CMA merger decisions are subject to judicial review by the Competition Appeal Tribunal (CAT). Challenges to the CMA’s substantive decision-making have, however, generally been unsuccessful. Although the CAT has been willing to intervene on matters of procedural fairness and errors of law, as recent decisions confirm, the CAT is reluctant to intervene in the CMA’s assessment of competitive effects and identification of remedies.

The wide margin of appreciation that the CMA enjoys derives from the CAT endorsing a broad “probabilistic” approach to the assessment of evidence together with its strict application of the judicial review standard of “irrationality.” As a result, merging parties have found it difficult to overturn the CMA’s substantive findings on appeal absent a clear and unequivocal error.

The cost implications of judicial review have also increased following a series of recent judgments. Unsuccessful applicants are likely to find themselves responsible for substantially all of the CMA’s litigation costs, including costs incurred by in-house lawyers, while successful applicants can no longer assume that the CMA will be liable for the costs that they have incurred.

The Standard of Review

The CMA must decide, on a balance of probabilities,[1] whether a transaction constitutes a relevant merger situation and whether it is expected to result in a substantial lessening of competition (SLC).[2] If it does, the CMA must also decide (on a balance of probabilities) what remedies will be effective in addressing that SLC and which of the available remedies is least intrusive.[3]

When applying this standard, the CMA is not obliged to find that any particular theory of harm is more likely than not to occur, or that any particular evidence is dispositive. Rather, it must give “ full and proper consideration to the evidence which it has gathered, and apply the ‘probabilistic test’ at the end-point[4] and may consider evidence “in the round.” In circumstances where the CMA identifies a catalogue of arguments against a merger, this “probabilistic” approach makes it difficult to challenge the CMA’s conclusions even if individual arguments are susceptible to criticism.

While the CMA can prohibit a merger based on a balance of probabilities, the threshold for challenging the CMA’s finding before the CAT is higher. In reviewing the CMA’s merger decisions, the CAT applies a judicial review standard.[5] There are four established grounds of challenge in judicial review in the UK: illegality, irrationality, procedural unfairness, and proportionality. The CAT has consistently held that, in questions of substantive assessment, CMA decisions will be reviewed under strict principles of irrationality under the Wednesbury unreasonableness test, i.e., was the decision so irrational that no reasonable competition authority could have reached the same decision? The CAT applies this test not only to questions of economics and competitive effects, but also when reviewing the CMA’s assessment of proportionality and matters of procedure. For example, the CAT does not ask itself whether a remedy is effective and proportionate but whether the CMA was irrational to reach the conclusion it did on the effectiveness and proportionality of remedies.

Even where applicants are successful, the CAT’s remedial powers are limited to setting aside the decision (in whole or in part) and, if appropriate, remitting the matter to the CMA.[6] The CAT cannot replace the CMA’s decision with its own.[7]

Recent Challenges to CMA Merger Decisions

Recent CAT judgments show how difficult it is for parties to succeed in applications for judicial review of CMA merger decisions. In Ecolab/ Holchem,[8] the CMA found that the merger would likely give rise to an SLC in the supply of formulated cleaning chemicals and ancillary services to food and beverage manufacturers in the UK, a market in which the parties had a combined market share of below 40%. The CMA rejected Ecolab’s proposed remedy to transfer customers and assets to a rival. Instead, it ordered Ecolab to divest the Holchem business, effectively prohibiting the transaction.

Ecolab challenged the CMA’s decision before the CAT on four grounds relating to the CMA’s SLC decision and its rejection of Ecolab’s proposed divestment remedy.[9] The CAT rejected all four grounds on essentially the same basis: it restated that the CMA has a “wide margin of appreciation” in deciding the extent to which it is necessary to carry out investigations to discharge its statutory duties and maintained that the CAT should intervene only if no reasonable public authority could have arrived at the same conclusion on the basis of the available evidence.[10] It concluded that the CMA had gathered sufficient evidence to reach an SLC finding and to reject Ecolab’s alternative remedy. It further found that the CMA did not have to carry out further consultation on Ecolab’s proposed remedy because, in the CMA’s view, there was no reason to suppose that modifications to that proposal would have been able to overcome the shortcomings already identified by the CMA.[11]

In Tobii/Smartbox,[12] the merging parties supplied hardware, software, accessories and related services to enable people with speech, language and communication difficulties to communicate (known as augmentative and assistive communication solutions (AAC)). The CMA found—on the basis of both horizontal and vertical theories of harm—that the merger would give rise to an SLC as a result of reductions in the existing product range and quality, less new product development, and higher prices. It ordered Tobii to sell Smartbox.

