Relaxing whilst doing Competition Law is not an Oxymoron

Be careful what you wish for: why discretion to fine-tune digital markets may not be in the interest of authorities (or the public interest at large)

with 5 comments

Thibault Schrepel has been kind enough to invite me to write a piece for his Concurrentialiste (see here). It was a good chance to explore some of the themes that are central to my ongoing research projects and to share my thoughts on some of the proposals to regulate digital markets that are doing the rounds. My message? Regulatory design is key, and there are instances in which giving discretion to an authority might not be in the latter’s interest (or the public interest at large). I reproduce my post below, and very much look forward to your comments. Thanks again for the invitation, Thibault!

Emerging regimes for the regulation of digital markets share a common philosophy. They are grounded on the belief that, if authorities enjoyed more discretion and, in the same vein, if the constraints to which they are subject were reduced, they would be in a position to intervene fast, and adopt the sort of far-reaching remedies which, the argument goes, digital markets demand. In this sense, the new instruments represent, first and foremost, a departure from the limits of competition law systems. Establishing the anticompetitive object or effect of a practice and ensuring that the theories underpinning intervention reflect mainstream views are seen, from this perspective, as a burden that may dangerously delay much-needed action.

There are reasons to question whether this philosophy will deliver on its promises. Granting discretion to an authority to fine-tune digital markets whenever it deems it necessary does not address the fundamental challenges raised by the measures proposed. The phenomenal difficulty that comes with the latter does not relate to the lack of discretion, but to the very nature of the intervention expected from authorities. Redesigning products (as in Google Shopping), altering business models (as in Android) and re-allocating rents across the supply chain (is a 30% commission too much or just about appropriate?) are complex tasks, prone to errors, which are not made any easier by doing away with the need to show the anticompetitive object and/or effect of a practice (or the need to weigh such effects against any pro-competitive gains). The lengthy aftermath of the Google saga, and the aura of limbo and uncertainty that surrounds the remedies implemented in those cases, provides eloquent evidence in this sense.

This piece, however, makes a different point. It is submitted that granting discretion to an authority to fine-tune digital markets does not necessarily do it any favours. On the contrary. Depending on the design of the regime, discretion may in fact weaken it vis-à-vis stakeholders. One of the most precious powers of an administrative authority (and all of us, really) is the ability to say no. The ability, in other words, to see off pressure from firms and governments and take a decisive stand on a particular issue (or not take a stand at all). Paradoxically, the problem with discretion is that it empowers the authority to reach virtually any decision it desires from a policy-making standpoint. Once this impression is created, it may be difficult for an authority to justify why it favours certain outcomes and why it does not opt for the maximalist (or minimalist) positions allowed under the regime.

It is not difficult, on the other hand, to anticipate the behaviour of stakeholders potentially benefitting from regulatory intervention. It is natural for them to try and secure the outcomes that, within the boundaries of what is possible, best serve their interest. Where, for instance, a break-up of Big Tech firms is allowed under the regime, there is no downside, from their perspective, in pushing for structural action. An authority that is serious about meeting the objectives set in legislation, stakeholders would claim, should seek the most effective remedy, and the one that genuinely preserves competition on adjacent markets. If the regime requires intervention, and allows for the break-up of firms, all the pressure lies with the authority.

The aftermath of Google Shopping and Android gives a flavour of how these stakeholders may react to authorities’ newly found discretion. In the first of these cases, the European Commission imposed a duty of neutrality regarding the way in which the relevant search results are displayed. Ostensibly, the firm’s chosen path complies with the obligation, even if it just one of many ways to align its behaviour with the decision’s requirements. Rivals and their advisers, however, have been repeatedly pushing – again, as one would logically expect – for interpretations of the non-discrimination duty that are more favourable to their own interests. These attempts have little chance of success (and so the European Commission has signalled) because of the very constraints that come from the competition law system. The point of remedies under Regulation 1/2003 is to bring the infringement to an end, not to reshape markets to optimise rivals’ chances of success. Now think of how different the picture would be – and how much more powerful rivals would – if this constraint disappeared and every approach – from the minimalist to the maximalist – were in theory possible and left in the hands of the authority.

