Tuesday, May 13, 2014

Salomons II

Last July, I got quite miffed about the Competition Appeals Tribunal’s judgement in Akzo Nobel v. Competition Commission. I argued that the CAT had bent the rule of Salomon v. Salomon, which says that shareholders and the company they own are legally treated as distinct entities, past the point of breaking. For this reason, I recommended an appeal.

Apparently, so did Akzo’s solicitors, because in April the Court of Appeal ruled in its version of Akzo Nobel v. Competition Commission, deciding unanimously to dismiss the appeal.

While the Briggs, LJ.’s judgement for the Court is clearly better than the CAT’s, I’m not quite sure that I’m entirely convinced. While he stays light years away from the Single Economic Unit case law invoked by the CAT, when you get right down to it the result isn't so easily distinguished.

The logic of the Court of Appeal is quite straightforward:

  1. The legally relevant criterion, from s. 86(1)(c) Enterprise Act 2002 is that Akzo Nobel must be “a person carrying on business in the United Kingdom”.
  2. Firstly, it is important to note that that doesn’t have to be the same business as the business that is seeking to merge or acquire. All that is needed in order to establish jurisdiction is that Akzo Nobel carries on any kind of business in the UK.
  3. The Akzo Nobel group “may fairly be described as carrying on business in the Netherlands and in the UK.” (par. 33)
  4. The only question is by whom, i.e. by which legal person, this business is carried on.
  5. Akzo Nobel NV, the Dutch parent company, is held to be managing the business “both strategically and operationally” (par. 36)
  6. For this reason, Akzo Nobel NV is carrying on business in the UK, meaning that the Competition Commission is entitled to make an enforcement order against it.
In order to distinguish Salomon v. Salomon, Briggs, LJ. makes a distinction between a parent who does nothing more than “to exercise its rights as shareholder in the traditional fashion, leaving the entire management of the business to the subsidiary’s directors” (par. 37) and “the exercise of the strategic and operational management and control of a manufacturing and sales business” (par. 34).

Even on its face, this strikes a huge blow against the rule of Salomon v. Salomon, given that something akin to Akzo’s management structure is used by “most modern corporate groups” (par. 12). Are we really to believe that the parents of all those groups have a sufficient presence in the UK for the Competition Commission to go after them?

While the judge contrasts his test with the presence test advocated by Akzo, which Parliament could have but did not enact (par. 30), where he seems to have landed is that almost any international group that trades in the UK is “carrying on business” there, which is also not what Parliament enacted. Parliament could have easily extended the jurisdiction of the CC to all companies that trade in the UK, i.e. that buy or sell goods or services there, but it did not. To do so would have made the CC’s jurisdiction to make enforcement orders essentially equivalent to its jurisdiction to investigate. That would have been convenient for the CC, but it is not the law. (As the Judge admitted in par. 25.)

A potential tie-breaker is the purpose of s. 86(1), which is to establish possible connecting factors between the addressee of an enforcement order and the UK, so that the UK does not violate international comity in seeking to regulate behaviour outside its borders (par. 26). This is certainly true, as far as it goes, but in order to apply this tie-breaker the judge would have had to survey the general international understanding of comity, which he did not do.

Briggs, LJ. might have referred to internet regulation cases in the European Court of Justice, like the Google Adwords case or today’s Google Spain, or to the US Supreme Court’s case on the Alien Tort Statute last year in Kiobel v. Royal Dutch Petroleum, where the Supreme Court sharply limited the jurisdiction of US courts over torts outside its borders. In competition law specifically, there seems to be quite a bit of patience with extraterritorial decisions, but there are limits. The € 1.06 bn fine against Intel, for example, drew criticism from the US on comity grounds. (The judgement from the General Court is expected on 12 June.) 

The same goes for the Commission’s decision to block the proposed merger between Honeywell and GE in 2001. (See also this research paper commissioned by the Commission.) If comity does not prevent the EU from blocking a merger between Honeywell and GE, presumably it does not prevent the UK from blocking the Azko/Metlac takeover either. But the Court of Appeals made no findings in this regard.

Returning to the rule of Salomon v. Salomon, there is one final question that troubles me: If Akzo Nobel N.V. is carrying on business in the UK because it “exercise[s] the strategic and operational management and control of a manufacturing and sales business” in the UK, where does that leave Mr. Salomon? As a shareholder/director, he surely did the same thing. So was he, as a private person, carrying on a business in the sense of s. 86 EA2002? To be sure, the criterion enacted there is particular to competition law, but that doesn’t mean the question isn’t important. If the shoemaker Mr. Salomon had lived in France while selling his shoes through his shop in the UK, would that have meant that Mr. Salomon himself – as opposed to his company – was carrying on business in the UK? Could the Competition Commission have prevented him from merging his business with the business of Mr. X, also resident in France with a shop in the UK? Given the Court of Appeal’s judgement, it seems clear that the CC could have. But I’m not sure where that leaves corporate legal personality.

Thursday, January 30, 2014

Ferries & Competition - Appendix

Good news: In the Terschelling Ferries case that I blogged about on Monday, EVT won on appeal! The Court of Appeals in The Hague ruled today that EVT gets to continue to operate between Harlingen and Terschelling for the time being. (A new regulatory scheme is currently the subject of parallel litigation. Don't ask.)

Competition Law
As I predicted, the Court of Appeals did not go near the competition law aspects of the case, other than to discuss a letter sent by the Dutch Competition Authority ACM in November last year, where the ACM expressed grave concerns about the Minister's proposed solution. In this letter, it confirmed that competition law probably applied, because the government would probably qualify as an undertaking in this matter, given the Höfner precedent I mentioned.

