Blog //
01st June 2019

Online platforms and vertical relations online platforms and vertical relations

When Claus-Dieter Ehlermann proposed the concept of these gatherings in Fiesole some 25 years ago, as “a European rival to Barry Hawk’s Fordham”, I could not have expected that the events would have had such longevity and such an impact on European law, or that I would have switched from being a troublemaking advocate to being a ruminant judge.  I realise that judicial oversight in European competition cases confronts a number of problems which are easier to identify from the bench than from the bar.  I want to recognise some challenges which are particularly relevant to the digital economy, and to speculate about possible evolutions.

There are two EU courts, within a single institution, in Luxembourg.   The Court of Justice has made major contributions to the law governing data collection, institutional oversights, transatlantic regulation of data flows, the right to be forgotten and others.[2] European law and the Court’s interpretation thereof have helped shape how the digital world works.  The rules on the custody and transfer of personal data, the status of new forms of business as conducted by Uber or AirBnB have been considered.  I suggest that the appellate function of the EU courts has contributed less to the development of competition rules in this sector, at least so far. 

The point of departure

A competition decision by the European Commission is taken as an administrative act.  It is formally adopted by a college of 28 Commissioners, each being formerly a prominent political leading figure, in his or her native country, of whom the competition Commissioner has knowledge of the file.  While in previous decades unchecked lobbying of the responsible Commissioner was routine, modern practice is much more structured and focused on issues of principle.  Any fines levied go into the budget of the EU.

The procedures are administrative, not judicial.  There is no “prosecution” of an “accused” before a “judge” in the process.  As the late director John Ferry used to remind people, the hearing is not to be conceived of as a trial. Today there is a hearing chaired by the Hearing Officer and attended by representatives of the Member States.  The Hearing Officer, though senior and well respected, is not a decision maker.  The entire process is not a bilateral debate at the end of which a decision-maker says proven or not proven.  The Hearing Officer’s role is to preside over an attempt by the accused enterprise to change officials’ minds; but is not to decide on factual controversies or legal debates.[3] 

The same officials may have been involved from the start of the case until the end, which means that the investigation of the complaint, the gathering of the facts and the framing of the accusation, the hearing of the oral argument and the drafting of the condemnation may well be entrusted to the same officials.

There has been an immense amount of academic writing in which practitioners and teachers and enforcers battle over the familiar ground of fairness, with a dose of ECHR law.[4] Wouter Wils has written exhaustively and trenchantly in support of the current regime’s fairness and legality.[5]  The propriety, fairness, wisdom, legality and desirability of this decision-making regime have been repeatedly challenged and repeatedly upheld in Luxembourg.  The institution is an administration and not a judicial body. 

National courts are of course co-competent to apply Articles 101 and 102, and regularly put questions to the Court of Justice about competition matters.  Some very important competition law principles have emerged from these references: Oscar Bronner,[6] Tournier,[7] Courage v Crehan[8] and many others. These preliminary reference cases offer the Court of Justice the chance to address big abstract principles. The General Court when hearing appeals has a rather different challenge, that of judicial review of administrative action.

Although it has regularly prevailed in court against challenges based on due process, the Commission is sensitive to the criticisms about procedures.  In the early 1980s IBM case[9], it recruited a retired director, Conor Maguire, to preside over the hearing as relations between the competent director and IBM had become too acrimoniously feisty.  Some years on, the role of the Hearing Officer had greatly expanded as the procedural intricacies of competition cases became more prominent.  That trend has continued.  Access to the file, confidentiality and other due process matters were the subject of quite elaborate rule-making.  The remit of the Hearing Officer was amended and formalised.  Criticisms and internal suggestions have been made over the years by holders of the office like John Temple Lang and Karen Williams, as well as by practitioners.

While the Commission prevails in the majority of competition appeals, it takes seriously every judicial setback. After three celebrated reverses in merger cases (Tetra Laval,[10] Airtours,[11] Schneider[12]) a much more prudent approach in merger matters was announced.  “Devil’s advocate” panels were established, so that the proponents of an enforcement measure must try to persuade their colleagues / superiors / expert advisors that action would be lawful and merited.  In addition, the Legal Service of the Commission is a traditionally powerful brake on over-enthusiasm.  Peer review is another means of internally testing a new case.  These are genuinely significant initiatives.   They are however invisible to parties on the outside.  The accused party will probably feel most aggrieved that its excellent arguments were not given proper weight.  The officials will probably feel that the arguments have been made and understood and rejected several times already.  Both sentiments can be eloquently justified.

But it remains the case that a company defending itself against a charge of infringement in Brussels is likely to feel more dissatisfied when emerging from a hearing than one in London or Berlin.  That doesn’t mean that the procedures are unlawful or lawful. 

An administrative process

We need to look at the process of testing an enforcement decision as a whole: it is not by studying whether one stage was imperfect but by looking at the entirety of the formal steps leading to a final judicial determination that one can reach a useful view on the fairness and adequacy and soundness of the actions of the administration.