If you want another example of an area of the law where the European Commission enjoys broad discretion and is, as a result, subject to pressure to reach certain outcomes, think of State aid. Under Article 107(3) TFEU, the authority is empowered to define, as it sees fit, the instances in which subsidies and similar measures are in the interest of the EU as a whole. And it is not necessary to explain at any length why lobbying by EU Member States awarding aid can be just as formidable as the calls for action coming from dozens of firms operating in the digital sphere.

Of particular interest are the insights to draw from the way in which the European Commission has chosen to exercise its discretion in the field. Years of enforcement have taught the authority that it is unlikely to withstand pressure from EU Member States with a stake in a particular measure. It just does not trust itself. In this sense, it knows that its discretion is most powerful and best used when it is constrained ex ante and thus leaves minimum (if any) scope for manoeuvre in individual cases. Just like Ulysses when he reached the Land of the Sirens, the European Commission has tied its hands tightly to ensure that its decisions are in the public interest (and it is able to say no).

If you have never done so, take a look at State aid instruments implementing Article 107(3) TFEU – say for instance the Guidelines on regional aid, or the General Block Exemption Regulation. The most apparent feature of most of this body of law is its infinite dullness: pages and pages of figures detailing to the millimetre the intensity of the aid that is allowed. Beneath the dullness, however, lies the fundamental lesson to draw for the future regulation of digital markets: once the European Commission is freed from the constraints coming from the EU competition law system, it will have to devise a set of sound, effective and meaningful constraints to make sure it acts in the public interest. If it does not do so, it risks losing control of policy-making altogether.

Written by Pablo Ibanez Colomo

26 January 2021 at 3:22 pm

Posted in Uncategorized

5 Responses

Subscribe to comments with RSS.

  1. As always, thought-provoking material from Pablo! A healthy reminder of the centrality of the trade-off between rules and discretion in the design of public policy, dating back at least to the 1977 seminal paper by Kydland and Prescott in the Journal of Political Economy. I agree with the general point made. I would add (or rather, reformulate) that any reasonable discussion of the current attempts at relaxing constraints on public intervention on markets should not view legal prescriptions as mere obstacles or constraints but ask why they were put in place to start with. Very often, legal rules have a rationale or an intended goal that is not altogether meaningless. That does not mean that, on balance, the costs they create might not overwhelm their intended benefits but, at least, the discussion should factor in those benefits! (There are exceptions: we all know legal provisions which have long lost their raison d’être and make for pure costs.)
    I would like to make two remarks.
    1/ The ability of policy-makers to commit through legal instruments is not perfect. State aid regulations and guidelines are indeed very detailed and dull (Pablo’s words!) but, when the financial crisis erupted in 2008, it took only a few weeks for the Commission to issue revised guidelines basically allowing the bailouts of financial institutions which Member-States were contemplating. Those new guidelines contained all sorts of reasonable and useful reservations and residual constraints but, eventually, they did serve the purpose of restoring discretion at a time when short-term pressure was very high. (This is not to say that this course of action was wrong; it may well have been a smart policy response; the point is that it is always possible for a legislator to undo its own commitments unless it is bound by some superior, slow-evolving norm.)
    2/ When it comes to the current digital markets agenda, I think that, for didactic purposes, Pablo may be exaggerating the amount of discretion that the Commission proposes to give themselves. The proposed DMA does not allow the Commission to impose any obligation on any firm. Firms would have first to be designated as gatekeepers for some core online services, in accordance with a detailed procedure, a substantive standard and some observable thresholds. Then, they will have to comply with a modifiable but exhaustive list of obligations.
    The proposed act does not leave it to the discretion of the Commission to break-up firms either, as in Pablo’s stunning example. This would come only after what is called ‘systematic’ non-compliance and a market investigation, and any structural remedy would have to pass the test of indispensability.
    Surely, with new powers would indeed come new responsibilities for the Commission. It may well face commitment problems, but the proposed legislation may contain enough bells and whistles to avoid the most blatant ones. Certainly, an informed discussion can take place about that.
    Yet, it is true that, the minute a new form of public intervention is contemplated, affected economic agents (here, potential gatekeepers and their rivals) will have an incentive to invest vast amounts of resources to affect the legislative process and discuss those bells and whistles. Moreover, once the act is adopted, they will have the incentive to continue investing millions in tilting regulatory proceedings one way or the other. Any high-stake public decision gives rise to rent-seeking efforts which, by themselves, do not increase GDP in the direction policy-makers wish.