The only difference between the ACM's tentative analysis and mine is that they approached this as a possible case of abuse of dominance by the State, while I preferred - and continue to prefer - to think of it as collusion between the State and TSM. In the end, that does not affect the analysis very much, except that the collusion frame would open the door to a fine for TSM. It is true that the State is almost certainly dominant in the market for Waddenzee port facilities, assuming the Harlingen port isn't an outright essential facility. Either way, the State's decision to terminate the contract with EVT clearly implicates one or more examples of abuse listed in the Treaty:
Article 101 TFEU (ex Article 81 TEC) 1. The following shall be prohibited as incompatible with the internal market: all agreements between undertakings (...) which have as their object or effect the prevention, restriction or distortion of competition within the internal market, and in particular those which: (b) limit or control production, markets, technical development, or investment; (c) share markets or sources of supply;
Article 102 TFEU (ex Article 82 TEC) Any abuse by one or more undertakings of a dominant position within the internal market or in a substantial part of it shall be prohibited as incompatible with the internal market in so far as it may affect trade between Member States. Such abuse may, in particular, consist in: (b) limiting production, markets or technical development to the prejudice of consumers;
(Obviously these provisions don't directly apply, since the required impact on trade between Member States is missing, but the Dutch Competition Act contains identical language for strictly domestic situations.)

Contract Law
The reason why I got my prediction wrong - I predicted that EVT would lose - is that I underestimated how much the Court of Appeals would be willing to second-guess the government's decision about what the public interest required, and how much the Court would be willing to do so in an expedited procedure. (Note the timeline: The District Court in The Hague heard this case on December 17 and ruled on December 31. EVT submitted its grievances with the Court of Appeals on January 10. There was an oral hearing on January 21, and the Court of Appeals ruled today.)

The District Court used a fairly deferential standard, holding essentially that the State could reasonably come to the conclusion that the public interest required termination. The Court of Appeals, on the other hand, relied on a Supreme Court decision from 2012 about a rental contract between a small business owner and a foundation specialising in playgrounds where the subject of the lease was a small recreational area. In that case the Supreme Court held that for long term leases a decision to terminate the contract must not necessarily be judged only against a deferential abuse of right standard, but that reasonableness ("redelijkheid en billijkheid") under art. 6:248 Civil Could could require that the court take a more general view of the circumstances of the case and the interests of each party.

I think that the Court of Appeals put too much weight on that precedent. The reasonableness standard of art. 6:248 Civil Code is particularly important to protect a weak party against a stronger one. While parties don't get any stronger than the State, and while EVT was faced with being put out of business, I don't think the circumstances of the case were analogous to those of the Supreme Court's case.  EVT knew what it was getting into when it first started the service, just like it knew what it was getting into when it asked and received permission to start using a larger ship. It was trying to enter a previously monopolised market, and it knew that the State (and the municipalities of Vlieland and Terschelling) would side with the incumbent in case of trouble. Hence the parallel litigation about the new concessions regime, aforementioned; EVT needs for this entire market to be organised differently if it is to have a fair chance. Making a competition law argument could have forced the State's hand, but this contract law approach is just too forgiving.

So why did the Court of Appeals go there? Well, I can only speculate, but it feels suspiciously like it was trying to hold the State to a public law standard even though this was a private law case. In administrative law, the standard would have been somewhere in between the District Court's abuse of right standard and the Court of Appeals' reasonableness approach. While the administrative law standard officially asks whether the government entity could have reasonably done as it did, in practice this allows for quite a bit of second-guessing. But this was not an administrative law case.

Finally, it should be mentioned that the Court of Appeals flagged up a much better contract law argument, without however relying on it. After holding that the State acted unreasonably in terminating the lease without sufficient cause and without consulting with EVT, the Court discussed EVT's chances in a full trial, i.e. a non-expedited one. It said that it is "not impossible" that EVT would win on its competition law argument (par 12), and then it noted that there may well be an issue with the contractual provision that the State relied on to terminate the contract. The provision in question requires, simply put, that continuing on with the contract should "prevent" a proper service with the islands. The District Court had interpreted that provision broadly, but the Court of Appeals proposed a more narrow reading. In par. 13-14 of its judgement, the Court argued that in a full trial EVT may well succeed in proving that its service is not actually preventing the service in any way, shape or form.

It may seem like a technicality, but I much prefer this argument to the one the Court actually relied on. Too bad it will probably never see the light of day, because while theoretically an expedited hearing is meant as a temporary measure until a full trial can be held, it is exceedingly rare for an expedited judgement to be followed by a proper trial. So unless the State takes this case to the Supreme Court, this litigation is now over.

Tuesday, January 28, 2014

Competition Law, Sustainability, and Geography

The Ferry post raised an interesting issue that I'd only casually discussed elsewhere: if we are going to weigh the damage done by collusion against one or more associated benefits, which benefits for which consumers do we get to include? Inconveniently, it seems like the answer is benefits for the same set of individuals who are the (ultimate) consumers of the product or service.

The Energy Pact Case
On 6 September 2013, the Dutch energy industry made a pact with the government and a variety of other stakeholders about the future of energy supply in the Netherlands, under the auspices of the allmighty SER, the Social-Economic Council. As part of this pact, the industry promised to close five coal-fired power plants that were particularly unsustainable in their output of SO2, NOx and particulates.
Given that such an agreement to reduce supply is arguably problematic under competition law, the industry body Energy Netherlands asked the competition authority ACM for its opinion. In a non-binding "note" just 20 days later, the ACM tentatively concluded that this part of the pact was unlawful as contrary to the Competition Act. Evaluating the social costs and benefits over the duration of the pact (2016-21), the Authority found an estimated negative price effect for consumers of € 450 million and a benefit - in the form of avoided carbon abatement costs - of € 180 million. Given the imbalance between costs and benefits, the conclusion inevitably followed, much to the chagrin of the government.