Tobii challenged the CMA’s decision before the CAT arguing that the CMA had: (i) breached its duty of procedural fairness by refusing to disclose to Tobii and/or its external advisers relevant evidence which formed the basis of the CMA’s findings, (ii) failed properly to define the market for AAC solutions, and (iii) failed to support its SLC findings with relevant, reliable, and sufficient evidence due to material errors in the CMA’s collection of evidence.[13]

The CAT rejected the majority of Tobii’s grounds. First, it rejected the submission that the CMA had failed to provide Tobii with sufficient evidence because, having reviewed the evidence, the CAT considered Tobii had enough information to understand the gist of the case. Second, it found that the CMA had properly defined the market for AAC solutions, including because, on the facts of the case, the CMA was not under a duty to ask customers how they would respond to a 5 to 10% price increase (the classic “SSNIP test”) or take into account product differentiation when defining the market. Third, the CAT dismissed Tobii’s challenge to the CMA’s approach to the collection of evidence to support its SLC, restating that “the question of precisely where the line is drawn in determining whether an inquiry has gone far enough is an issue for the relevant authority to evaluate and the Tribunal will need to be shown a strong case to show that the relevant authority manifestly drew the line in the wrong place.”[14]

Tobii succeeded, however, in demonstrating that the CMA’s finding of harm to competition due to partial input foreclosure did not have a sufficient evidential basis. The CAT first restated that the CMA has a “wide margin of appreciation” as to the extent to which it is necessary to carry out investigations.[15] It nevertheless concluded that the CMA had failed to establish that the merged entity would have an incentive to engage in partial input foreclosure, in particular because it did not calculate the likely diversion ratio that would arise in a partial foreclosure scenario, or whether it would exceed the minimum diversion ratio that would make a partial foreclosure strategy profitable. This finding did not alter the outcome or legality of the CMA’s decision, meaning that Tobii remained under a duty to sell Smartbox.

Challenges to the CMA’s Timetable

The CAT has shown greater willingness to intervene on matters of procedure. In Sainsbury/Asda v CMA,[16] the CMA sent the parties 19 Working Papers[17] between 9 and 28 November 2018, asking for responses to all of them by 7 December 2018. The parties asked for an eight-week extension to the CMA’s statutory timetable[18] and proposed to respond to all the Working Papers by 4 January 2019. The CMA said that this proposed timeline would jeopardise subsequent stages of the process, even accounting for a possible eight week extension. It asked the Parties to submit their responses by 17 December at the latest and offered to hold a main party hearings in the week ending 14 December.

On 12 December 2018, the parties applied to the CAT, challenging the CMA’s refusal to grant an extension to respond to the CMA’s Working Papers, as well as its decision to schedule the main party hearings on a date that did not give the parties time to explain their position on important points in the Working Papers to the decision makers. The CAT agreed, finding that the original 7 December deadline was unreasonable and unfair, given the volume and complexity of the Working Papers. It also ruled that the CMA had not given the parties sufficient time to prepare for the main hearing. It did not, however, stipulate new deadlines, leaving this to the discretion of the CMA.

Comparison with the EU General Court

The EU’s courts, the General Court and Court of Justice, have historically subjected European Commission merger decisions to more rigorous review than the CAT in respect of CMA decisions, and have overturned around 20% of the prohibition decisions rendered by the European Commission since the European Merger Regulation came into force.

Most recently, on 28 May, the General Court in Three/O2 overturned the European Commission’s prohibition of a transaction that would have combined two of the four UK mobile telecoms providers.[19]

The facts are complex, but there are two clear differences of approach between review of this decision by the General Court and judicial review before the CAT: (i) the European Commission was held to a higher standard of proof (namely a “strong probability” standard) than has been required of the CMA, and (ii) the General Court broadly applied a “manifest error” standard of review, while the CMA continues to apply a strict irrationality standard.