It is a fact that administrative enforcement of the competition rules has become the norm in much of Europe.  There is no doubt that competition law is regarded as a necessary protection against collusions, market failures and economic thuggery.  Competition law could in earlier days serve as a means of thwarting robber barons in a semi-penal process, where huge fines reflect the stigma of wrong-doing which the decision attributes.  Some of the earliest cases saw moral as opposed to economic sentiments being given prominence.  The first Roche Vitamins case was heavily coloured by a sense of wrongdoing, exacerbated by the extraordinary story of Stanley Adams.  Theories have evolved: there was an early very high concern with the use of competition law to encourage market integration by challenging contractual obstacles to cross border trade.  Rebates, economic effects and technology licensing have each evolved in terms of economic and competition theory.

Modern competition law controversies can be very sophisticated, involving a choice between closely balanced alternative market definitions, and consideration of whether a lawful choice by a large player should be prohibited to serve the greater good. Decisions consider why or how the dominant player refused to deal, or stopped dealing, or chose to upgrade its product or tweaked its business model to respond to a rival’s emergence.  To what extent may a dominant player run its business to favour its own business model, partners or objective?

Judicial scrutiny of administrative action can give rise to vigorous debate.  I have written that the intensity of judicial review by the Luxembourg EU courts has been variable.  While in some cases, going back as far as Geitling[13] (joined cases 36, 37, 38 and 40/59) the court has been highly interventionist, rewriting the Commission’s remedy comprehensively, in many cases the Court stated that it would in essence respect the Commission’s findings in a technical field. A previous chapter written for this series of workshops[14]  reviewed the progress of “light judicial review” from Consten/Grundig[15] to Remia Nutricia[16] and Microsoft[17] while not leaving out the judicial rigour in Woodpulp[18].

An administration needs judicial oversight in order to be legitimate, and to avoid arbitrary action. The General Court was created to deliver a more appropriate level of appellate review to face heavy decisions in the fields of antidumping and competition.  Its remit to handle appeals against the acts of the European institutions and agencies covers many subjects; asset  freezes of those deemed a security risk; trademark and plant variety decisions; state aids; financial regulations and competition;  as well as access to documents; petitions; staff claims; contractual disputes involving  EU institutions and private parties.  Competition cases were a large part of the court’s diet in the early days of the CFI, but are less numerous today.  The court is naturally concerned to render judgments which are timely, and has taken care of the backlog of arrears of ten years ago.   

Lapse of time

Hamlet listed “the law’s delays”[19] as one of the aspects of the human condition which could make a sensitive mind contemplate self-harm.  The EU courts are faced by considerable challenges while proceeding with all diligence.  Moreover, the market place develops faster and faster.  Delay is especially relevant to the application of the competition rules.  There is a real risk that the enforcer addresses past conduct and that the court’s contribution will relate to almost irrelevant (though not forgotten) conduct.  And these concerns seem especially relevant to the digital world. 

I suggest that the general impression that competition cases take too long is attributable to several factors: the speed of events in the real world; the proper care of the institution to do a careful job; the court’s inevitable delays.  I wish to reflect on these and consider if any changes can be considered.

The pace of change

The celebrated chemical firm that became DuPont de Nemours was founded as a gunpowder producer in Delaware in 1802. Its growth to be an industrial powerhouse evolved over centuries.  Twenty-one years ago Google was a start-up (1998) and the word internet was a future concept.  According to my colleague Judge Marcoulli, the word “internet” was first mentioned in a 1998 judgment called Svenska Journalistförbundet[20].  Twenty-one years in terms of digital market development is an immensely long period, equivalent to millions or billions of years in the geological history of the earth.  By contrast, Uber and AirBnB involve rags to riches development in an astonishingly short time. Uber reached a turnover of $11 billion in less than ten years. 

In November 2012, Nick d’Alasio from London, then aged 15, developed an app called Summly which summarises news stories.  In March 2013 he received some $30 million from Yahoo which has built its features into its web offering.  Scan is an app designed by three college friends for a university student competition (they came second).  It offered better scanning of QR codes.  There had been 1 million downloads in three months, 10 million by March 2012 (when Scan had a staff of 9 people) and 25 million nine months later.  And then in 2014 Snapchat acquired the company for a reported $50 million.[21]  These rags to riches acquisitions sound charming.  Greater competition law concern has related to what might be called strategic acquisitions by giants. On the one hand, why should competition law discourage big firms from offering richer products, whether by acquisition or internal growth?  On the other hand, can a pre-emptive strategic investment chill the future for competitors?[22] Facebook acquired the social media company Instagram in 2012 for $1 billion as well as the messaging company WhatsApp in 2014 for a reported $19 billion. Google acquired the video platform YouTube in 2006 for $1 billion 650 million and the GPS navigation software application Waze in 2013 for a reported $970 million. It is suggested that high tech giants buy up promising start-ups with too much enthusiasm.[23] Is the current practice of handling such acquisitions of potentiality adequate?  Should we continue to assume that a merger between a large digital company and a promising start-ups has positive effects on competition? Or should we rather take the view that, at least in some cases, such behaviour is likely to harm competition in the long term by reinforcing dominance, and adopt a sort of precautionary principle for mergers?