The question of "which consumers" was actually relatively easy in this case: Given that for SO2, and NOx there are nationwide emissions limits, the ACM simply assumed that a given reduction in NOx output would save Dutch society the € 9.40 per kg that it would otherwise have had to spend on other kinds of abatement. Similarly, for SO2 it assumed a the "shadow price" of € 5.40. Both these values were taken from a study by the CPB. For particulates, where no emissions limit exists, the CPB study estimated a benefit to society from reduced output of € 44.30 per kg. While the shadow prices methodology makes perfect sense, the story on particulates raises an important question: Why count only benefits enjoyed by Dutch citizens? Is this even OK under the ECJ's precedents on the Temelin nuclear power plant (see this article) and on environmental impact assessments?

EU Law
Well, it is difficult to get around the clear text of art. 101(3) TFEU, which requires that any agreement, in order to pass muster, must "[allow] consumers a fair share of the resulting benefit". In its decision in CECED, an important precedent for the Energy Pact case, the European Commission talked about checking whether "the agreement is likely to deliver both individual and collective benefits for users and consumers" (par. 51) and "allowing users a fair share of the benefits" (par. 57), without explaining the distinction - if any - between users and consumers. (That case dealt with an agreement in the Belgian appliances manufacturing sector aimed at improving energy efficiency.) This interpretation of primary Treaty law seems difficult to square with the idea that benefits enjoyed by consumers in a different (geographical) market, or by non-consumers, should also be included. This is also the conclusion reached by the Commission in its Guidelines on the application of Article 81(3):
43. The assessment under Article 81(3) of benefits flowing from restrictive agreements is in principle made within the confines of each relevant market to which the agreement relates. The Community competition rules have as their objective the protection of competition on the market and cannot be detached from this objective. Moreover, the condition that consumers must receive a fair share of the benefits implies in general that efficiencies generated by the restrictive agreement within a relevant market must be sufficient to outweigh the anti-competitive effects produced by the agreement within that same relevant market. Negative effects on consumers in one geographic market or product market cannot normally be balanced against and compensated by positive effects for consumers in another unrelated geographic market or product market. However, where two markets are related, efficiencies achieved on separate markets can be taken into account provided that the group of consumers affected by the restriction and benefiting from the efficiency gains are substantially the same.
By way of exception, the Commission quoted Case T-86/95, Compagnie Générale Maritime and others:
130 (...) Regard should naturally be had to the advantages arising from the agreement in question, not only for the relevant market, namely that for inland transport services provided as part of intermodal transport, but also, in appropriate cases, for every other market on which the agreement in question might have beneficial effects, and even, in a more general sense, for any service the quality or efficiency of which might be improved by the existence of that agreement. Both Article 5 of Regulation No 1017/68 and Article 85(3) of the Treaty envisage exemption in favour of, amongst others, agreements which contribute to promoting technical or economic progress, without requiring a specific link with the relevant market.
However, the Commission explained away that seeming oddity by pointing out that "importantly, the affected group of consumers was the same" (Guidelines, op cit, fn 57). That seems right.

Product Market
Taking the Commission's guidelines at face value, even the ACM's analysis of the Energy Pact is too generous, since it weighs a price increase in the energy market against reduced abatement costs elsewhere in Dutch society, i.e. not necessarily in the energy market. After all, the Commission clearly said that an increase in the deadweight loss "cannot normally be balanced against and compensated by positive effects for consumers in another unrelated (...) product market". Since the ACM made no finding of where the reduction in abatement costs might occur, much less whether that other product market is somehow related to the energy market, the Authority seems to have fallen foul of the Commission's guidelines.

Fortunately, the Commission itself doesn't seem to go that far either. In CECED, it started by estimating the recoupment period for the average consumer, weighing the increased purchase price of the appliances against the resulting reduction in energy costs (par. 52-54). Arguably, that is a weighing of costs and benefits in the same or in related markets. (See p. 14 of the ACM's draft Position Paper on Competition and Sustainability.) But next it looked at "collective environmental benefits", using a methodology much like the ACM's relying on estimates of the Europe-wide (!) social benefits of reduced SO2 and NOx emissions. (For comparison: in 1999 the Commission put the shadow prices at € 4-7 per kg for SO2 and at € 3-5 per kg for NOx.) More generally, throughout the case law of the Commission and the European Courts, I am yet to encounter an example of a justification that is rejected because it concerns the same inviduals but on a different product market, or on no product market at all. (The consumer benefit may come in the form of reduced government spending, for example.)

Geographical Market
Geography is a different story. The problem with geograpy is that it is often (but not always, see the Compagnie Générale Maritime case above) a proxy for the actual individuals we are talking about. As long as the "fair share of the resulting benefit" is somehow enjoyed by the same individuals, competition law is unlikely to object. But if the beneficiaries are different individuals from the individuals paying the proverbial and literal price for the reduction in competition, we have a problem. As we saw in the Ferries case, competition law does not look kindly on cross-subsidisation. Even if such a scheme is (Kaldor-Hicks) efficient, from an economics point of view, there needs to be some kind of compensation, enough to achieve Pareto efficiency. If it is impossible to achieve this in the context of an agreement between market participants, there is no alternative to government intervention: it will either have to become a party to the pact, offering to take care of the compensation, or it will have to turn the agreement into a statutory scheme with compensation. (Without compensation, the statute would arguably constitute an unlawful infringement of EU law.)

