Standard of review for the CMA's DMU: will JR actually be quicker and more effective?

The Digital Markets, Competition and Consumers Bill (DMCC Bill) provides that judicial review (JR) is the review standard applicable to decisions of the CMA’s Digital Markets Unit (DMU) under the new Strategic Market Status (SMS) regime.¹

This approach restricts both the grounds on which a decision can be challenged and the remedial action that the Competition Appeal Tribunal (CAT) can take if it finds that the DMU has gone wrong: the CAT will be able to overrule the DMU only on grounds such as procedural fairness, error of law, and irrationality, and not on the basis that it considers the decision to be substantively wrong and would itself have reached a different conclusion. In terms of remedial action, under the JR standard proposed in the DMCC, regardless of how easily the CAT might be able to correct the error of law, the failure of due process, or the irrationality of decision-making logic, it has no freedom to substitute its own decision for that of the DMU and settle the matter there and then (or as the President of the CAT has put it, to “rescue” the decision²). Its only remedy is to send the decision back to be made again.

The DMCC is currently going through committee stage in the UK Parliament and the standard of review of DMU decisions has been the subject of much debate. One key issue in the debate is how “flexible” JR may be in terms of contesting decisions and the arguable conceptual overlap between merits and JR (e.g., a “merits” ground might instead be framed as a failure to take relevant considerations into account in a JR context). Proponents of JR cite this flexibility as a reason why it should be deemed fit for purpose for review of DMU decisions and argue that it is not too “light touch” notwithstanding what is, on any view, an enviable quantity of agency discretion to exercise new powers. 

We have written elsewhere³ about why, if the intention of the government is for such “flexibility” to exist, there is no substitute for making the relevant “enhanced” JR intended to apply clear on the face of the legislation. The purpose of this article is not to engage further in that debate, but to examine whether two assertions made by proponents of the JR standard are correct:

1. That appeals on a JR basis are faster than appeals on a merits basis; and
2. That making DMU decisions subject to merits review would stifle the efficacy of the SMS regime.

In this paper, we argue neither of these propositions are accurate, for the reasons we will explain.

Is JR faster than merits?

In response to the consultation on the DMU regime, the government concluded that JR was the most appropriate standard of review. Part of its reasoning was that JR procedures would be quicker than a merits-based system of appeals, and hence would support “the delivery of robust outcomes at pace”.⁴

This is an oversimplification. It is critical that any assessment of timing considers not just the length of the judicial process, but how long it takes for the matter finally to be resolved. This needs to take account of the administrative decision, the judicial process, and the time it takes to reach a final resolution following a successful appeal. 

JR vs merits timing considerations: zeroing in on relevant comparators

At the front end, we would not expect a lower standard of review to lead to faster administrative decisions: it should be assumed that any responsible regulator will wish to take lawful and fully reasoned decisions whatever standard of review applies to them. Put simply, the DMU will not “cut corners” because it “only faces JR” at the end. Within statutory deadlines, that process will take as long as it takes, particularly in the case of a new regulator operating in emerging and fast-changing markets.

In terms of the time to reach a final resolution however, while it is true that historically, appeals of CMA decisions on a JR standard have had a faster judicial process than those on a merits standard, they have not been faster end-to-end. 

Other commentators have argued that CAT reviews into mergers and market investigations on JR principles are typically quicker than reviews of Competition Act 1998 (CA98) cases on the merits.⁵ We do not take issue with the averages computed from (a relatively small sample of) Enterprise Act 2002 (EA02) JR cases and CA98 merits appeals, where they note that CAT reviews into mergers and market investigation take in the region of six months and CA98 cases have taken from 11 months to two years. We would however query the extrapolations and analogies involved in using these examples for the DMU context. 

First, as a matter of principle, the speed of any appeal process will inevitably depend on the court’s determination of questions of urgency and prioritisation, and availability of judicial and court resources. For this reason, the timeline for review of merger control decisions is not a suitable comparator. The merger appeal deadline is one month (vs. two for CA98 appeals), and once lodged, while the timetable for such reviews has generally been shorter, it is because the CAT will typically case manage to hear merger appeals more urgently, given they relate to pending or completed corporate transactions: at one extreme, the CAT heard the procedural fairness challenge in Sainsbury’s vs CMA essentially one working day after the appeal notice was lodged, because the matter was urgent.⁶

A closer analogue for the kinds of decisions the DMU will be taking is the market investigation regime, under which CMA decisions are subject to JR, or indeed CA98 decisions that, while subject to merits review, have in rare cases involved remittal of the decision to the CMA. Depending on its caseload, the CAT may seek to prioritise review of DMU decisions ahead of CA98 cases (and other matters), or it may not. This is and will remain a question of CAT case management discretion. 

What the numbers actually say

We have analysed the timetables in MIR appeals, and CA98 appeals – in the case of the MIR regime since its inception in 2003, and in the case of CA98 appeals, since the creation of the CMA in 2014. 

This analysis – summarised in the table below and in more detail in the Annex – demonstrates that:

• Looking only at the judicial part of the appeal process (i.e., the time from the authority decision until the court decision), cases appealed on a JR standard are determined more swiftly than those on a merits basis – in an average of 296 versus 354 days (i.e. around two months faster).
• However, when an appeal is successful and the relevant decision is remitted to the CMA (or predecessor authority), the judicial part of the process is generally significantly shorter than the subsequent remittal.
• When considering the length of the overall process, a comparison of JR and merits on normal cases shows that merits is far quicker (333 days versus 541 days). Even an inclusion of Pfizer – an outlier case – shows that the average length of the overall process in merits cases is around one month longer.
• A remittal therefore lengthens the process significantly – on average in markets cases by ten months (and longer if the Pfizer remittal is included).

Average process lengths for markets cases and CA98 appeals

The outlier Pfizer remittal case

The most cited example of lengthy merits review under CA98 is Flynn Pharma and Pfizer, a case which is rightly held up as the “poster child” of a lengthy process. However, the reasons for the lengthy process were very case-specific.

In that case, the CAT overturned a CMA infringement finding and remitted the case back to the CMA. While the CAT had the power to substitute its own decision for that of the CMA, it did not have the evidence before it to enable it to do so based on its interpretation of the law and thus decided to remit the case to the CMA for the missing evidence to be gathered⁷ and assessed against the correct legal standard. In the counterfactual where a JR standard had applied to that case, it is not unreasonable to assume (indeed the inference seems extremely strong) that one ground of challenge on classic JR principles would have been that the CMA had failed to take relevant considerations into account – namely, the critical missing evidence. 

The timeline was further extended by the fact the remittal only began after an unsuccessful further appeal by the CMA of the CAT’s decision in the Court of Appeal (the remittal has since concluded and is subject to a pending appeal). Critically, CA98 investigations are not subject to any statutory timeframes (unlike most DMU decisions) and the remittal process – which is the most significant driver of the lengthy timeline – took 791 days. But JR remittals of markets and merger cases have also not been subject to judicially-imposed deadlines.⁸ Rather, the CMA sets its own administrative deadlines. 

The Pfizer case is an outlier case that “biases” the sample of merits cases. But to the extent that Pfizer outlier timing is relevant at all to the JR vs. merits standard of review debate for DMU, it is best thought of as a wash: it is an outlier than would apply equally to a JR setting as a merits setting. This is because a remittal is a remittal: there is no reason in principle that we can see why the CMA’s 791-day administrative process would have been faster if the remittal had been based on a JR ruling not a merits ruling. 

CAT remedial flexibility can increase end-to-end efficiency

In terms of speed and efficiency, there is no doubt that a typical merits appeal will take longer in terms of hearing days than a JR (because it involves fact and expert witness examination and cross-examination, etc.) and also in terms of preparation time for the parties. But this is a distinction in the realms of weeks to several months at most. It is not a significant driver of end-to-end timing compared to remittal length. 

Conversely, there is efficiency, where the case permits, in giving the CAT the flexibility to resolve the matter itself in the judicial procedure where the interests of efficient justice point to this outcome. This is an efficiency upside of merits review that does not apply to JR, however flexible it is asserted to be with respect to the intensity of review. If there were any doubt, is notable that in a recent subsequent merits appeal under CA98, the CMA’s counsel has explicitly requested the CAT not to remit the case to the CMA. 

