies of the survey evidence.

57. The Commission summarised its conclusion that allocative efficiency provided no clear answer in a further paragraph on which EE relies :-

"Overall, in light of the available evidence we find certain aspects of the reasoning of EE, Vodafone and Telefónica convincing and prefer it to Ofcom's, particularly regarding the form of price changes that are likely to follow a reduction in MTRs. We believe that Ofcom's reasoning has led it to underestimate the negative effect on mobile ownership of adopting LRIC in preference to LRIC+. We also consider that there are no good grounds to expect LRIC to cause an increase in mobile usage (an increase or decrease are both possible); and Ofcom may have overstated the increase in fixed usage. However, the appellants have not provided convincing evidence that the scale of decline in the number of users would be significant; and the appellants have not demonstrated that this constitutes a significant negative effect on allocative efficiency. Most of the evidence available relates to the number of subscriptions, and we treat it with caution for three reasons:

(1) most of the available evidence is not robust, is not aimed at the difference between LRIC and LRIC+, or both; (2) it is not clear how a decline in subscriptions translates into a loss of users; and (3) as we set out above, the loss of a subscription that was being subsidized (i.e. its owner valued being on the network less than the cost of being on the network) is not necessarily allocatively inefficient. To the extent that there is some loss of 'efficient' users, that has to be set against all the other effects of higher MTRs (such as higher FTM prices). Therefore we agree with Ofcom that allocative efficiency grounds alone do not provide a clear answer as to whether a LRIC or LRIC+ cost standard should be preferred. For these reasons, bearing in mind the statutory framework within which Ofcom was required to make its decision and the burden being on the appellants to prove that Ofcom erred in its conclusion that LRIC was, in particular, appropriate for the purposes set out in section 88(1)(b) of the Act, and notwithstanding those matters on which our conclusions differ from the conclusions reached by Ofcom under this part 2(a), we do not believe that Ofcom was mistaken, in respect of the appropriateness or otherwise of its choice for promoting efficiency,724 in choosing a LRIC cost standard." [2.823]

58. This paragraph shows that the reason the Commission took the view that allocative efficiency provided no clear answer as to whether LRIC or LRIC plus was appropriate was not because it lacked sufficiently robust survey evidence as to customer responses to price changes. Only the first of the three reasons for caution relates to survey evidence. That paragraph, read as a summary of all of the discussion which went before, contradicts any suggestion that the Commission was unable to reach a conclusion as to allocative efficiency because it thought it lacked necessary evidence. It was able to reach a conclusion: its conclusion was that allocative efficiency provided no clear answer as to which of the two cost standards it should choose. That was the reason why it concluded that the arguments as to allocative efficiency did not support the conclusion that Ofcom was wrong in its choice of LRIC:-

"(b) We summarize our views of the challenges to Ofcom's allocative efficiency assessment in paragraph 2.823. Though we do not agree with all aspects of Ofcom's reasoning on allocative efficiency, we agree with Ofcom that allocative efficiency grounds alone do not provide a clear answer as to whether a LRIC or LRIC+ cost standard should be preferred. For these reasons, and notwithstanding those matters on which our conclusions differ from the conclusions reached by Ofcom, we do not believe that Ofcom was mistaken, in respect of the appropriateness or otherwise of its choice for promoting efficiency,852 in choosing a LRIC cost standard. In addition we agreed with Ofcom's conclusion that the adoption of LRIC is likely to have little effect on dynamic efficiency.853" [2.929(b)]

59. After concluding that the effect on vulnerable consumers would not be significant, the Commission rejected the appeals. Its final paragraph shows that it well understood that mere disagreement with Ofcom on some of the issues should not necessarily lead to disagreement as to the conclusion:-

"2.931 There are issues where we find some force in the appellants' arguments. However, in order to find that Ofcom erred in adopting LRIC rather than LRIC+ as a cost standard, we would need to find errors that would materially affect Ofcom's judgement. We do not hold Ofcom to be wrong simply because we consider there to be some error in its reasoning on a particular point—the error in reasoning must have been of sufficient importance to vitiate Ofcom's decision on the point in whole or in part.854 Having regard to our conclusions on the four limbs of assessment, set out in paragraph 2.929, and to the additional support that the Recommendation provides for Ofcom's conclusion we do not believe that it has been demonstrated that Ofcom was wrong in deciding that the LRIC cost standard was appropriate by reference to the statutory duties and considerations in sections 3, 4 and 88 of the Act. Nor do we believe that the appellants have demonstrated that these statutory duties and considerations would have been better served by the setting of the price control by reference to a LRIC+ methodology."