Ancillary Restrictions
One final possibility that is mentioned in the ACM's draft Position Paper on Competition and Sustainability is the notion of "ancillary restrictions" mentioned by the Court of Justice in a few cases, which are restrictions of competition that are sufficiently "unavoidable" if some other beneficial thing is to be done that they are outside the ambit of art. 101 TFEU. The notion has its roots in the law on concentrations:
104 In Community competition law the concept of an `ancillary restriction' covers any restriction which is directly related and necessary to the implementation of a main operation (see, to that effect, the Commission Notice of 14 August 1990 regarding restrictions ancillary to concentrations (OJ 1990 C 203, p. 5, hereinafter `the notice on ancillary restrictions', point I.1), the notice on cooperative joint ventures (point 65), and Articles 6(1)(b) and 8(2), second paragraph, of Regulation No 4064/89). 105 In its notice on ancillary restrictions the Commission rightly stated that a restriction `directly related' to implementation of a main operation must be understood to be any restriction which is subordinate to the implementation of that operation and which has an evident link with it (point II.4).
Case T-112/99, Métropole. In its draft Position Paper , the ACM cites some more recent case law that concerns associations of undertakings, such as the Dutch Bar Association in Wouters and the international swimming association FINA in Meca Medina. In the former case, the Court said:
97. However, not every agreement between undertakings or every decision of an association of undertakings which restricts the freedom of action of the parties or of one of them necessarily falls within the prohibition laid down in Article 85(1) of the Treaty. For the purposes of application of that provision to a particular case, account must first of all be taken of the overall context in which the decision of the association of undertakings was taken or produces its effects. More particularly, account must be taken of its objectives, which are here connected with the need to make rules relating to organisation, qualifications, professional ethics, supervision and liability, in order to ensure that the ultimate consumers of legal services and the sound administration of justice are provided with the necessary guarantees in relation to integrity and experience (see, to that effect, Case C-3/95 Reisebüro Broede [1996] ECR I-6511, paragraph 38). It has then to be considered whether the consequential effects restrictive of competition are inherent in the pursuit of those objectives.
And in the latter:
42 Next, the compatibility of rules with the Community rules on competition cannot be assessed in the abstract (see, to this effect, Case C-250/92 DLG [1994] ECR I‑5641, paragraph 31). Not every agreement between undertakings or every decision of an association of undertakings which restricts the freedom of action of the parties or of one of them necessarily falls within the prohibition laid down in Article 81(1) EC. For the purposes of application of that provision to a particular case, account must first of all be taken of the overall context in which the decision of the association of undertakings was taken or produces its effects and, more specifically, of its objectives. It has then to be considered whether the consequential effects restrictive of competition are inherent in the pursuit of those objectives (Wouters and Others, paragraph 97) and are proportionate to them.
43 As regards the overall context in which the rules at issue were adopted, the Commission could rightly take the view that the general objective of the rules was, as none of the parties disputes, to combat doping in order for competitive sport to be conducted fairly and that it included the need to safeguard equal chances for athletes, athletes’ health, the integrity and objectivity of competitive sport and ethical values in sport.
44 In addition, given that penalties are necessary to ensure enforcement of the doping ban, their effect on athletes’ freedom of action must be considered to be, in principle, inherent itself in the anti-doping rules.
While this comes dangerously close to adopting some all-purpose rule of reason, it seems like the ACM was right to conclude that this doctrine is insufficiently clear for it to be included in the Position Paper. At least, it is insufficiently matured outside the context of concentrations and associations of undertakings. Moreover, it is clear that the Court will not easily consider a rule to be sufficiently inherent or necessary to qualify. Comparing the Wouters precedent with the Commission's CECED decision, for example, it is difficult to see how the scheme at issue in the latter case could fall under the doctrine of ancillary restraints.

Where competition for one group of consumers is reduced for the benefit of another group of individuals, competition law is likely to pose significant problems.  In that regard, the ACM's Energy Pact "note" was a useful shot across the bow. Environmental lobby, beware!

Ferries & Competition

Recently, the District Court in The Hague tackled an interesting case about competition between ferry companies. While some of the legal details make the Court’s decision quite obviously right, the economics of the case are worth digging into a little further.

Starting on 18 November 2008, a company called EVTEigen Veerdienst Terschelling – operated a ferry service between the Frisian mainland in Harlingen and the Frisian Wadden Island of Terschelling. It did so in competition with the incumbent operator TSM, the Terschellinger Stoomboot Maatschappij. Critically, TSM, and only TSM, also provided a service to the neighbouring island of Vlieland. (Although there is a separate service connecting both islands directly.)

Until the spring of 2012 this arrangement caused few problems for anyone. The docks in Harlingen and Terschellingen, which are the property of the State, had more than enough capacity for both companies. However, from that point onwards EVT rocked the proverbial boat by operating its service with a bigger ship. It had permission from the State to do that, but that didn’t change the fact that EVT’s rental agreement with the State for the docks in Harlingen were subject to the condition that it could be terminated by the State at any time for reasons related to the public interest. (Par. 2.13)

In October 2013, the State considered that the deteriorating profitability of TSM posed a considerable threat to the continued existence of the main ferry service to Terschelling and Vlieland, and for that reason it terminated EVT’s rental agreement effective 1 February 2014. Starting next month: no more competition.

Legally, the matter is quite straightforward. The relevant term in the rental contract is cast in quite general terms, and the construction proposed by EVT which would exclude a situation like this from “the public interest” seems unreasonably awkward. EVT’s arguments under the Competition Act and state aid law are likewise either incorrect or at least not suited for consideration in an expedited procedure (“kort geding”). So EVT lost before the District Court, and presumably it will lose before the Court of Appeals as well.