Now, it can happen that the passage of time cures the problem; dominant players become merely strong; technologies age or disappoint, or are overtaken.  Fortunes and barometers of tension go down as well as up.  The digital market is a dynamic and fast moving sector where market power may not be that durable. Whole technologies can quickly become obsolete. In the 1970s and 1980s a huge area of antitrust concern related to the mainframe computers of IBM.  In the 1990s and 2000s Microsoft’s ubiquitous PC software was a matter of big concern.  I remember the wars about access to IBM mainframe technology as yesterday’s battles.  Google was in its infancy, Uber and AirBnB were unknown.  Many bright ideas fail to proceed to prosperity.  The worst may not happen.  But there is always a risk that enforcement activity will be too late.  The remedies imposed on Microsoft in its battles about the design of Windows PC and server technology in 2004 (launching an emasculated version of Windows without media technology, and describing under licence how its servers communicated with one another for certain functions) had little marketplace impact.  Does the historic assumption that market dominance will cure itself after a few years need reviewing?  It might seem not to be the case consistently.  At what point should we consider that the acquisition of an innovative start-up by a digital giant is a so-called “killer-acquisition” that poses a threat to competition?  How to analyse the potential for a start-up to become a competitor on the market irrespective of the acquisition?  Should competition authorities be systematically tougher, more intrusive, scrutinising every purchase by big digital companies, and decide to prevent potentially a merger where it appears that the targeted start-up is likely to succeed independently and thus provide customers with another choice?  If so, what legal instrument is more appropriate?  Should we adopt more intrusive regulations to fill a current gap or interpret current rules in order to adapt to the new market dynamics?

Complainants routinely warn that by the end of next month, commercial disaster will ensue unless something is done.  (I used to be fluent in sincerely singing that song: clients genuinely feared for their survival). Moreover, it is said that market places can “tip” in favour of a dominant player such that competition law may be the only way to restore normal market condition.  So there may be urgent need for a remedy issued by an administration.  Yet the administrative process will necessarily be controversial.  Suppose that the enforcer orders an important change in the business model of an allegedly dominant player.  The measure may be hopelessly inadequate for the complaining party.  It may be disastrous for the dominant one.  The findings on relevant market, dominance, abusive behaviour, remedy and so on will normally be controversial.  The remedy may not obviously match the abuse needing cured and may be far more burdensome than the fine. The relevant market may turn on highly controversial categorisations.  Parts for one brand of cash registers?  Electric cars, hybrid cars, all cars?  All servers or inexpensive servers? Search functions or other functions?  The decision will involve successive findings of fact and will probably involve novel legal theories or at least the adaptation to the care of existing theories.  There will be a bundle of factual points and then a bundle of legal ones. How can the court do an efficient and useful job?

Getting clarity on the theory of the case can take time.  Developing a case may take several attempts.  In Microsoft there were three Statements of Objections over four years, as the Commission reformulated its reproaches in function of Microsoft’s explanations ant its own enquiries.  In Google Shopping the proceedings started in November 2010.  After several years of discussions commitments were apparently offered but were regarded as insufficient. There were two Statements of Objections, in April 2015 and July 2016. In June 2017 the Commission took a decision and imposed a fine of 2.42 billion euros.  The Google Android mobile device decision was considerably faster, with the opening of formal investigations in April 2015, the sending of the statement of objections one year later and the fining of Google €4.34 billion in July 2018.  The Mastercard II investigation was opened in April 2013. A statement of objections regarding Mastercard’s cross-border card payment services and its inter-regional interchange fees was sent two years later. The Commission came to a conclusion in terms of the cross-border rules and fined Mastercard €570 million. Concerning inter-regional interchange fees, the Commission accepted the commitments proposed by Mastercard in April 2019.  Such a period does not seem unusual.  Diligent enquiry takes time, as the questions are important, very technical, and fiercely debated.

What was called a courageous decision against Facebook taken by the Bundeskartellamt earlier this year was later blocked by the Düsseldorf Oberlandesgericht.[24]  It is no secret that digital companies stock and use data for marketing and commercial purposes.  In the digital world, information has been called the new gold.  Every piece of information users give can be used, whether it is to target them with proper advertising or to sell as data to third parties, even political parties.  Users largely lose control over the collected data.  In its decision, the Bundeskartellamt imported the values of contract and data protection principles into antitrust analysis.[25]  It concluded that Facebook was exploiting users’ data without their consent and had imposed inappropriate contractual terms and conditions.  It also indicated that such behaviour could have negative consequences on competition, as competitors unable to amass data in a similar measure would face an additional barrier to market entry.  This decision is only one – important – element in the wider discussion surrounding data handling and its impact on competition in the digital market.  It also supports the argument that existing antitrust laws are not obsolete and can be used in a flexible and open way to answer to challenges brought on by the digital economy.  The fact that it was so quickly overturned however could stimulate legislators wishing to regulate digital companies dominating the market.  The question remains open.  But let us remember the merit of short, equivocal, timeless terminology which like a constitution can be readily applied to new situations.

So the pace of change in the marketplace seems to be accelerating, while the complexity of the facts, the novelty of the legal questions and the sharply opposed views of the facts make reaching a conclusion slow and difficult.

The administration’s task

On the other hand is the factor of carefulness.  The European Commission is not a reckless institution.  It has to make careful assessments about very controversial subjects.  It is well aware of the possibility of judicial review. Internationally within the framework of the ICN and the OECD it preaches for best practices in antitrust enforcement around the world.  Even though its procedures may be criticised, its decisions take time, preparation and checking.  Every market participant will be keen to be heard.  Every step of a Commission case is likely to be closely watched and very controversial.  Reaching a conclusion may be arduous.  The institution knows about judicial review and drafts accordingly.

