UK Competition Appeal Tribunal Judgment Derails CMA Mobile Browsers and Cloud Gaming Market Investigation
On 22 November 2022, the UK Competition and Markets Authority (CMA) decided to refer the supply of mobile browsers and mobile browser engines and the distribution of cloud gaming services through app stores on mobile devices for an in-depth ‘phase 2’ market investigation. The market investigation regime enables the CMA to investigate competition issues that affect an entire sector of the economy and provides it with extensive powers to remedy concerns that are found to exist.
Since the launch of the market investigation, the CMA inquiry team has published a 17-page issues statement and initiated wide-ranging research into consumer behaviour with respect to mobile browser use and the experience of web developers regarding browsers and browser engines. It has also sent out a large number of requests for information to industry participants. According to the CMA’s administrative timetable, it was due to publish working papers at any time and was in the process of scheduling hearings and site visits. In other words, by the start of this year, the investigation was in full swing, as it needed to be to meet the CMA’s case timetable for publishing its provisional findings in September or October this year, in time for its final report to be published before the statutory deadline of 21 May 2024.
On 18 January, Apple responded to the reference by applying to the Competition Appeal Tribunal (CAT) for a review of the CMA’s decision to launch the market investigation, claiming that it had not complied with the statutory timetable and was hence unlawful. Unfortunately for the CMA, in its judgment of 31 March, in Apple Inc. & Others v Competition and Markets Authority, the CAT agreed with Apple, rendering both the reference decision and the entire investigation on which it was based null and void. The resulting detonation of the market investigation clearly creates a major headache for the CMA. Given that the statutory framework within which the CMA operates generally gives the authority broad discretion on the exercise of its functions, it is interesting to examine the specific legal and factual context that led to this surprising outcome.
In the words of the CMA’s press release accompanying its decision to launch a full market investigation, the reference was justified by concerns arising from Apple and Google’s “effective duopoly on mobile ecosystems” that allowed them to “exercise a stranglehold over operating systems, app stores and web browsers on mobile devices”. According to the CMA, its proposal to make a reference had been supported by browser vendors, web developers and cloud gaming services providers, who claimed that the status quo was “harming their businesses, holding back innovation, and adding unnecessary costs”. The CMA was concerned that Apple and Google’s domination of the mobile browser market and Apple’s restriction of cloud gaming through its App Store “limit choice and may make it more difficult to bring innovative new apps to the hands of UK consumers”. A market investigation was thus justified to investigate the impact of the restrictions Apple and Google imposed on businesses and users, to assess Apple and Google’s claimed justifications for these restrictions and, if required, to remedy any adverse effects on competition.
Notwithstanding the substantial reasoning to justify the reference set out by the CMA, however, this outcome was not always the plan. As is usual in such cases, the decision to launch a ‘phase 2’ market investigation was preceded by a ‘phase 1’ market study, which was initiated by a market study notice on 15 June 2021. Six months later, on 14 December 2021, the CMA issued a 445-page interim report on the market study. The interim report indicated that the CMA had reasonable grounds to suspect that features of the markets for mobile operating systems, app stores and app distribution and browsers and browser engines restrict or distort competition in the UK. As such, its provisional conclusion was that the statutory test for it to make a full reference was met. So far so good.
Nevertheless, the interim report explained that the CMA had decided to treat the market study in this case as an input into the new digital regulatory regime, under which its new Digital Markets Unit (‘DMU’) was poised to acquire new statutory powers to regulate technology companies with strategic market status (including Apple and Google) on an ex-ante basis. Not unreasonably, the CMA considered that the newly empowered DMU would “in principle be best placed to tackle the competition concerns identified by this market study” using its new regulatory powers. As a result, while the CMA expressly reserved the ability to revisit its position “if the legislation required to bring the proposed new regime into force is not laid before Parliament for some time, or its anticipated scope materially altered” (paragraph 9.12 of the interim report), it ultimately decided not to make a reference. Since the market investigation regime is essentially a tool to enable the CMA to implement sector-specific regulation, following in-depth investigations lasting 18 months, a decision to bypass a full reference when new regulatory powers for the whole technology sector were apparently within reach seemed reasonable in the context.
The formal decision not to make a reference was duly published on the same day, alongside the interim report. It is a short document, at only one and a half pages. The decision itself made no reference to the CMA’s reasons for not making a reference, simply stating that the “decision not to make a market investigation reference should not in any way be interpreted as the CMA finding no concerns in the sector, only that it considers that any potential concerns would not be best addressed through a market investigation at this time”. The decision focused more on the fact that the CMA had received no representations to the effect that a reference should be made before the deadline specified in the market study notice, i.e., 26 July 2021. Crucially, this specific fact entitled the CMA to decide not to make a reference without any further consultation, under the relevant statutory provisions. Given the pressure on the CMA to investigate concerns over mobile ecosystems, the lack of any such response to the notice is surprising, suggesting that the CMA took a very narrow view of the form that such “representations” needed to take to be counted.