However, we can wonder whether this is really the result we should prefer from a policy point of view. Is a monopoly service by TSM – or a near-monopoly by TSM constrained by a limited service by EVT – really the economically optimal outcome? The Court’s judgement contains two little doors to arguments to the contrary.

Firstly, the studies quoted in the Court’s statement of facts provide significant evidence for the proposition that the real problem here is that the ferry service to Vlieland is loss-making, meaning that TSM needs to turn a profit on its service to Terschelling in order to stay afloat. (Sorry.)

If that is the case, literature from other transport modes – particularly railways – suggests that a cross-subsidy of this kind is an inappropriate tax on the inhabitants and visitors of Terschelling for the benefit of the inhabitants and visitors of Vlieland. If the Dutch government considers that the public interest requires that Vlieland should be reachable at a price lower than the cost of operating the service, it should subsidise that service from general taxation. There is no reason to place the burden only with the inhabitants and visitors of Terschelling. This is the whole logic behind taking different parts of the rail network away from the incumbent and tendering them.

The second little door is in par. 4.20, where the Court rejects EVT’s competition law argument. The Court does so by bypassing the question of whether the State acted as an undertaker in this case and holding instead that any arguable infringement of the ban on collusive practices was objectively justified by the public interest. However, it is unclear what the analysis behind this conclusion is.

Whether the State acts as an undertaker in renting out port facilities is an interesting legal question. Presumably, in EU law under Höfner, the answer is that it is. There is nothing inherently sovereign about renting out port facilities. Under Dutch competition law, the special section of the competition act that deals with public sector bodies and their behaviour as market participants does not apply to the State or to ministries. (Cf. art. 25h(3) Mw.) The general definition of an undertaking in art. 1, on the other hand, simply refers to art. 101 TFEU.

The Court bypasses that question and holds that there is no (tacit) collusion aimed at removing competition, because what collaboration existed (the commissie bootdiensten – ferry services committee) was aimed at achieving the best possible transport infrastructure for the islands instead.

This misses the point between collaboration that is anticompetitive by object – which arguably did not occur here – and collaboration that is anticompetitive by effect. Given that the ferry services for each island form a separate market, there is no question that the state’s collaboration with TSM in the ferry services committee and elsewhere had the effect of reducing competition on the Terschelling market.

(The only way the different ferry services could be considered a single market is if a change in the ticket price for one island would cause passengers to choose to go to another island in significant numbers. While this may be remotely plausible for the tourism sub-market – depending on the size of the ferry relative to the total cost of the trip – it is certainly unlikely for residents and other non-tourists.)

The twist at the end is that the alleged justification – making sure that a ferry service would continue to exist for every island – does not suffice under competition law, because the “beneficiaries” of the collusion are not the customers of TSM and the state on the market in question (Terschelling), but rather their customers on a market where no anti-competitive collusion has been alleged (Vlieland).

This is the same problem that the Dutch competition authority encountered recently in its opinion on the Energy Accord. Given that the relevant market in that case was defined as the Dutch market, benefits enjoyed by foreign customers could not be taken into account. Even worse, it is doubtful whether reductions in global warming could be taken into account at all, given that these benefits are enjoyed by energy customers in their capacity of home owners, tax payers, etc., but not in their capacity of energy customers. (For more details, see the discussion here.)

However broadly we define “technical and economic progress” under art. 101 TFEU, the customers on the Terschelling market are worse off as a result of the State’s decision to terminate its rental agreement with EVT. So under any rule of reason-adjacent test, the State’s plea for justification must fail. Unfortunately, the evidence needed to apply such a test cannot be considered by a court in an expedited procedure. So EVT still loses.

Wednesday, October 16, 2013

Prisoner Voting

Something that made me mad in today's UK Supreme Court prisoner voting judgment. Per Lord Sumption:
115. From a prisoner’s point of view the loss of the right to vote is likely to be a very minor deprivation by comparison with the loss of liberty. There are no doubt prisoners whose interest in public affairs or strong views on particular issues are such that their disenfranchisement represents a serious loss, just as there are prisoners (probably more numerous) whose enthusiasm for active sports makes imprisonment a special hardship. The severity of a sentence of imprisonment for the convicted person will always vary with a wide variety of factors whose impact on him or her will inevitably be arbitrary to some degree. It has been said, for example, that disenfranchisement may bear hardly on someone sentenced to, say, a short period of imprisonment which happens to coincide with a general election. For some prisoners, this will no doubt be true. But I decline to regard it as any more significant than the fact that it may coincide with a special anniversary, a long anticipated holiday or the only period of fine weather all summer.
So much for a proper respect for democracy...

In order to heal the soul, here is a paragraph from Baroness Hale's judgement in the same case:
90. To take an obvious example, we would not regard a Parliament elected by an electorate consisting only of white, heterosexual men as uniquely qualified to decide whether women or African-Caribbeans or homosexuals should be allowed to vote. (...) Given that, by definition, Parliamentarians do not represent the disenfranchised, the usual respect which the courts accord to a recent and carefully considered balancing of individual rights and community interests (...) may not be appropriate.

Thursday, July 04, 2013

Phoebe Putney

An increasing problem in the competition law world is the relationship between competition law and varying degrees of state action in other realms. In Europe, this problem is artificially easy because of the hierarchy of laws involved. In Deutsche Telekom and the other margin squeeze cases, the European Court of Justice relied on the principle of primacy to hold that competition law, which is written in the main body of the Treaties, trumps sectoral regulation, which isn’t. When there is no Community dimension, and outside the EU, the answer isn’t quite that straightforward.