“At every stage in a competition investigation, we’re aware that our reasoning and our choices have to satisfy the courts.  So for us the rule isn’t just an abstraction.  It’s a constant presence, in the work that we do”.  (Commissioner Margrethe Vestager, 10th May 2019)

Thus there are consultations, notes, meetings, advice from the Chief Economist, the Legal Service, maybe Member State submissions, certainly extensive filings by the main parties, and probably voluminous reports from economists and technical experts.  Getting to a conclusion from all that takes time.

The Courts

The third phenomenon to bear in mind is the structure, tradition and method of the European Union’s courts.  Like other international courts, ours has particular challenges: we receive cases in 24 languages; we work in one language; we are in theory not bound by any doctrine like stare decisis but we do in fact accord very great weight to precedent.  Ours will be the first judicial word on the controversy.   Although it would be very interesting to hear prosecutions of accused wrongdoers and decide with the apparatus of the criminal courts’ processes, the Treaties do not so provided. 

Article 263 of the TFEU establishes for the EU courts a jurisdiction to hear appeals “on grounds of lack of competence, infringement of an essential procedural requirement, infringement of the Treaties or of any rule of law relating to their application, or misuse of powers”.  Unlimited jurisdiction is granted by Article 261 of the TFEU in the case of “penalties provided for in …. Regulations”.

The General Court’s remit is to apply administrative law.  French administrative law principles have had a particularly strong influence on the shaping of the court’s practices.  The French language, in which we deliberate, naturally contributes to that phenomenon.  Aware of the possibility of appellate review, and aware of the raw importance of the topics at stake, the judges of the General Court are naturally cautious.  So we have a fast-moving market place which regularly presents competition law questions, and those questions often appear to require rapid emergency intervention.  We have a public enforcement body which has to respect a considerable number of procedural formalities before taking a decision.  And we have a court which is at first instance compelled to get a grip upon a very large file. 

Competition cases before the General Court may be very voluminous.  Among pending cases before the Court are files with over 10,000 pages.  One decision by the Commission had nearly 1000 pages. In another case the application had 60 pages with about 5000 pages of annexes (a big chunk of which was the Commission’s decision). Another application has 100 pages and thousands of pages of annex, including a Commission decision of over 300 pages.  By contrast, cases like Magill[26] which addressed big principles of competition law were far more manageable.  In the cases of Magill, the Commission decision had 9 pages, the CFI judgments had 46 and 37 pages and the Court of Justice 31 pages.

This first instance competence in competition cases may be contrasted with our jurisdiction over trademark and plant variety decisions.  A dissatisfied seeker of a registered trademark can seek reconsideration in Alicante, and then can go to a Board of Appeal in Alicante, and then challenge the decision of the Board of Appeal in Luxembourg.  Each of these steps reduces the distractions, focusses on the key issues, and leaves the topics in contention clearer when the case arrives in Luxembourg.  By contrast, in a competition case the winnowing of the facts and focussing of the controversies has not occurred.

Following a period of uncertainty,[27] it is now clear that the courts must deliver a certain level of intensity of inquiry into the well-foundedness of the competition measure: was the evidence reliable and complete, and did it support the conclusions reached?  In the cases of Halcor / Chalkor, the Court of Justice confirmed that regardless of the Commission’s margin of discretion the courts must perform a thorough review of law and facts in a competition appeal.  Following Chalkor[28] and Menarini[29] and accompanying developments we are obliged to make a review of fact and law which is in principle complete and thorough.  Those developments are of course important.  But they do not deliver a universal answer to the profound question of how intense should be the judicial scrutiny of a competition agency’s decision, especially one which is based on a long succession of factual determinations.

The number of competition appeals is declining.  The court was created in 1989 to deliver a more thorough degree of judicial scrutiny in competition cases and antidumping cases than the Court of Justice had been able to give in the 1980s.  Since 1989 the number of appeals on competition matters has fluctuated: in 1995 65 cases; 1999 34 cases; 2001 39 cases, 2010 79 cases, 2015 17 cases; 2018 28 cases. The raw numbers can be deceptive: some cases are hugely important; some are important for the parties but do not present new principles; quite a number deal with comparative justice in fining cases; but it does seem to be the case that the trend in competition appeals to the General Court is declining.

The phenomenon of settlements

The Commission has encouraged settlements in cartel cases, which nowadays  usually involve on appeal the question of whether two companies seeking leniency were equally treated in the setting of the fines.  Thus a leniency application involves oral descriptions of infringing conducts recorded on tape on behalf of companies which wish to confess to participation in cartel-like activity.  A very large number of enterprises have admitted to participating in illegal or highly imprudent behaviour. In exchange for accepting to sign a detailed written confession, the fine can be considerably lower than otherwise planned by the Commission.

It is interesting to note the Commission’s practice of settling cartel cases through private confession, negotiation and public acknowledgement of guilt is being extended mutatis mutandis to help conclude abuse cases, by offering a discount of a certain percentage from the fine which would otherwise have been imposed, a conclusion which involves a public acknowledgement of guilt.