Would merits review compromise the efficacy of the regime?

It has been argued that the DMCC is subject to more urgent timing imperatives than CA98 enforcement, and that introduction of a merits review will impact the efficacy of the regime. It is hard to see how this would be the case. Under the DMCC as drafted, an appeal does not suspend the application of a DMU direction or order. The DMCC provides that an undertaking will be expected to comply with a decision until the outcome of an appeal is known, unless the CAT makes an interim order to the contrary. The only exception is in relation to penalties, where the obligation to pay is suspended when a challenge is brought against the penalty decision.⁹

This has significant implications, which are best illustrated through a practical example. Consider an appeal of a DMU decision on the content of an SMS firm’s Code of Conduct. Unless the CAT makes a contrary ruling, the conduct requirements will continue to apply to the SMS firm while the appeal is pending – the SMS firm will be legally obliged to follow them and risks fines for breaches if it does not. 

In a scenario where the DMU decision is upheld by the CAT, the standard applied in the appeal makes no difference to the operation of the SMS regime: the conduct requirements apply while the appeal is pending, and continue to apply when the appeal is determined. As noted above, the judicial process on a merits standard could be expected to take a few months longer than an appeal on JR only (our analysis suggests merits appeals take – on average – 58 days more than JR appeals), but in terms of whether the SMS firm is subject to conduct requirements, there is no difference between the two.

By contrast, in a scenario where the DMU decision is quashed by the CAT, the standard of review could make a significant difference. In a merits review, where sufficient evidence is before the CAT to allow it to reach what it considers an appropriate outcome, it can do so: the conduct requirement therefore applies while the appeal is pending, and the hypothetical amended conduct requirement will apply from the date of the decision (or whatever other date is specified by the CAT). By contrast, when the CAT considers a decision on JR and finds a problem, as noted, a remittal is the only possible remedy. The CAT cannot substitute its own decision for that of the CMA, even when it identifies a minor procedural error or holds all the evidence (unlike in Pfizer) to correct the decision. In this case, the impugned conduct requirement would cease to apply and the DMU would need to go back and run its process again.

Conclusion

We have previously explained why we believe the nature of the decisions taken by the DMU should be subject to a higher level of judicial scrutiny than traditional JR, and that by far the best way to achieve that is to make the intention clear on the face of the DMCC itself. 

This paper has dealt with the much narrower factual question of whether JR appeals under the DMU would in fact be quicker than if a higher appeal standard was specified, and whether – given the broader provisions of the DMCC – a higher standard would affect the efficacy of the SMS regime. 

We have concluded that, on closer inspection, the better view is “no” on both counts: relative to JR, merits would not slow down the DMU, nor make it less effective.

Annex        Timing of Judicial Review and Merits Review appeals to the Competition Appeal Tribunal¹⁰

¹ There is a minor exception for certain procedural penalties in relation to the merger reporting obligation, for which the CAT is able to substitute its own decision in accordance with the standard under section 114 Enterprise Act 2003.

² Speech of Sir Marcus Smith at the CMA Judicial Review Conference, 18 April 2023, at para 43: DISCRETION (catribunal.org.uk). 

³ Sir Jonathan Jones KC and Verity Egerton-Doyle, The New Digital Markets Unit: What test should apply to appeals against its decisions? Competition Law Insight (March 2023). 

⁴ Consultation response, at para 93.

⁵ Tom Smith and David Gallager, In Defense of Judicial Review: The established UK appeal standard is the best approach for a dynamic digital economy, page 4 (May 2023).

⁶ Sainsbury’s v. CMA [2019] CAT 1. 

⁷ This being the candidate comparator product for purposes of the abuse of dominance assessment. 

⁸ This can be contrasted with cases like Floe Telecom v Ofcom, where the CAT ordered that Ofcom re-investigate the matter with a view of issuing a non-infringement decision or a statement of objections within five months of the order. 

⁹ Clause 101, DMCC.

¹⁰ This table excludes cases that were subject to referrals to the European Court of Justice, given these do not have any read across post-Brexit.

¹¹ Time between receipt of applicant’s application to the CAT and the CAT’s final decision.

¹² Time between remittal by the CAT and the CMA’s Final Report following remittal, unless otherwise stated.