Power without accountability – UK Constitutional Law Association

Clause 19 of the Digital Markets, Competition and Consumers Bill (“the DMCC Bill”) gives the Competition and Markets Authority (“CMA”) wide powers to impose conduct requirements on undertakings that the CMA has designated as having strategic market status (“SMS”) in respect of a digital activity (essentially, “big tech”).  The condition for the exercise of the power is that the CMA considers it “appropriate” to do so having regard to broadly drawn statutory objectives – the “fair dealing” objective, the “open choices” objective (essentially, enabling consumers to choose freely and easily between providers) and the “trust and transparency” objective (essentially, having the information to make properly informed decisions).  The types of conduct requirement that may be imposed are set out in clause 20: essentially, there is no statutory limit to the types of requirement that may be imposed save that (whatever they are) they must be for the purpose of securing various objectives or of preventing various widely defined types of conduct (for example “using data unfairly”).  Breach of a conduct requirement may result, if the CMA so decides, in an enforcement order that requires action to remedy the breach or payment of damages.

It is worth comparing clause 19 with the provisions of Articles 5 to 9 of the EU Digital Markets Regulation (Regulation (EU) 2022/1925).  Articles 5-7 of the EU regulation set out a long list of specific obligations imposed on “gatekeepers” (the equivalent of those enjoying SMS), with the Commission having power (under Article 9(2)) to adopt subordinate rules to “specify the measures that the gatekeeper concerned is to implement in order to effectively comply with the obligations laid down in [those Articles]”.  In the EU Regulation, therefore, the task of setting the basic conduct rules for undertakings identified as “big tech” is done in EU Parliament and Council legislation.  But, under the DMCC Bill, that task will be left to the CMA. 

The DMCC Bill’s wide delegation of rule-making power to the CMA is particularly striking given that it is in an area which is, and is likely to continue to be, of great public interest and public policy importance, as a result of the enormous economic power enjoyed by big tech and the overspill of that power into the political sphere, especially in relation to social media.

Despite that, however, the CMA – whose Board is a body of persons appointed by the Secretary of State for a term of up to five years– will have, at best, only limited political accountability for its decisions.  Parliament has no role in scrutinising the CMA’s exercise of these rule-making powers, apart from an ability, through its select committees, to subject the CMA to questioning after the event.  Although the Secretary of State can make regulations to define the edges of the CMA’s powers (for example, in adding to the list of types of enforcement order that may be made, or filling in certain elements of the definition of SMS), she has no power to direct or veto the making of a conduct requirement, or even to set parameters for the CMA’s exercise of these powers.  

Further – an aspect of clause 19 that has generated considerable debate and controversy – the only procedure that the CMA needs to go through before imposing a conduct requirement is a consultation exercise (clause 24), and the only route to challenging its decisions to impose or not to impose a conduct requirement is an application to the Competition Appeal Tribunal (“CAT”) under clause 102 for judicial review on ordinary principles.  

The combination of lack of political accountability with thin procedural requirements and limited rights of appeal (both for big tech and for anyone concerned that the CMA has not done enough to tackle big tech) is unprecedented and – in the present writer’s view – both important and deeply problematic from a constitutional and democratic perspective.


There are two possible precedents for the (in effect) legislative powers given to the CMA by clause 19.

Market investigations

Powers to investigate the conduct of businesses with monopoly or substantial market power date back to the Attlee government (the Monopolies and Restrictive Practices (Inquiry and Control) Act 1948): they were subsequently developed in the Fair Trading Act 1973 and the Enterprise Act 2002.  

The common thread running through those provisions over time is that they set up a form of public inquiry conducted by an expert body (the Monopolies Commission, then the Monopolies and Mergers Commission (“MMC”), then the Competition Commission (“CC”) and, since 2014, the CMA itself through independent panels).  That inquiry examines whether features of an industry operate against the public interest, with a final remedy (if such problems are identified) of sector-specific legislation to deal with those problems.  

What has changed over the years is that (a) the concept of what the public interest is has narrowed, so as to focus on distortions of competition that adversely affect consumer welfare (as opposed to, for example, aspects of conduct within an industry that affect employment or regional development) and (b) consistent with the greater focus on what might be seen as a technocratic exercise (analysing the extent of competition and its impact on consumer welfare, particularly in terms of prices rather than a wider public interest one), the involvement of elected politicians has steadily reduced.  