Costs Awards in CAT Appeals

A series of recent judgments show that the CAT will be generous to the CMA in cases where it successfully defends its decisions, while applicants may not recover any costs even when they are successful.

Unlike the position in most civil courts,[20] there is no general rule in the CAT that the unsuccessful party must pay the costs of the successful party. Rather, the CAT has a discretion to make “any order it thinks fit in relation to the payment of costs in respect of the whole or part of the proceedings.”[21] The CAT’s starting point in regulatory appeals had historically been that costs follow the event.[22] Following the Court of Appeal’s judgment in BCMR, that position has changed. In BCMR, British Telecom successfully appealed a decision by Ofcom and was awarded costs. The Court of Appeal overturned this award. It held that, where Ofcom was acting in a purely regulatory capacity and its actions in defending its decision were reasonable, the CAT should not start from an assumption that Ofcom would be liable for costs.[23]

The Court of Appeal restated this principle in Flynn/Pfizer (Costs), discussed in more detail below.[24] In this case, the appellants successfully appealed an abuse of dominance decision by the CMA and were awarded costs. The Court of Appeal overturned this award. It held that—even in competition cases—there should be no assumption that the CMA will be liable for costs when it loses on appeal. This principle has not yet been tested in judicial review of a merger decision.

In contrast, the CAT has been generous in awarding costs to the CMA. In Ping (Costs)[25] and Tobii (Costs),[26] the CAT held that the CMA could recover costs incurred by in-house lawyers as well as costs incurred by outside counsel. The CMA has sought to recover in-house legal costs at Government solicitors’ guideline hourly rates (GHRs), even though these rates greatly exceed the costs of employing CMA staff.[27] Following Re Eastwood,[28] the CAT took the view that attempting to produce a comprehensive analysis of all the costs attributable to the work of in house solicitors would entail an immensely complex investigation. As there was no better way to calculate the CMA’s costs, the CAT considered that it was not in a position to postulate alternative rates. It expressed misgivings about the CMA’s GHRs but nevertheless allowed it to recover its costs at that rate, concluding that in almost all cases … disbelief must be suspended and strained logic must be tolerated “ for the merit of simplicity and of avoiding the burden of detailed enquiry.”[29]


Recent cases confirm that the CAT is likely to intervene in the CMA’s substantive assessment of mergers only where the CMA has clearly acted irrationally. A combination of the CMA’s probabilistic approach to assessing evidence and the CAT’s strict application of the irrationality threshold mean that applications for review face an uphill battle absent a clear and unequivocal error by the CMA Recent rulings also place the cost risk of litigation firmly on applicants. The CMA is likely to recover substantially all of its costs, including in house costs, if it wins, and may not be liable for the merging parties’ costs even if they are successful in their appeal.

[1]      The Court of Appeal has endorsed the approach of expressing an expectation as a more than 50 per cent chance. See IBA Health Ltd v OFT [2004] EWCA Civ 142, paragraph 46.

[2]      Enterprise Act 2002, sections 35 and 36.

[3]      Enterprise Act 2002, section 41.

[4]      Intercontinental Exchange v CMA [2017] CAT 6, at 245.

[5]      See sections 120(4) and 179(4) of the Enterprise Act 2002 (the Act). Note that a different standard applies to penalties imposed by the CMA in the course of merger investigations. The CMA may impose penalties for failure to respond to information requests (section 10(1) and (3)) and for breaches of interim measures (section 94A). These decisions are excluded from the general rights of review under ss.120 and 179 of the Act and are subject to a dedicated avenue of appeal (section 114). In the case of Electro Rent Corporation v CMA [2019] CAT 4, the CAT decided that, although the Act does not set out the standard of review that should be applied, it is not limited to judicial review (at paragraph 68).

[6]      Sections 120(5) and 179(5) of the Enterprise Act 2002.