In September 2016, ARA, a dominant recycling company in Austria, received a reduction in its fine of 30% in recognition of its cooperation and admission of guilt. And again in the case of AB InBev in May 2019 the company publicly announced that it had reached a settlement with the European Commission whereby the beer producer admitted hindering cross border trade (by labelling beer cans with national football motifs which were meant to impede trade between Belgium and Netherlands) and promised to reinforce its future compliance efforts.

In each of these cases the parties of course were able to point to an early conclusion to the controversy with a lower fine. They assumed, presumably, that if the decision had been taken and the heaviest fine imposed, the appeal would have been unsuccessful.  The leniency regime is being deepened with eLeniency,[30] a new online tool offered by the European Commission to facilitate the submission of statements and documents to DG Comp.  Instead of dictating oral corporate statements at the DG Comp’s premises one can now type in these statements directly to DG Comp’s secure server and upload supporting documents.  This new tool does not change the familiar leniency procedure, but is meant to make it easier to companies to file a leniency application. 

I cannot possibly regret the reaching of settlements whereby cases are closed and life moves on.  But it is interesting to observe the trend to resolve a competition law controversy in Brussels rather than in Luxembourg.  In one sense, there is much to be said in favour of avoiding litigation.  Each side gets less than it hoped for, but the matter is concluded on terms which could have been worse.  In another sense, the absence of judicial input could be seen as a missed opportunity to get guidance on novel theories in a new market place.  The legitimacy of an enforcement system which in practice eschewed judicial review might be diminished. Bill Kovacic remarks that administrative action gets stronger roots and better respect when it is subject to genuine judicial oversight.  Sir Robert Megarry, a wise and learned judge, used to ask visitors to name the most important person in court.  They never guessed correctly: his answer was always the same - the person who is going to lose.

A modern competition decision takes a position on a succession of factual disputes.  Suppose the question is which is the relevant market: hybrid cars and electric cars, small servers or all servers.  Each determination will be capable of being justified, and probably will agree with one set of expert reports and disagree with another set.  While it is true that the determination is finally a factual question to be decided in light of the criteria endorsed in previous cases[31], it may be a very delicate question.  Other distinctions may be equally delicate.  Should it be an obvious restriction of competition to prohibit a reseller to make internet sales or certain kinds of internet sales?

It is certain that predicting economic outcomes or public demand, or success or failure in the digital world, is exceedingly difficult.  Market developments can be swift, highly lucrative or severely anti-competitive.  It does not follow that we need a new rule book.  There may be decisions which are not obviously parallel, and where the lines of distinction may be fine.  We can note in the online world the Pierre Fabre[32] and Coty[33] cases.  Pierre Fabre, a moderate sized producer of hygiene products argued unsuccessfully that maintaining a prestige image for its products was not adequate justification for restricting passive sales.  By contrast Coty was able to promote the standing of its luxury perfumes by prohibiting online sales via third-party platforms.  The Pierre Fabre court took the view that maintaining a prestigious image for cosmetics albeit with a modest market share was not a legitimate aim which justified restricting competition

The Coty Court held that cosmetic and body hygiene goods were not luxury goods and that by contrast Coty’s perfumes, being luxury items, could properly be distributed pursuant to contractual restrictions to prevent them being mass-marketed.

It is not difficult to imagine the arguments which could be made on each side in each case. The distinction is a fine one, which could suggest that the underlying rules are flexible enough to keep being relevant.  The challenge is to grasp the realities of the market, to isolate the core controversy from the peripheral ones and to reconcile speed and thoroughness.

Despite the requirement of Menarini and Chalkor, there is a real challenge in effecting a proper review of the dozens of factual determinations which make up an important decision.  There will never be no factual basis for the finding.  How robust and interventionist should the court be?  At what point does a judicious conclusion of fact become an error of law?  How relevant are factors such as reasonableness and proportionality?

The European Commission enjoys, for the moment, the reputation of being an innovator in competition policy.  Novel theories prosper more readily in Brussels than in Washington DC or before federal judges in the US district courts.  The explanation is in part a reflection of government public enforcement policy which naturally fluctuates with elections.  Another explanation is that private enforcement is conducted before federal judges who are cautious about legal innovations.  The extent of judicial scrutiny of administrative action is an interesting topic of comparative enquiry.  In the common law tradition there is certainly respect and deference, but the notions of reasonableness and (maybe) proportionality are rather different than would apply in continental Europe.[34]

Some alternative approaches

I tentatively suggest that judicial review is a useful element in a lawful and effective system of competition law enforcement. 

We are encouraged to take more account of effects and market reality, rather than theoretical notions of per se offences.  That enquiry means fitting the assessment within the Treaty’s provisions on review of legality, as was commended in Chalkor.

The number of competition appeals coming to our court has fallen sharply over the past dozen years: from 79 to 28 in 2018, and 17 in 2015.  Many more competition problems are being disposed of in Brussels rather than in Luxembourg. The incidence of cases which are concluded in Brussels can be regarded as good news, in that settlements are never undesirable.  But it might also be a source of concern in that they mean the competition law is advanced and enforced without input from Luxembourg.  The legitimacy of an enforcement system which tends not to involve judicial review might appear to be diminished.

The lapse of time between distortive actions in the market and the final judicial word after a complaint, an investigation, a decision and an appeal is very long.  In a market where change is very swift and where it seems that market capture of “tipping” is a real possibility, such a delay is especially dangerous.  Intervening or not intervening can be equally controversial.