    Cedric Argenton

    27 January 2021 at 9:30 am

    • Thanks a million for the thoughtful comment, Cedric! I look forward to discussing these matters in the near future

      Pablo Ibanez Colomo

      27 January 2021 at 9:59 am

    • Dear Cedric, I tend to agree with Pablo. As DMA Art. 10 suggests, the scope of “fairness” and “contestability” are quite broad and can be said to include a wide range of diverging public policy objectives from data protection, consumer protection, exploitative behaviour, exclusionary behaviour, and even simple product design decisions that may hinder interoperability, and without regard for possible consumer welfare justifications or the protection of legitimate business interest.

      The lack of a normative reference point (e.g. based on competition law principles) was also criticised by the authors of the German Competition Law 4.0 Report (Schallbruch, Schweitzer and Wambach) in a recent piece published by FAZ (“Europa stutzt die Datenmacht der Digitalkonzerne”, available here: )


      28 January 2021 at 2:00 pm

  2. Pablo,

    Always interesting to read your thoughts! Here, though, I think the issue is what is meant by a remedy. We are not looking at compensation- which is for those harmed and they may be able to obtain redress in damages through the courts.

    We are now looking at the role of the public authority in the enforcement of the rule of law for the benefit of all in society. The public interest is in having a law that operates to affect behavior where the old law has demonstrably failed. The remedy needed is for enduring harm to the process of dynamic competition; for the day we never saw because it was monopolized. We are all harmed by abuse of dominance whether from the raising of monopoly rents or the sharp elbowed exclusion of competitors from sight or reduction in innovation or the loss of jobs and opportunities in our societies.

    What some have called the refeudalization of society is probably over the top but if it exists it is at least in part caused by monopolization.

    The first thing we need to reject is Easterbook’s famous type I and Type II errors. framing the issue in that way ignores the fact that law is intended to be normative. Law is not economics – it is about ensuring that all comply- and that does not mean avoiding the enforcement of the law for fear that the world would possibly be better off where the law is not enforced. Law needs enforcement without fear or favor because that drives confidence in law- whether for economic ends or because it shapes people’s individual and collective business behavior.

    Another issue is the argument “first do no harm” because the market may be self righting. If it is or isn’t is something that needs to be decided before deciding on the remedy. Since the new legislation is being brought in because the market is not self righting this needs to be appreciated in enforcing it.

    Laws are not neutral when it comes to standards of behavior – some things are unacceptable, others are criminal, many are ok. When a business breaks the law a question then arises for the public authority about how to remedy the position and ensure compliance with that law in the future; both for the party in breach and the system as a whole. Depending on how the law is written a component of intent may be relevant to the sanction that is imposed. In my view that would be worth including here. Willful breach of the law is the basis on which punitive damages are allowed to be awarded in many courts- provided that the public authority has not already addressed the issue inn its fines and remedies. if he fines do not strip the wrongdoer of the benefits of wrong doing then it remains open to the courts to award damages that do so. Otherwise we do not have a law worthy of the name.

    Here, the remedies need to allow the authority to take that into account and to therefore strip the wrongdoer of the fruits of their wrongdoing; and thereby encourage compliance with the law by others that may be tempted to transgress in similar ways. As currently written it may not go far enough and may need to be changed.

    How do you evaluate and address the harms done in lost opportunity, lost life chances, lost businesses, lost innovation? The Texas vs Google case is interesting here. As Google is accused of willfully manipulating and misleading activities – then it has to be dealt with as a willfully misleading manipulator of personal and commercial data. Texas is seeking jury trail and damages in the public interest.

    There are many difficult issues. Monopoly profits are ill gotten gains. How is society compensated for them? Why should Google be allowed to keep them? How could they be distributed? Given they are tainted with the illegality of trading illegally, and doing so over a generation, can they be put to one side for the benefit of the current generation or should they also benefit the generations whose horizons have been limited by foreclosure and blight?

    Where we are dealing with the issue of data mining and misuse of personal data for commercial gain or commercial data for personal gain there are also issues of the extent to which- for suppliers in the chain – the gain is one that involves gain of commercial benefit that is enduring and can be returned. Those may more properly be mattes for the courts but maybe, in assessing what should be in the normative law, some account needs to be taken of the private system of enforcement that now exists- doesn’t it?.

    Putting things right under a new law should not be limited by the lack of imagination with which the old law has been enforced – or maybe you disagree?

    Tim Cowen

    29 January 2021 at 7:44 pm

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: