The UK's Competition Appeal Tribunal (CAT) has overturned a decision of the Competition and Markets Authority (CMA) which found that "most favoured nation clauses" used by the price comparison website Compare the Market infringed competition law. In so doing, the CAT made important observations regarding the correct approach to market definition – both generally and specifically as regards 'two-sided' markets – and the correct approach to bringing a 'by effects' case. The CAT also made several criticisms of the manner in which the CMA approached and relied on evidence, both in the decision and on appeal. The judgment will likely have implications beyond the case at hand.

Key issues

• The CAT's judgment set aside the CMA's Decision, overturning the CMA's findings both as regards market definition and as regards anticompetitive effect.
• The judgment contains important observations on how market definition should be approached in 'two sided' markets, finding that each side of the 'market' might itself form separate product markets, with other sales channels posing important competitive constraints.
• It also contains important observations on the anticompetitive effects of wide MFNs. In the present case, the CAT found that there were no anticompetitive effects and, in doing so, doubted that wide MFNs are necessarily injurious to competition. This calls into question the recent inclusion of wide MFNs as a category of "hardcore" restrictions in the UK's Vertical Agreements Block Exemption Order.
• The judgment also contains criticisms of the way in which the CMA approached the decision, advanced and weighed evidence, and defended this decision on appeal, which may have implications for the approach the CMA takes in future appeals.

Background

In November 2020, the CMA issued the Decision against CTM finding that CTM infringed Chapter I of the UK Competition Act 1998 and Article 101 of the Functioning of the European Union 'by effect' between 2015 and 2017. The CMA imposed a c.£18m penalty.

The Decision took issue with contractual obligations known as wide most favoured nation clauses (wide MFNs), which were imposed by CTM in its agreements with a number of home insurance providers. These clauses required home insurance providers to provide to the price comparison website (PCW) the lowest (or equal lowest) prices on offer anywhere for that product, whether on other PCWs or via their own direct channels. In contrast, narrow MFNs prevent the home insurance provider from undercutting the prices quoted by it on the PCW only on its own website or other direct marketing channels.

The Appeal

CTM appealed the Decision.

The focus of the appeal was to challenge the CMA's market definition and the CMA's finding that the wide MFNs had an anticompetitive effect, and the penalty imposed. In its judgment of 8 August 2022, the CAT substantially agreed with the appeal and set aside the Decision.

The CMA erred in defining the market

As the CMA brought this as a 'by effects' case, it was required to define the relevant market. The Decision characterised the market as 'two-sided' comprising the supply by PCWs of:

• customer introduction services to home insurance providers, for which price competition arose in respect of the commissions charged by PCWs to home insurance providers when a customer purchased insurance from the insurer via the PCW (commissions); and
• price comparison services to consumers, for which the focus of price competition was the premium charged by the insurer to the customer (premiums).

Nevertheless, the Decision found that the relevant market was the provision of PCWs for home insurance products in the UK; i.e. there was a single overall product market that included both 'sides' of the 'two-sided' market. The CAT overturned this finding. In so doing the CAT made two important observations.

First, the CAT's essential criticism was that the CMA's market definition exercise had pre-determined a finding of anti-competitive effects. In the Decision, the CMA recognised that there are multiple ways in which 'twosided' markets can be defined but said that this meant it was free to choose which approach is “appropriate” in the given case. The CAT criticised this as being unpredictable and importing judgmental factors into market definition.

Second, regarding 'two-sided' markets, the CAT rejected the CMA's approach of defining a single overall market. Rather, the CAT held that the different sides of the market have different constraints, which in turn required separate examinations. The CAT held (contrary to the consensus of the expert economists that gave evidence) that the CMA should have examined the supply of customer introduction services to home insurance providers and, separately, the supply of price comparison services to consumers. On assessing these separate markets, the CAT found that "the CMA’s market definition is not fit for purpose". For example, it found that PCWs not only compare prices, but also enable customers to conclude insurance contracts (by clicking through the quote to the insurer's website). Accordingly, PCWs are merely a form of intermediation, and that, ultimately, a customer merely wishes to contract for insurance, whether directly or indirectly. Moreover, the CAT found it "remarkable and odd" that the Decision did not even examine the other available purchase channels, including direct sales by home insurers (e.g. through their own website, or renewals). On examination, the CAT found these other channels acted as a constraint on PCWs.