[7]      The standard of review in the UK differs from that of the EU, where the General Court can conduct a “full judicial review” of the EC’s merger decisions, meaning that it can review the factual basis of the decision, the interpretation of the law on which it is based, and the grounds on which a merger is authorised or prohibited. Although the General Court has said that the EC has “a certain discretion, especially with respect to assessments of an economic nature,” its treatment of EC merger decisions makes it clear that it will subject the EC’s analysis to rigorous scrutiny.

[8]      Completed acquisition by Ecolab Inc. of The Holchem Group Limited, 8 October 2019.

[9]      Ecolab Inc. v Competition and Markets Authority [2020] CAT 12.

[10]    Ibid., at paragraphs 58 and 110.

[11]    Ibid., at paragraph 110.

[12]    Completed acquisition by Tobii AB of Smartbox Assistive Technologies Limited and Sensory Software International Ltd, 15 August 2019.

[13]    Tobii AB (Publ) v Competition and Markets Authority [2020] CAT 1.

[14]    Ibid., at paragraph 199. See also Akzo Nobel N.V. v Competition Commission [2013] CAT 13 at 160.

[15]    Ibid., at paragraph 431.

[16]    J Sainsbury plc and Asda Group Limited v. Competition and Markets Authority [2019] CAT 1.

[17]    By para 52 of Schedule 4 of the Act, a group must have regard to any guidance issued by the CMA Board. The CMA Board adopted the guidance previously issued by the Competition Commission entitled Chairman’s Guidance on Disclosure of Information in Merger Inquiries, Market Investigations and Reviews of Undertakings and Orders accepted or made under the Enterprise Act 2002 and Fair Trading Act 1973 (CC7 Revised). Para 7.3 states that there is no general obligation to disclose any of the many internal working papers produced in the course of an inquiry by and for the Group. It then continues: “However, Groups may disclose some working papers (or extracts from them) during the course of an inquiry or review, where they consider that to do so would assist parties to understand their developing thinking. Whether it is appropriate or practical to do so may depend upon timing considerations; for example, it would not be sensible to do so when the CC is soon to disclose that thinking in an annotated issues statement or provisional findings. However, parties will have the ability to comment following disclosure.”

[18]    The CMA is required to prepare and publish its report within a period of 24 weeks from the start of Phase 2. Under section 39 of the Enterprise Act 2002, the CMA may extend that period by no more than eight weeks if the CMA considers that there are “special reasons” why the report cannot be prepared and published within 24 weeks. See, e.g., Tobii/Smartbox Notice of extension of inquiry period under section 39(3) of the Enterprise Act 2002) (paragraph 76).

[19]    Case COMP M.7612 Hutchison 3G UK/Telefonica UK (O2), Commission decision of 11 May 2016.

[20]    Civil Procedure Rules, Rule 44.2(2)(a).

[21]    Rule 104(2) of the CAT Rules 2015. Rule 104(4) sets out a list of factors that may be taken into account when making an order under rule 104(2) determining the amount of costs which includes, amongst others, whether a party has succeeded on part of its case, even if that party has not been wholly successful.

[22]    See, e.g., Tesco plc v Competition Commission [2009] CAT 26 and PayTV [2013] CAT 9.

[23]    British Telecommunications plc v Office of Communications [2018] EWCA Civ 2542, paragraphs 72 and 83. The Court of Appeal applied a principle established in Perinpanathan [2010] EWCA Civ 40.

[24]    Competition and Markets Authority v Flynn and Pfizer [2020] EWCA Civ 617.

[25]    In Ping (Costs), for example, the CMA proposed to recover over £170,000 for work carried out by its Assistant Legal Director. This was equivalent to two and half years of that individual’s salary. Ping argued that these rates were in breach of the indemnity principle, which provides that a litigant may only recover the cost that have actually been incurred in the course of the litigation. Ping Europe Limited v. Competition and Markets Authority [2019] CAT 6.

[26]    Tobii AB (Publ) v Competition and Markets Authority [2020] CAT 6.

[27]    Ping Europe Limited v. Competition and Markets Authority [2019] CAT 6, at paragraph 24.

[28]    Re Eastwood (dec’d); Lloyds Bank Ltd v Eastwood and others [1975] Ch 112.

[29]    Ibid., at paragraph 50, citing Sidewalk Properties v Twinn and others [2016] 2 Costs LR 253.