  • the 1948 Act made provision for a broad inquiry by the Monopolies Commission, started by the Board of Trade (i.e. a Cabinet minister), which resulted (if public interest detriments, widely defined, were found) in an order-making power exercised by the Board and subject to the usual Parliamentary process for scrutinising statutory instruments;
  • the 1973 Act maintained that model, save that the power to start an inquiry was also given to the Director General of Fair Trading (“DGFT”: a non-political appointment); 
  • the 2002 Act made a more fundamental change, as it removed the Secretary of State from the process, at the same time as focusing the inquiry on competition and resulting consumer welfare issues: in particular, the power to order remedies, including the power to make legislative orders, was passed to the CC (without any Parliamentary scrutiny); and
  • the Enterprise and Regulatory Reform Act 2013 consolidated the Office of Fair Trading (the successor to the DGFT) with the CC to form the CMA – though independent panels within the CMA, appointed separately from the CMA Board and not answerable to it, continue to be responsible for market investigation inquiries, after an initial enquiry phase conducted by CMA officials.  

The markets investigation regime and its predecessors have resulted in highly consequential legislation – examples being the making of orders that abolished “tied houses” (transforming the beer and pub industries) and the imposition of the Groceries Code and other trading rules on supermarkets.  

The fact that, since 2002, potentially highly consequential legislative power has been held and exercised by an unelected and politically unaccountable body – the CC and latterly the CMA – subject to no judicial scrutiny save judicial review on standard principles to the CAT, has been justifiable in constitutional and democratic terms for essentially two reasons: 

  • first, that the basis for the exercise of the powers has been defined in terms of competition and associated consumer welfare – plausibly seen as a technocratic exercise suitable to be carried out by experts; and
  • second, that the powers may be exercised only after a thorough inquiry process, conducted by an independent body, which allows all those affected significant procedural rights to make their views known. 

In contrast, the clause 19 power is not tied to any defined concept of competition or consumer welfare and cannot plausibly be regarded as a technocratic exercise.  On the contrary, the whole impetus behind the taking of these powers is a widespread concern (sometimes referred to as the “new antitrust”) that the competition/consumer welfare analysis that has dominated thinking about competition law and policy since the 1980s is inadequate to deal with the accretion of economic and political power by big tech: in particular, that competition analysis is challenged by the typical big tech model of the “free” provision of services and deep “network effects” and does not tackle the accretion of political power, for example, via control of social media.  (For an excellent and persuasive non-technical account of the “new anti-trust” and its application to big tech, see Jamie Susskind, “The Digital Republic” (2022), chapters 32-33.)

The DMCC Bill reflects that thinking: thus, the extent to which legislative restrictions on big tech are justified in the name of preventing it from “using data unfairly” (to take an example from the DMCC Bill’s clause 20 objectives) is not a technocratic exercise based on a competition/consumer welfare analysis, but a public policy judgement where different political perspectives will produce very different answers, and which involve questions of distribution of resources and power.  (Indeed, it is notable that clauses 19 and 20 do not refer to “consumers” at all, but rather to “users”.)  Rather than providing any precedent for clause 19, the history of the market investigations regime suggests that, just as the conferring of decision-making on unelected regulators accompanied, and was acceptable only because of, the narrowing of policy focus to competition/consumer welfare concerns, the corollary of widening the basis on which legislative interventions should be made ought to be accompanied by placing the relevant powers into the hands of accountable politicians rather than unelected officials. 


Another precedent might be the various licensing regimes that have operated in several utility sectors since privatisation.  Indeed, the concept of SMS as the basis for liability to the imposition of conduct requirements has echoes of telecommunications or airport regulation, where a finding of significant market power is or has been the basis for liability to a licensing regime (including price control). 

The parallel is, however, inexact.  The licensing regimes differ and have differed in various ways: but common features have been, first, an emphasis on price control and specifically on using price control to (in broad terms) replicate the results of competition, had that been achievable (a matter plausibly regarded as better suited to decision-making by technocrats rather than politicians), and, second, where broader public policy objectives have been in play (for example, water) a power is typically given to the Secretary of State to direct the regulator in various ways.  

Further, a common feature of the licensing regimes has been the provision of an appeal mechanism, typically to an independent CMA panel, that allows the licensee (and sometimes interested third parties) to challenge a regulator’s licensing decision on the basis that it is “wrong” or contains an error of fact or error of discretion.  In such appeals, CMA panels have emphasised that such an appeal is not a re-hearing but rather a focused examination of specific grounds of appeal, taking the regulator’s decision as a starting point and not interfering merely because the panel might have taken a different approach (see, for example, §§3.14ff of the CMA’s recent decision on the appeals from the Civil Aviation Authority’s setting of price controls on Heathrow Airport).  However, the appeal mechanism does mean that, if an interested party wants it, a “second pair of eyes”, independent of the regulator, is cast over the decision before it is finalised, and policy mistakes or dubious analyses can be corrected even if those mistakes fall short of the irrationality standard (which, of course, is often set particularly high in areas of economic policy).  


The powers given to the CMA under clause 19 are therefore unprecedented in their combination of: 

  • wide rule-making discretion in an area of intense and legitimate public controversy and importance (and which cannot be characterised as matters of technocratic judgment and expertise) – rule-making power that under the EU equivalent legislation has been exercised by the EU legislature and set out in the governing Regulation, with only detailed implementation left to the Commission;
  • absence of political accountability or Parliamentary oversight; and
  • lack of any effective appeal mechanism or independent scrutiny (either on behalf of big tech or of interested parties who think that the CMA has not been tough enough).

The combination of decisions that are not mainly about technical matters but about important matters of public policy involving questions of value and distribution, combined with the absence of precise parameters set by elected politicians for the exercise of those powers, is not a promising basis for conferral of extensive powers on an independent regulator: as Stuart Hudson argues, proponents of independent regulation would generally agree that it is appropriate only where the decisions being delegated are ones mainly of technical expertise rather than social policy issues, and where elected politicians set clear parameters.

What is to be done?

As pointed out at the start of this post, the debate on clause 19 has largely focused on appeal mechanisms (judicial review or merits appeal to the CAT).  The above analysis suggests, however, that that dispute focuses on the wrong question.  The prior, and even more important, question is whether the clause 19 power is suitable for exercise by the CMA at all – which, for the reasons set out above, it is not.

One option for addressing that would be to follow the EU lead and set out in the DMCC Bill a detailed set of rules that SMS undertakings (big tech) would be bound by.  Another, more flexible, approach would be for the Secretary of State to have power by subordinate legislation to set out either detailed rules or strong parameters for setting such rules, for which legislation or parameters the Secretary of State would be accountable to Parliament.  

If, however, that battle has been lost, the lack of any independent review of CMA decisions is still a matter that needs addressing, as a “second best”.  But – contrary to the arguments being made by big tech – that improvement does not consist of creating some form of merits appeal to the CAT: that is because the CAT – or any court – is simply not institutionally competent to take what are, ultimately, important decisions of public policy involving value and distribution.   

A better proposal is that there should be recourse (either by big tech or by users or competitors who believe that the CMA is not being tough enough on big tech) to an independent second pair of eyes, in the form of an independent CMA panel.  The panel  could receive appeals from decisions taken by the CMA Board and consider whether the Board has made errors of fact or discretion, and do so on the same basis as CMA panels currently act as an appellate body in certain licensing decisions (see above).  Since the CMA’s panels stand outside the CMA’s decision-making and are independent of the CMA Board (see above), there is no difficulty in principle with providing for a statutory appeal to a CMA panel, even if in some respects it looks like an “internal” appeal.  It is however important to distinguish this proposal from assurances that the CMA Board will, as part of the CMA’s internal decision-making processes, review every proposed clause 19 decision: such an internal review (which those affected may not even know about, and will have no right to make representations to) is essentially an administrative process and is not a substitute for an appeal.  

The proposal for an appeal to an independent panel – put forward by the Joint Working Party of the UK Bars and Law Societies on Competition Law, of which the present writer is co-chair – would at least make it easier to deal with policy errors and analytical mistakes by the CMA without having to go through the process of judicial review and the inherent difficulty of persuading a court that such errors meet the standard of irrationality.  And it would provide some protection against the dangers of confirmation bias, regulatory capture, or overreach. 

In thinking about improvements to the Bill, it is important to respect the point that – to be effective – regulation of big tech has to be fleet of foot: in such a fast-moving industry, the price of regulatory delay is often irreversible harm.  But there is no reason why allocating the basic policy responsibility to the Secretary of State rather than to the CMA, or providing for her to set clear parameters for CMA intervention, should slow down decision-making.  Nor, in the case of the “second best” proposal of an appeal to an independent panel, would the proposal necessarily slow the process down.  Regulatory appeals to the CMA panels in complex matters typically take no more than six months: and to the extent that they avoid a judicial review challenge (which even if misplaced leads to months of delay) by correcting procedural errors and errors that could meet the irrationality standard, no time is lost. Moreover, an appeal panel can re-take a decision after correcting any identified errors rather than being confined to remittal and resulting further delay (which is usually all that a court can do when it finds that a ground for judicial review has been made out).  

The DMCC Bill returns to Parliament shortly.  It has almost completed its passage through the Commons.  But the House of Lords needs to consider very carefully not just the economic and industry implications, but also the constitutional implications, of this part of the Bill.

George Peretz KC is a barrister practising in public law at Monckton Chambers.

(Suggested citation: G. Peretz, ‘Clause 19 of the Digital Markets, Competition and Consumers Bill: Power without accountability’, U.K. Const. L. Blog (15 November 2023) (available at https://ukconstitutionallaw.org/))

Leave a Reply

Your email address will not be published. Required fields are marked *