Third, the CAT provided its own market definition, distinguishing between the two 'sides' of the platform. The CAT agreed that the upstream market was limited to customer introduction services to home insurance providers. However, as regards the downstream market, the CAT found that the market should include all channels, including direct channels. In reaching this conclusion, the CAT noted that customers do not 'pay' for the use of PCWs and hypothesised that, even if there were a monopolist for the supply of PCWs, it would still be unable to introduce a charge to consumers for its, because other channels pose a competitive pressure. Accordingly, these other channels fall within the market.

The Decision was wrong to find anticompetitive effects

The Decision found that CTM's wide MFNs had anticompetitive effects. CTM was guaranteed to have the lowest premiums from the relevant insurers without ever having to lower its own commission fees or provide some other benefit to the insurers. In addition, CTM was able to increase its commission without the relevant insurers being able to fully reflect that increase in the premiums they quoted on CTM compared to the premiums quoted on other PCWs. The CAT overturned the Decision's finding of anticompetitive effects, making a number of both general and specific observations.

First, the CAT noted that the Decision's analysis of the effects of the wide MFNs was limited to qualitative evidence (e.g. contemporaneous documents and responses to requests from the CMA). The CAT accepted that a case can be brought on such a basis. However, it noted that, in the present case, the evidence brought simply showed that the wide MFNs were effective and constrained the relevant home insurance providers' ability to quote lower premiums on rival PCWs. That says nothing, however, as to whether those effects were anticompetitive, which is the essential question in a 'by effects' case. Rather, what the Decision needed to prove, said the CAT, was that, in the counterfactual world without the wide MFNs, there would be greater incentive for home insurance providers to reduce the premiums that they charged to consumers. The CAT was strident in its criticisms in this regard, stating that a "great deal of the analysis [in the Decision] operates at the level of theory or (less helpfully) bare assertion" (CAT's emphasis) and noted that the CMA's qualitative evidence was "anecdotal (at best)", lacked "depth (bald explanations, without detail, are the order of the day), and was inconsistent with the CMA's theory of harm". Moreover, the CAT found that such qualitative evidence – which necessarily was limited to the views of the relevant authors – was of limited utility, since a person could not opine on the effect of wide MFNs on commissions and premiums across portions of the industry.

Second, the CAT held that there were also several general features of the market that militated against the existence of anticompetitive effects.

• Inter-brand competition between the home insurance providers was not constrained by the wide MFNs. Rather, the most a wide MFN clause could do was eliminate intra-brand competition, i.e. competition between different channels for the sale of the same home insurer's products. The CMA had not considered the impact of such inter-brand competition. Yet it was significant. For example, home insurance providers want to encourage customers to renew their insurance without searching for alternatives, whereas PCWs compete to try and ensure that customers search for alternatives. Moreover, PCWs compete to ensure that they have a significant number of brands on their websites, thereby enhancing interbrand competition.
• There were limits to the degree to which wide MFNs could constrain intrabrand competition in this case. The wide MFNs in question only applied to differential pricing in respect of the same product to consumers with the same risk profile. However, home insurers assess a consumer's risk profile on the basis of the answers to set questions that are posed by the PCW. As PCWs often ask different questions, the same consumer can end up being offered different premiums on different PCWs without any PCW having breached its wide MFN obligation. In any event, the CAT noted that home insurance providers considered it critical to appear as high as possible on PCWs' search outputs, by offering the lowest premium. Accordingly, it was unclear why a home insurance provider would actively wish to price differentially across different PCWs even absent the wide MFNs. The CAT was not persuaded by the qualitative evidence the CMA provided in this regard.
• Moreover, narrow MFNs prevailed throughout the market and were not challenged by the CMA. The CAT referred to an economic paper which had found that narrow MFNs could produce similarly harmful effects to wide MFNs in certain circumstances. If that were the case here, CTM's wide MFNs may have made no difference, but the CMA had not addressed that question.

Third, the CAT accepted that the quantitative evidence – advanced by CTM – suggested that the wide MFNs did not have anticompetitive effects. Importantly, the CMA did not rely on any quantitative evidence, a decision the CAT found "prima facie odd and difficult to justify". This was particularly so given that, after the CMA began its investigation, the wide MFNs were withdrawn, thereby providing a natural experiment by which to conduct a "before and after" test for the effects of the wide MFNs. Moreover, the CMA itself had relied on quantitative evidence to examine wide MFNs in another case.

The CAT levelled a number of important criticisms at the CMA's approach to the handling of evidence

When dealing with the question of whether there were any anticompetitive effects, the CAT levelled several criticisms at both the Decision's approach to the use of evidence and the CMA's approach to evidence on appeal.

First, the Tribunal faced a practical difficulty in understanding what evidence the CMA was seeking to rely on. While the Decision contained cross-references to the evidence on which it was said to be based, it was "extremely difficult […] to get a sense of the true nature of this evidence." The CAT was not in a position to find each of the documents in every footnote and to chase down each cross-reference. While the Tribunal made various requests to the CMA to understand its evidential position, ultimately, the Tribunal found that "despite the superficial specificity in the Decision – it was 2,708 footnotes and 794 pages – there is no meaningful corpus of material, capable of being considered, that constitutes the evidence on which the CMA relies, and which constitutes the foundation for the factual findings in the Decision."

Second, the Tribunal outlined the approach that it expected to be taken in CA98 decisions. They should draw a "hard-and-fast distinction" between: (i) evidence; (ii) analysis of that evidence or inferences being drawn from it; and (iii) conclusions of fact drawn from (i) and (ii). Each of these elements should have their own place in the decision and need to be properly, and separately, set out. In practice, a decision should articulate factual findings and the evidence on which they are based should be specifically referenced and (ideally) contained in an altogether separate part of the decision. This evidence should not be synthesised but should be the primary material on which that particular factual conclusion is based. The fact is that unless the factual basis for a decision is properly stated, it can neither be properly attacked nor defended. The CAT found it impossible to identify the CMA’s primary facts; and so impossible to understand the analyses and inferences from those facts. This issue was compounded by extensive repetition of generalised summaries of evidence and assertions. As a result, the CAT considered that the conclusions drawn in the Decision rested on shaky evidential foundations.

Third, the CAT accepted that documentary evidence was admissible without the CMA putting forward witnesses. The CAT does not operate in accordance with the strict rules of evidence and does not require the CMA to call each and every person it has interviewed for cross-examination. However, the CAT noted that, without some important witnesses who could speak to the documents in question, they will not, automatically, be accorded weight or substance. Indeed, the CAT noted that where oral or other evidence might be material, the principle is that issues must be resolved against, rather than in favour of, the CMA.

Fourth, the CAT confirmed that, while it was best practice for a company under investigation to put forward all of its points during the investigation phase, there was no bar to an appellant (as opposed to the CMA) adducing entirely new evidence on appeal provided it is articulated and identified in the notice of appeal. The CMA will be entitled to respond to such evidence, and often will.

Conclusion and Take Aways

The CAT's judgment is an important reminder that, where a UK competition authority is required to prove anticompetitive effects, it is not enough to explain how the alleged anticompetitive effects might arise in theory, without showing that such effects were likely to have actually occurred.

The judgment is arguably a considerable set-back for the CMA's comparatively more strident approach to wide MFNs. On the advice of the CMA, the Government recently included wide MFNs as a category of "hardcore" restriction under the UK's Vertical Agreements Block Exemption Order (VABEO). Such hardcore restrictions are not only excluded from the benefit of the VABEO, but are also, according to the CMA's guidance on the VABEO, "generally" considered to be restrictions of competition "by object". Where a restriction is qualified as being "by object", the CMA is not required to demonstrate anti-competitive effects, given the presumed propensity to harm competition. The fact that the CAT has found that CTM's wide MFNs had no proven anticompetitive effects calls into question the logic behind their legal characterisation as "hardcore" in the VABEO and their presumptive "by object" status. In contrast, the European Commission has opted to treat wide MFNs as "excluded restrictions" that must be assessed individually for anticompetitive effects under the equivalent EU block exemption for vertical agreements (see our briefing on the new EU and UK competition regimes for distribution arrangements). As such, the CAT's ruling is arguably more in line with EU legislation than with the UK's new VABEO.

The judgment may also cause the CMA to reflect more widely on how it approaches infringement decisions. One key area will be as regards market definition. This is most obviously the case as regards 'two-sided' markets, with the CAT finding that such markets should comprise two distinct product markets. In this context, the CAT also provided guidance on the requirement for authorities to not pursue 'outcome based' definitions. In practice, the CAT's approach to market definition may have implications for many other industry sectors, particularly where intermediation services play a role: the CAT's judgment provides support to arguments that other sales channels, including direct sales channels, form part of the same relevant market or at least constitute important competitive constraints that must be reflected in the assessment.

Finally, the CMA may need to reflect on how it drafts its decisions, and how it chooses to defend its decisions on appeal. In the CAT's view, any decision needs to clearly identify the corpus of material that constitutes the evidence on which the CMA relied, and which constitutes the foundation for the factual findings made in the decision. According to the CAT, that material cannot simply be all the documents relied on in the decision – the material needs to be capable of being considered. Moreover, if the decision relies on documentary evidence, the CMA will need to consider the potential benefits of producing witnesses that can be cross-examined on such evidence.