If we were starting afresh the structure of the system would be different, but that does not change the fact that the current regime has been repeatedly challenged and repeatedly upheld.  Naturally, the fact that a system has been endorsed as legal does not mean that it cannot be improved.  One route which has been proposed by several authors, including Advocate General Nils Wahl, is that the Commission’s procedure should involve a contradictoire stage, whereby someone like an administrative law judge (the Hearing Officer would be an obvious candidate) could preside over a hearing and make findings of fact.  This could or should narrow the areas of controversy and focus what is in dispute and what is secondary.  It might address the genuine problem of vast decisions which are hundreds of pages long and which contain hundreds of factual allegations on both sides.  (One of my colleagues unkindly said that one such decision was a vomit rather than a legal act).  So one reform which would be welcomed by many would be to ensure that a decision is capable of being read by a human being of adequate attentiveness in less than two hours.

Another approach would be to explore the greater use of interim measures decisions by the Commission.  The report by Professor Jason Furman in the UK[35] recommended greater use of interim measures in UK competition cases.  I was surprised to observe that Article 8 of Regulation 1/2003 had never been used. I note that the judgment in La Cinq[36] suggest that the Commission’s approach to interim measures was too conservative. The Commission rejected a request for interim measures submitted by La Cinq, a French TV operator, concerning complaints it had made to the Commission regarding the European Broadcasting Union. La Cinq appealed.  The Court of First Instance found that the Commission had taken a too restrictive line.

“[…] [T]he reasons adopted by the Commission for the decision at issue […] amount to requiring that, for a grant of interim measures to be possible, the existence of a clear and flagrant infringement must already be established at the stage of the mere prima facie appraisal which has to serve as the basis for the grant of such measures.” (para 61)

“It follows that, by identifying the requirement of a 'prima facie infringement' with the requirement of a finding of a 'clear and flagrant infringement' at the stage of interim measures, the Commission based its reasoning on an erroneous interpretation in law of the condition relating to the probable existence of an infringement.” (para 62)

“In this respect it must be held that, in stating in its decision that 'the only damage that can be regarded as irreparable is that which cannot be remedied by any subsequent decision', the Commission adopted a legally incorrect conception of irreparable damage the existence or risk of which could justify the adoption of interim measures.” (para 79)

“Moreover, the Commission's interpretation would make it almost impossible to verify the fulfilment of such a condition, and this would, in practice, amount to depriving of all substance the power granted to it to adopt interim measures.” (para 81). 

The Commission took interim measures in IMS[37] but withdrew them after the court suspended their effectiveness, so the auspices are equivocal. 

The merit of interim measures is that the decision is taken relatively quickly, and identifies the key problems.  The decision will usually be quite short and the judicial examination can also be accelerated.  Priority can be given to identifying the correct legal principle.

Competition cases on big novel controversies present issues of fact and of law.  There is an understandable wish to be thorough in the Commission’s presentation of the facts in its decisions, for fear of being criticised by the General Court.  But that prolixity lends itself to the drowning of the facts in a sea of detail and a dilution of the clarity of what exactly is being legally reproached.  Thus the appeal has to consider a mass of factual disputes, and the justification for the novel, or at least particular, application of the existing law.  Yet the two exercises can get entangled.

Every case is different and of course it is not guaranteed that a provisional assessment is the best reaction to a competition law controversy.  But it is interesting that it seems Article 8 of the basic regulation is awaiting a test.  The Broadcom case this week looks as if interim measures will again be used in a new and controversial setting.[38]

Another possibility about which to speculate is the practice of amicable settlements negotiated with the assistance and supervision of the General Court.  These were provided for staff cases brought under Article 270 TFEU.  President Vesterdorf occasionally invited parties to review their differences in private after the oral argument.  I participated in such an exercise in the case of Tillack v Commission.[39]  Might there be interest in expanding that provision?

To be clear: the General Court certainly has the capacity to deal with important competition cases and certainly has the stamina to handle huge files.  The legal challenge of reconciling the public interest and the private interest remains important and difficult.  The increase in the number of judges, and the diminishing number of cases mean that there is human capacity.  Necessarily it takes time to master a large dossier and to deal with all the issues arising and producing an answer relatively swiftly.  The General Court is indeed open for business.


Each of us here present must have the sensation of having witnessed extraordinary and profound events since our first meetings 24 years ago.  Google was then a start-up.  Uber was not even a glimmer in an eye.  Smartphones were unknown.  Though it is tempting to say that our age has seen unique levels of change, I can observe that a group of elites meeting in Florence 100 years ago would have witnessed the emergence of the motor car, aeroplanes, air-conditioning, trans-atlantic communication and the Geneva Conventions as well as revolution, conflict and immense bloodshed.  We are not unique in facing change. That makes me think that our economic constitution, the EU Treaties, do not need modification to face new circumstances.