Given the clear decision not to refer, it came as somewhat of a surprise that the publication of the CMA’s final market study report on 10 June 2022 was accompanied by a separate document consulting on a new proposal to make a market investigation reference into mobile browsers and cloud gaming. According to this consultation document, the CMA’s change of course was a response to “multiple calls for further direct action by the CMA since publication of [its] market study interim report in December 2021, along with [its] updated understanding of the potential timing of legislation to place the DMU on a statutory footing”. As the CMA politely put it, although the Government (after a painful delay and rumours that it might even abandon the planned legislation altogether) had finally confirmed that it intended to legislate to give powers to the DMU, the bad news was that “this will not be in the current parliamentary session (ie within the next year)”. Faced with this unexpected delay to its hotly awaited new powers, the CMA changed its mind and that the concerns identified would be better tacked through the exercise of its non-existent regulatory powers and duly published its decision to launch a full market investigation on 22 November 2022.
The CAT’s judgment, in this case, turned on complex points of statutory interpretation. Before getting into these, the important issue to bear in mind, and the most likely source of the CMA’s error, is that it originally decided not to make a market investigation reference in this case and subsequently changed its mind. To echo Keynes, it did so because the facts had changed since its promised powers had not materialised. The question for the CAT was whether the law allowed it to do this.
While the CMA’s original decision not to make a reference was based on the assumption that it was about to acquire new powers that would have rendered a market investigation wasteful and unnecessary, the Government’s interminable delay in bringing forward the legislation for those powers revealed that decision to be ill-founded. As in many other areas of British life, the cumulative impact of Brexit and the Covid-19 pandemic on the machinery of Government and the increasing dysfunction within the Cabinet in the run-up to Prime Minister Boris Johnson’s resignation in July 2022 threw a bag of spanners into the works, leading amongst other things to a log-jam of delayed legislation. While the CMA’s decision to change course in light of events is understandable in this context, as Apple correctly identified, the specific statutory framework governing the conduct of market studies preceding market investigations significantly constrained the CMA’s ability to do so.
In a crisp and well-reasoned judgment, the CAT examined the CMA’s decision to make the reference in light of the specific statutory provisions governing references following a prior market study, namely sections 131A and 131B of the Enterprise Act 2002 (EA02). These sections were introduced by the Enterprise and Regulatory Reform Act 2013 (ERRA) specifically to limit the duration of market studies. In the March 2011 consultation paper entitled ‘A Competition Regime for Growth’, which preceded the ERRA, the then coalition Government noted concerns over “lengthy market studies and market investigations” delaying remedies and causing prolonged uncertainty, with phase 1 market studies lasting up to 21 months and end to end times for cases referred to a full phase 2 market investigation of up to 67 months. With this in mind, one of the proposals for consultation was the introduction of a statutory time limit to “ensure that market studies are not extended”.
In the Government’s Response to its consultation, it noted that a “majority of respondents considered that market studies that are to be referred to a phase 2 investigation should be done so promptly”. It, therefore, confirmed that “The CMA will be required to consult on a decision to make a MIR [Market Investigation Reference] within 6 months where such an outcome is being considered as a result of a market study; and will not be permitted to take longer than 12 months to make a MIR” (at page 33). This requirement was duly carried over into the legislation, in what became sections 131A and 131B EA02. The extracts from the policy documents make it clear that the asymmetry faced by the CMA when applying these sections was intentional. In other words, whereas the CMA has a full 12 months to make a phase two reference, including the initial six-month period for the decision to consult on a potential reference if the CMA is minded not to refer then it must formally decide not to make a reference within six months of the start of the study.
As the CAT noted in its judgment, the effect of these two sections (which it described as constituting a “clear code”) is that, once the CMA has published a market study notice, it has six months to decide either: (i) not to make a market investigation reference, or (ii) to consult on a proposal to make a reference. As the CAT succinctly observed, in December 2021 the CMA opted for option (i).