To begin with the easy one: When EU competition law applies, i.e. when someone is naughty in a way that “may affect trade between Member States”, the only question is who’s to blame:
  • If the offending behaviour is mandated by EU sectoral legislation, the secondary legislation in question is null and void, something that can be established through the prejudicial question procedure or through an action for annulment, as the case may be. 
  • If the offending behaviour is mandated by Member State sectoral legislation, the Member State is the culprit and it is up to the Commission to start an infringement procedure. In an appropriate case, the Member State legislation in question can also be slapped down by a national court, with or without a prejudicial question. 
  • If, on the other hand, the relevant sectoral law merely permits the offending behaviour without mandating it, the onus is on the company to make sure it obeys both competition law and sectoral law, meaning that it can be punished for failing with regard to the former. This is what happened in the margin squeeze cases.

At the Member State level, when there is no Community dimension, saying that your naughtiness is expressly permitted by sectoral legislation is fundamentally a sound argument. Of course, in a given case the question remains whether this is a correct interpretation of the law, which will depend on which statute was enacted first, whether there is express saving language somewhere, the way the US Telecommunications Act expressly says that it is without prejudice to competition law, and it will depend on the various canons of interpretation. 

In the British Albion Water competition law case, for example, Ofwat, which under the UK system is both the sectoral regulator and one of the possible competition authorities, argued that Welsh Water’s pricing structure (the so-called “costs principle”) was expressly permitted by section 66E WIA 1991, an argument that was obviously gratefully echoed by Welsh Water itself. The Competition Appeal Tribunal went along with that approach to some extent, but concluded that it was not enough to save Welsh Water from competition law liability under a margin squeeze theory. (Albion Water v. Water Services Regulation Authority, [2006] CAT 23, par. 920-980. Cf. also Ofwat’s decision, par. 324-331.)

The reason why all of this is interesting now is that the problem has cropped up in the US again. There, the situation is a mixture of the previous one. On the one hand, the relationship between competition law and sectoral regulation has a clear hierarchical dimension, because American competition law is enacted at the Federal level in the Sherman and Clayton Acts (15 USC 1-38), on the other hand, the Federal authorities are much more reluctant to have this law displace State law, the Supremacy Clause notwithstanding. The result is something of a half-way system, as summed up in the 1943 Supreme Court case of Parker v. Brown.

In Parker, the Court held that the competition laws don’t displace state law. The Court argued that the Federal legislator would not do something so drastic without explicitly saying so, and in this case it didn’t. (Nor did the sponsors of the Sherman Act.) If the states want to create a monopoly or a cartel somewhere, they are allowed to. At the same time, however, the Court didn’t take back its earlier holding that the States can’t shield private actors from antitrust liability. (Cf. Northern Securities Co. v. United States from 1904.) So all sectoral legislation that does not create an outright statutory monopoly or cartel winds up somewhere in the middle.

This can produce some pretty unsatisfying results in either direction. Earlier this year, in FTC v. Phoebe Putney Health System, the Court held unanimously (Justice Sotomayor writing for the Court) that some hospital shenanigans in rural Georgia were not shielded by Georgia’s health care legislation, because the naughtiness in question was not “clearly envisaged” by the legislator. This is odd because, as the Court summarises it, 

the Law authorizes each county and municipality, and certain combinations of counties or municipalities, to create “a public body corporate and politic” called a “hospital authority.” §§31–7–72(a), (d).Hospital authorities are governed by 5- to 9-member boards that are appointed by the governing body of the county or municipality in their area of operation. §31–7–72(a).
Under the Law, a hospital authority “exercise[s] publicand essential governmental functions” and is delegated “all the powers necessary or convenient to carry out and effectuate” the Law’s purposes. §31–7–75. Giving morecontent to that general delegation, the Law enumerates 27powers conferred upon hospital authorities, including the power “[t]o acquire by purchase, lease, or otherwise and tooperate projects,” §31–7–75(4), which are defined to include hospitals and other public health facilities, §31–7–71(5); “[t]o construct, reconstruct, improve, alter, andrepair projects,” §31–7–75(5); “[t]o lease . . . for operationby others any project” provided certain conditions are satisfied, §31–7–75(7); and “[t]o establish rates and chargesfor the services and use of the facilities of the authority,”§31–7–75(10).

I find it difficult to see how that does not include the right to authorise a private actor to monopolise hospital services in a given county. In that regard, the standard applied by the 11th Circuit, who asked whether the hospital authority’s actions were “foreseeable” in the sense of “reasonably anticipated” by the legislator, makes more sense to me:

[I]t is not necessary, the court reasoned, for an anticompetitive effect to “be ‘one that ordinarily occurs, routinely occurs, or is inherently likely to occur as a result of the empowering legislation.’” (…) [Applying that standard], the court reasoned that the Georgia Legislature must have anticipated that the grant of power to hospital authorities to acquire and lease projects would produce anticompetitive effects because “[f]oreseeably, acquisitions could consolidate ownership of competing hospitals, eliminating competition between them.”

If you are going to protect the States’ right to organise regulated sectors in whatever way they please, you have to give a fair reading to their statutes. By applying the “presumption against state immunity” (p. 7) as drastically as it did, without properly considering what the Georgia legislature most likely intended, the Supreme Court made the state immunity doctrine a sham. 

Even more recently, the 4th Circuit backed the FTC’s decision against the North Carolina State Board of Dental Examiners, who had set out to prevent non-dentists from offering teeth-whitening services. Here, the basic legal approach is eminently sensible; being a private body consisting predominantly of dentists, the state board is asked to show that their behaviour was “actively supervised by the State itself”. (Cf. the 1980 Supreme Court ruling of California Retail Liquor Dealers Association v. Midcal Aluminum.) Since it wasn’t, the state board was nailed to the wall by the FTC and the court.

But again I’m not completely convinced by how it goes on. The problem is what the 4th circuit and the FTC are doing with regular competition law. After all, just because the state immunity doctrine of Parker does not apply, does not mean that the state board is automatically at fault. You still have to tick all the usual competition law boxes. And the box that trips them up, I think, is the one that we Europeans would call “justification”, and which in the US goes to the question of whether the “restraint of trade” in question is “unreasonable”, I requirement that is read into the Sherman Act in order to keep it from invalidating every contract ever (15 USC 1).

In the US, in order to decide that question, the courts first consider whether a “per se rule” applies – or as it is called now: a quick-look test – or whether the case should be treated under the rule of reason. Quoting the 1984 Supreme Court case of NCAA v. Board of Regents of the University of Oklahoma, the 4th Circuit explained:

The rule of reason applies “if the reasonableness of a restraint cannot be determined without a thorough analysis of its net effects on competition in the relevant market.”

The court does some hand waiving to argue that the distinction doesn’t matter much, before observing that

the Supreme Court has cautioned that we should be hesitant to quickly condemn the actions of professional organizations because “certain practices by members of a learned profession might survive scrutiny . . . even though they would be viewed as a violation of the Sherman Act in another context.” Nat’l Soc’y of Prof’l Eng’rs v. United States, 435 U.S. 679, 686 (1978).

So here’s my question: Where is this scrutiny? Immediately after this observation, the 4th Circuit quotes some more language that says the same, and then it “conclude[s] that substantial evidence supports the FTC’s factual findings regarding the economic effects of the Board’s actions and that those findings support the conclusion that the Board’s behavior violates § 1” (p. 31) and tells the state board to get the North Carolina state legislature to pass some legislation if there are legitimate public health concerns about non-dentists providing teeth whitening services (p. 32). 

This seems wrong to me. Surely it should be lawful for a group of companies to form a cartel if the alternative is to allow a market failure – in this case a threat to public health – to persist? If the public health threat truly exists, the net economic answer would be that the restraint of trade is the lesser of two evils, meaning that a rule of reason approach would conclude that there is no violation of the competition laws. And so it follows that the state board should be allowed to prove that this public health threat is real. I appreciate that the court’s role is to “uphold [the FTC’s] factual findings if supported by substantial evidence” (p. 26), but then at the very least I’d like for the court to say what that substantial evidence is. Instead, all I have is this, where, in its final opinion, we find the FTC categorically rejecting such a justification without even examining the facts. Instead, it quotes the Professional Engineers case already quoted above by the 4th circuit, where the Supreme Court said:

The Sherman Act reflects a legislative judgment that ultimately competition will produce not only lower prices, but also better goods and services. (…) The assumption that competition is the best method of allocating resources in a free market recognizes that all elements of a bargain—quality, service, safety, and durability—and not just the immediate cost, are favorably affected by the free opportunity to select among alternative offers. (…) The fact that engineers are often involved in large-scale projects significantly affecting the public safety does not alter our analysis. Exceptions to the Sherman Act for potentially dangerous goods and services would be tantamount to a repeal of the statute. In our complex economy, the number of items that may cause serious harm is almost endless.

Yeah, that’s just wrong. If you want to rule out self-regulation in favour of bureaucratic micro-management, that’s fine, but then you don’t get to call your approach a “rule of reason”.

Tuesday, July 02, 2013


In last month’s Akzo Nobel v. Competition Commission, the Competition Appeal Tribunal treaded a number of fine lines to end up with what would have seemed at first glance to be a no-brainer of a result. Yes, the Competition Commission (“the CC”) has jurisdiction to forbid Akzo Nobel from acquiring a company called Metlac, given its finding that the merger would result in a substantial lessening of competition in the UK.

Judging from the language of the judgment, the difficulty of this question seems to have surprised everyone involved. The legal trickery involved is as follows: While the familiar question is whether the competition authorities have the right to look at a merger in the first place, here the question is only whether the CC has the right to impose this particular remedy. After all, parties did not dispute that the proposed merger might affect market conditions in the UK (par. 21), and that the relevant turnover exceeded the £ 70 million threshold of section 23 of the Enterprise Act 2002 (“the Act”). No-brainer.

Things went horribly awry, however, when the CC decided to forbid the merger, relying on section 41 and section 84 of the Act. Forbidding, per se, was not the problem. Forbidding an anticompetitive merger is the most obvious remedy imaginable, which is why it is listed first in Schedule 8 of the Act. No, the problem was whether the CC had the power to forbid Akzo Nobel from performing an agreement, given that arguably neither that company nor the agreement was located in the UK. The key language is in section 86 of the Act:

86 Enforcement orders: general provisions
(1) An enforcement order may extend to a person’s conduct outside the United Kingdom if (and only if) he is—
(a) a United Kingdom national;
(b) a body incorporated under the law of the United Kingdom or of any part of the United Kingdom; or
(c) a person carrying on business in the United Kingdom.
(3) An enforcement order may prohibit the performance of an agreement already in existence when the order is made.
(4) Schedule 8 (which provides for the contents of certain enforcement orders) shall have effect.