I conclude by recalling a celebrated case from the time of Queen Victoria - competition controversies are not at all new.  In Nordenfelt v Maxim Nordenfelt Guns and Ammunition[40], the inventor sold his business for about a quarter of a  million pounds and £50,000 in shares and accepted a non-compete obligation whereby he would not ‘either directly or indirectly, in the trade or business of a manufacturer of guns, gun mounting or carriages …  or ammunition’.  The relevant market was worldwide, and the potential customers were “Governments and potentates, great and small, civilized and savage, who for purposes offensive or defensive desire… Nordenfelt guns with suitable ammunition.”  The plaintiff/inventor/patentee/seller (sadly ruined and arguing his own case before the House of Lords) relied on the absolute notion of restraint of trade: his contract to keep out of the market was indeed a restriction on competitive freedom.  The Lord Chancellor commented: 

“…it must not be forgotten that the community has a material interest in maintain the rules of fair dealing between man and man.  It suffers far greater injury from the infraction of these rules than from contracts in restraint of trade.  I think it is now generally conceded that it is to the advantage of the public to allow a trader who has established a lucrative business to dispose of it to a successor by whom it may be efficiently carried on.”

“In the age of Queen Elizabeth all restraints of trade, whatever they were, general or partial, were thought to be contrary to public policy, and therefore void.  In time, however, it was found that a rule so rigid and far/reaching must seriously interfere with transactions of every-day occurrence.  So the rule was relaxed.

When the series of cases, from the earliest to the present time, are carefully considered, I think they will be found to record the history of a protracted struggle between the principle of common honesty in private transactions, on the one hand, and the stern rule which forbade all restraints of trade on the other.”

A good rule of reason analysis.  So there was a defeat for the individual who invoked the basic, absolute principle and tried to escape his self-imposed duty not to compete.  A worldwide ban on post-sale competitive activity was, on balance, upheld by the House of Lords.  The courts’ discussion of whether to follow a rigid but narrow rule or a “rule of reason” approach was intriguingly modern.  The courts were aware of the need to avoid formalist rigidity, but equally they attached high importance to the binding effect of business decisions. The facts were carefully considered and merits were eloquently debated.  The outcome may have disappointed Mr. Nordenfelt.  But I do not doubt that he felt he had been listened to.  125 years later, my Court aspires to be equally respectful to all parties, including the person who is going to lose.

[1] Ian Forrester is Honorary Doctor of Laws and Professor of Law at Glasgow University and is a judge of the General Court of the European Union.  These remarks are obviously personal comments, which are in no way attributable to anyone else, certainly not the institution where I have the pleasure of working.  The questions are truly questions, not hints as to answers.  Thanks are due to Sarah Langenstein and Julia Kohler, PhD candidates, Nathalie Leyns and Vivien Terrien for their help and ideas.

[2] See for example Case C-131/12, Google Spain v AEPD [2014]; Case C-361/14, Schrems v Data Protection Officer [2016]; Joined Cases C-465/00, C-138/01 and C-139/01, Rechnungshof and Others v Österreichischer Rundfunk [2003] ECR I-4989; Joined Cases C-293/12 and C-594/12, Digital Rights Ireland Ltd and Seitlinger and Others [2014].

[3] The Hearing Officer has on a few occasions informally made recommendations to the Commissioner, suggesting that one or other aspects of the case were unpersuasive.  Such representations are informal and are not notified, certainly not made public.  But they can be effective.

[4] Donald Slater, Denis Waelbroeck and Sébastien Thomas, “Competition Law Proceedings before the European Commission and the Right to a Fair Trial: No Need for Reform?” (2009), 5 European Competition Journal 97; Denis Waelbroeck and Denis Fosselard, “Should the Decision-Making Power in EC Antitrust Procedures be left to an Independent Judge? – The Impact of the European Convention of Human Rights on EC Antitrust Procedures” (1994) 14 Yearbook of European Law 111; Denis Waelbroeck and Catherine Smits, "Le droit de la concurrence et les droits fondamentaux", in Les droits de l'homme dans les politiques de l'Union européenne (Brussels, Larcier, 2006); Ian Forrester, (2009) Due process in EC competition cases: A distinguished institution with flawed procedures. European Law Review. 34. 817-843; Arianna Andreangeli, “EU Competition Enforcement and Human Rights” (Cheltenham, Edward Elgar, 2008).

[5] Wouter Wils, ‘The Combination of the Investigative and Prosecutorial Function and the Adjudicative function’ (2004), 27 World Competition 201; id., ‘Powers of Investigation and Procedural Rights and Guarantees in EU Antitrust Enforcement: The Interplay between European and National Legislation and Case-Law’ (2006), 29 World Competition 3; id., ‘The Increased Level of EU Antitrust Fines, Judicial Review and the ECHR’ (2010), 33 World Competition 5; id., ‘Discretion and Prioritisation in Public Antitrust Enforcement, in Particular EU Antitrust Enforcement’ (2011), 34 World Competition 353; id. The Compatibility with Fundamental Rights of the EU Antitrust System in Which the European Commission Acts Both as Investigator and as First-Instance Decision Maker’ (2014), 37 World Competition 5.

[6] Case C-7/97, Oscar Bronner GmbH & Co. KG v. Mediaprint Zeitungs- und Zeitschriftenverlag GmbH & Co. KG [1998] ECR I-7791.

[7] Case C-395/87,  Ministère Public v Jean-Louis Tournier [1989] ECR 2521.

[8] Case C-453/99, Courage Ldt. v Bernard Crehan and Bernard Crehan v Courage Ldt and Others [2001] ECR I-6297.

[9] Case C-60/81 IBM v Commission [1981] ECR 2639, and the ensuing procedures.

[10] Case T-5/02 Tetra Laval v Commission [2002] ECR II-4381.

[11] Case T-342/99 Airtours v Commission [2002] ECR II-2585.

[12] Case T-310/01 Schneider v Commission [2002] ECR II-4071.

[13] Joined Cases C-36 to 38 and 40/59 Geitling and Others v High Authority [1960] ECR 423.

[14] Ian Forrester , ‘ A Bush in Need of Pruning: the Luxuriant Growth of “Light Judicial Review” ’, in Claus-Dieter Ehlermann and Mel Marquis , eds, European Competition Law Annual 2009: Evaluation of Evidence and its Judicial Review in Competition Cases, Hart Publishing, 2011, 423 et seq.

[15] Joined Cases 56 and 58Etablissements Consten SARL and Grundig-Verkaufs-GmbH v Commission of the European Economic Community [1966] ECR 299.

[16] Case 42/84, Remia BV v Commission [1985] ECR 2545.

[17] Case T-201/04, Microsoft v Commission [2007] ECR II-1491.

[18] Joined Cases C-89/85 etc., A. Ahlström Osakeyhtiö and Others v Commission [1993] ECR I-1307.

[19] W. Shakespeare, Hamlet, Act 2, Scene 3 (“the insolence of office” was another)

[20] Case T-174/95 Svenska Journalistförbundet v Council of the European Union [1998] ECR II-2289.

[21] I am indebted to Friedel Van Peer for these references. See also,

[22] For the avoidance of doubts, I note these interesting questions and express no conclusion about their answers.

[23] For a more detailed list of high value acquisitions by large digital companies, please see “Unlocking Digital Competition”, Report of the Digital Competition Expert Panel, March 2019 available at

[24] Facebook, 7 February 2019, Bundeskartellamt; Facebook v Bundeskartellamt, 26 August 2019, Düsseldorf Higher Regional Court

[25] For more detailed opinions on the cases, I recommend Eleanor M. Fox, “Platforms, Power and the Antitrust Challenge: A Modest Proposal to Narrow the U.S.-Europe Divide”, Symposium Issue, Nebraska, L. rev. vol. 98; Anca D. Chirita, “Global Platform Dominance: Abusive or Competition on the Merits?”, 2019

[26] Joined Cases C-241/91 P and C-242/91 P, Radio Telefis Eireann and Independent Television Publications Ltd v Commission [1995] ECR I-743.

[27] Ian Forrester, A Bush in Need of Pruning: the Luxuriant Growth of “Light Judicial Review”, in C-D Ehlermann and M. Marquis ‘eds), European Competition Law Annual 2009: Evaluation of Evidence and Its Judicial Control in Competition Cases (Oxford Hart Publishing, 2011) 407-452

[28] Case C-386/10 P Chalkor AE Epexergasias Metallon v Commission [2011] ECR I-13085.

[29] Menarini Diagnostics SRL v Italy, No 43509/08, 27 September 2011.


[31] M. Sousa Ferro, JECLAP 2015, 6 “Determining whether a certain degree of interchangeability between two products is “sufficient” to include them in the same market is a legal issue: it depends, not on economic theory or analysis, but on the criterion adopted in the case law to decide how much is enough.  Otherwise, there would be no legal certainty or predictability in the application of competition law, as the degree of “sufficiency” would always be subject to scholarly debate between economists.  It would also render judicial review impossible”.

[32] Case C-439/09, Pierre Fabre Dermo-Cosmétique SAS v Président de l’Autorité de la concurrence, Ministre de l’Économie, de l’Industrie et de l’Emploi [2011] ECR I-9419.

[33] Case C-230/16, Coty Germany GmbH v. Parfümerie Akzente GmbH [2017].

[34] The French tradition of judicial review of administrative action gave weight to the notion of objective, neutral control of official action, from which central authority would be checked and reinforced. By contrast German notions accorded more weight to the protection of individual rights and the aims being pursued. Proportionality is a contribution to European law from German law, which balances the appropriateness and necessity of the measure against the burdens imposed on the individual. Würtenberger/Neidhardt, “Distance et rapprochement entre le droit administratif allemand et le droit administratif français”, available under [Accessed 14 October 2019]; Schmidt-Assmann/Dagron, “Les fondements comparés des systèmes de droit administratif français et allemand”, available under [Accessed 14 October 2019].

[35] Jason Furman and Others, “Unlocking digital competition, Report of the Digital Competition Expert Panel”, available at:

[36] Case T-44/90, La Cinq SA v Commission [1992] ECR II-1.

[37] Case T-184/01 R, IMS Health Inc. v Commission [2001] ECR II-2351.

[38] European Commission Press Releases, “Antitrust: Commission opens investigation into Broadcom and sends Statement of Objections seeking to impose interim measures in TV and modem chipsets markets”, available at: and “Antitrust: Commission imposes interim measures on Broadcom in TV and modern chipset markets”, available at:

[39] Judgment of the Court of First Instance (Fourth Chamber) of 4 October 2006, Hans-Martin Tillack v Commission of the European Communities.

[40] Nordenfelt v Maxim Nordenfelt Guns and Ammunition [1894] AC 535.