Although both sides acknowledged that the CMA remained able to launch ‘standalone’ market investigations without undertaking a market study in certain circumstances, once it had triggered a market study by publishing a market study notice, sections 131A and 131B set out a binding timetable. That timetable required that any notice of a proposal to launch a market investigation had to be published (and the mandatory consultation period commenced) within six months of the original market study notice, i.e., by 15 December 2021. While the CMA’s original decision not to make a reference was in time, the subsequent notice of its proposal to make a reference and to commence consultation on 10 June 2022, i.e., almost a year after the market study notice, was well out of time. Since the timetable was set by the date of the original market study notice, the decision not to make an investigation (which the CAT described as “unequivocal”) was binding “in relation to the matter specified in the notice” and could not be substituted with a later decision to make a reference after all. The CAT, therefore, concluded that the decision was ultra vires, rendering the entire market investigation void. While the position would have been entirely different if the CMA had decided that it was minded to make a reference unless it acquired regulatory powers that provided it with better options, the CAT noted pointedly that it had not taken this approach.
Following Apple’s successful demolition of the legal basis for the current market investigation, the urgent question for the CMA and all interested parties to address is what happens next. The CMA responded on the same day of the judgment with an update on the case page, in which a spokesperson noted its disappointment with the judgment. The spokesperson went on to observe that the judgment placed “material constraints on the CMA’s general ability to refer markets for in-depth investigations” and risked “substantially undermining the CMA’s efficiently and effectively [to] investigate and intervene in markets where competition is not working well”. The statement concluded with confirmation that the CMA would be “considering our options including seeking permission to appeal”. No appeal has been launched at the time of writing.
While it would be natural to assume that the next best alternative to a successful appeal would be for the CMA either to make a new standalone reference under its wider section 131 powers or to start again and publish a new market study notice, either course of action appears to be blocked. As noted above, the fact that the CMA has made a market study notice in relation to mobile browsers and cloud gaming (ie the “matter specified” in that market study notice) means that the procedure specified in sections 131A and 131B applies and a standalone reference is excluded. Although the CAT accepted that the CMA would have the vires to issue a “fresh” market study notice that built on an earlier notice, issuing one on the same matter as a previous notice (as would be the case here), in the absence of a change of circumstances, may in the CAT’s view be challengeable on other public law grounds.
Given this, the CMA will no doubt be closely scrutinising the CAT’s suggestions in the judgment itself for resolving what it accepted was an “undesirable” outcome that “without good reason” constrained the CMA’s ability to make a reference. While the CAT’s guidance is (possibly intentionally) somewhat gnomic, it appears to suggest that, by purporting to issue a decision not to make a reference that was on its face categorical but in reality was subject to the qualifications that were set out in the CMA’s Interim Report, the CMA acted unlawfully. Specifically, the CAT indicated that “In concluding that the test for making a market investigation reference (…) was met but in declining to make such a reference only in the expectation of receiving further powers on the basis of a preliminary and (as it transpired) mistaken view of the potential for intervention, it might well be said that the CMA erred in law and/or took into account immaterial considerations”.
In other words, the CMA “did not have the option to decide not to make a reference at all with a reservation entitling it to revisit that decision at a later date”. Presumably, if the decision not to refer was unlawful and hence void (or the CMA chose to treat it as such), the block on the CMA revising its conclusions created by that decision would be removed and the CMA would be able to proceed with a new market study notice and, ultimately, a new referral. Such a course of action would hardly be quick, however, and would likely be fraught with additional legal uncertainty.
Fortunately for the CMA, the bill that promises finally to grant the CMA its long-awaited regulatory powers will also repeal the six-month time limit that was the authority’s undoing in this case. In the Government’s consultation document on its latest wide-ranging reforms to the competition and consumer regimes, entitled Reforming Competition and Consumer Policy, which was published in July 2021, it proposed removing the requirement to consult on market investigation references within the first six months of a market study, on the grounds that “the current statutory process for consulting on market investigation references (…) is unduly restrictive and does not materially speed up the making of a market investigation reference”. The Government confirmed in April 2022 that it would take this forward. There could be no better confirmation than the outcome of this case that the Government’s position was justified.
Although the Government’s plans to bring forward the bill to implement this proposal were confirmed in the Queen’s Speech on 10 May 2022 and in the Autumn Statement, at the time of writing, the draft bill has still not been published. In an irony that will not have been lost to the CMA, that same bill (the Digital Markets, Competition and Consumer Bill) will finally grant it the regulatory powers that were its original basis for deciding not to make a market investigation in this case. Assuming the bill finally makes it into Parliament this year, it remains possible that the CMA will simply abandon its market investigation altogether and take forward its work under its regulatory powers, as it originally intended. If it does so, the main value of this case may be to serve as a salutary lesson of the unexpected consequences of legislating for binding timetables for competition investigations and, more widely, the importance of good government for the entire administrative ecosystem.