60. Part of the respondents' resistance to the appeal relied on the Tribunal's view that it was EE's own fault if the evidence was insufficient. It is unnecessary to dwell on this point in the light of my conclusion that the Commission did not find that it lacked sufficient evidence to support Ofcom's decision. I merely reiterate the principle I endeavoured to express previously. It was for EE to marshal such evidence and material as it could to demonstrate that LRIC plus and not LRIC was the appropriate price control measure. Once Ofcom had dismissed its survey, it was open to EE to obtain a new survey. It chose to rely on that deployed by Vodafone. The Commission was entitled to comment on the inadequacies of that survey. It was not open to EE thereafter before the Commission or the Tribunal to seek to make good any deficiencies. There was some suggestion by the respondents that once EE had been shown the provisional determination it could have sought to make good those deficiencies. But by then it was, as EE argued, far too late. The provisional determination which criticised the survey evidence was published on 21 December 2011, the deadline for publishing the final determination was 9 February 2012. It was too late for EE to obtain another survey. But it had no right to a third attempt.
61. All parties face the difficulty in relation to surveys that their efficacy will depend upon accurate foresight of the "facts" (more accurately, predictions) adopted by Ofcom or the Commission on the basis of which customer responses are sought. For example, a survey which asks questions on the basis of lower price rises than that which the regulator expects will be deprived of much utility. In this case, the Commission's criticisms of the surveys, whilst not as dismissive as Ofcom, were criticisms of evidence adduced by and relied on by the appellants. There is no warrant for giving the appellants another chance of curing them.
62. But EE's responsibility for inadequate survey evidence is beside the point. The Commission did not feel inhibited, by any absence of evidence, from making judgments on an array of issues, all of which went to its rejection of the appeal. Still less can it be inferred that it only upheld Ofcom's conclusion because it felt that it had, as a matter of law, to choose one cost standard or the other. I would reject EE's first ground of appeal.

Ground 2: Procedural unfairness

63. In the course of its assessment as to allocative efficiency, the Commission concluded that the loss of users resulting from the change to LRIC would be smaller and not as harmful as the appellants, EE and Vodafone, would have had it believe [2.808, 2.809]. On the contrary, the loss of some users might improve efficiency. In this respect the Commission disagreed with Ofcom, which had taken the view that any loss would be detrimental to efficiency.
64. Ofcom's view had not been disputed on appeal to the Tribunal. EE contends that the point as to beneficial effect had never been put to it and that it never had a fair opportunity to deal with it.
65. During the course of the appeal there were bilateral hearings with both of the two appellants, Vodafone and EE. Bilateral hearings are an important part of the Competition Commission's procedure when hearing price control appeals for the reasons explained in its Guidelines (CC13) of April 2011. They present an opportunity, amongst other things, for the Commission to explore key issues with the parties (5.6). They take place after core submissions and proceed on the basis that the parties have already made developed arguments (5.8). Transcripts of the hearings, with the identification of confidential matters, are distributed to the parties afterwards (5.5).
66. As any advocate will tell you, questions from a tribunal or bench usually indicate the points which most trouble the interlocutor. The point was clearly made to Vodafone during the course of its bilateral hearing on 19 October 2011, as the Commission recalled in its determination [2.807]. EE had the transcript on 8 November 2011 and had until 23 November to respond. Of course, it is easy to miss points when reading through transcripts but Vodafone was a fellow appellant. Much was at stake.
67. The point was, in any event, previously raised in Three's written submissions in support of a change to LRIC. It was also raised by a member of the Commission, Professor Cubbin, at EE's own bilateral hearing (18 October, 2011, p.79). EE had the opportunity to respond to the point through its witnesses at that hearing and, on my reading of the transcript (of course, it is not possible to get the full flavour of the exchange), did so. It had the opportunity to respond later, an opportunity of which it availed itself on 2 November 2011 in relation to some of the issues raised at the bilateral hearing. EE has no basis for complaint. I would dismiss the second ground.

Ground 3: Inconsistency

68. Mr Turner QC accepts that EE cannot succeed on this ground unless it also succeeds on the first. In my view EE has failed to identify any error in the Commission's determination which would have justified the Tribunal in setting it aside on judicial review grounds. I would dismiss EE's appeal.

Lord Justice Patten:

69. I agree.

Lord Justice Longmore:

70. I also agree.