Given that the parent company Akzo Nobel Holding, the legal entity that would perform the offending agreement, is incorporated in the Netherlands and has its headquarters there, the only possible way for the CC to get at it is to show that Akzo Nobel Holding is carrying on business in the United Kingdom, something it didn’t have to show in order to get jurisdiction over the merger as such, and something it wouldn’t have to show if its remedy had instead focused only on Akzo Nobel’s conduct within the UK. (In other words, if they’d chosen a behavioural remedy instead of a structural one.) And that is a problem, because – as one would expect given the name – Akzo Nobel Holding NV doesn’t really carry on any real business whatsoever. (Although given the definition of section 129 of the Act, it carries on business in the Netherlands. Cf. par. 111.) The holding company just owns shares, directly or indirectly, in about 450 subsidiaries, several of which do carry on business in the UK.

Which brings us, much to everyone’s great surprise, to Salomon v. Salomon, an 1897 case so fundamental that every British and Irish law student – and probably many more in the rest of the world – has to study it on the first day of Introduction to Company Law. In that case, the House of Lords was asked to say that Salomon the majority shareholder was liable for the debts of Salomon Ltd, given that the company in question was nothing more than a sole proprietor shoe maker doing business as a company with the six other shareholders being nothing more than placeholders. (Under the Companies Act 1862 a company had to have at least seven shareholders.) The Lords, however, declined to “pierce the corporate veil” and held that plaintiff person and the defendant (bankrupt) company were distinct entities, meaning that the liquidator of the company could not recover the company’s debts from Mr. Salomon.

And now, 116 years later, the CC needed some way around that doctrine in order to be able to say that Akzo Nobel Holding NV – as opposed to its subsidiaries – was carrying on business in the UK. The solution was to take refuge in an idea that suggests that no piercing of veils is in fact happening, the idea of the “Single Economic Unit”. Like so many bits of creativity in English law, this one comes to us ultimately courtesy of Lord Denning, although the CAT does not cite him. The CC itself seems to have been reluctant to go there explicitly (cf. par. 99), but the CAT seizes on it with abandon. It lists several pages of authorities to the effect that the rule that a Single Economic Unit can be treated at law as such does not exist because the whole story amounts to nothing more than piercing the corporate veil in a way that is forbidden by Salomon (par. 100-107) before concluding that, in this case, none of those concerns matter. 

It must be said that if ever there was a case where the Single Economic Unit argument fit like a glove, it was this one. Akzo’s de facto organisational structure involved an Executive Committee and a slew of Business Areas, Business Units and Sub-Business Units, none of which corresponded even remotely to the legal structure of the group (par. 49-50). Individual corporations tended not to have individual strategies, or even natural persons as directors or secretaries. But then again, much the same could be said for A. Salomon & Sons Ltd; while it at least had human beings for shareholders, its corporate strategy was determined entirely by its majority shareholder. By checking against the Salomon case, it becomes clear that the CAT comes dangerously close to distinguishing the Salomon precedent – and the many other precedents that build on it – on no firmer basis than that the Akzo case involves shareholders who are themselves corporations. And even though from an economics point of view, that might be sensible, legally it is extremely dicey. As the CAT itself said:

82. In our judgment, the appeal to the economic purposes of the Act and the apparent irony in that context of allowing technical legal concepts to limit the achievement of those purposes is, in the present context, misconceived. It is, of course, true that the subject-matter of the Act comprises the assessment and regulation of economic issues but that subject-matter is realised through a legally constituted framework of procedure and enforcement. (…) There can be no special dispensation from those general principles, in the absence of any statutory provision to the contrary, simply because the substance of the issues under consideration is economic.

There is another reason why the CAT’s reasoning might be considered weak. The quoted language comes from a discussion of why the jurisdictional basis for considering a merger in the first place should be different from the jurisdiction to forbid it. Essentially the Tribunal’s reply is that the different language used in the two provisions is quite deliberate, that Parliament had clearly made the distinction on purpose, and that therefore Parliament’s will should be respected: Not all mergers that are within the jurisdiction of the CC can be forbidden by the CC.

And yet, looking at this Single Economic Unit approach, it is not obvious that there are in fact many cases left that fit that description. As the CC itself concluded, Akzo’s governance structure is typical for large multinationals (cf. par. 66). And small multinationals that operate in a UK market will normally have a more direct presence there. So, realistically, does the CAT’s interpretation of section 86(1)(c) really preserve the difference with section 23? A disinterested reader may well conclude that the CAT in fact did the opposite of what it said it would do in par. 82: it put the economic logic of the case, and of merger review in general, before the law.

For this reason, if I were advising Akzo Nobel, I would suggest taking this case to the Court of Appeals. Given that this case touches on important issues of jurisdiction over mergers and company law, with the former being tackled in this case for the first time (par. 74), there is no reason why the CAT or the Court of Appeals should not give leave to appeal.

Thursday, June 27, 2013

This Week and Last Week in Luxembourg

The Grand Chamber (Judge Juhász) reaffirmed that in competition law the actual facts on the ground are all that matter, regardless of anyone’s intentions. In this case, both the lawyers and the Austrian Kartellgericht said that the cartel in question was permitted, but that does not prevent the Austrian authorities from fining them anyway 15 years later. Bundeswettbewerbsbehörde and Bundeskartellanwalt v. Schenker et al. Cf. European Law Blog

The Grand Chamber (Judge Ó Caoimh) also gave the Czech Republic a € 250.000 lump sum fine for failure to comply with an – otherwise boring – infringement judgment. Commission v. Czech Republic

Guillermo Cañas’s attempt to marshal the forces of EU competition law against the world anti-doping agency WADA and against the ATP failed before the Court of Justice (Judge Bay Larsen) as it had before the Commission and the General Court. The problem continues to be that the applicant, having retired, no longer has an interest in fact in the dispute. Cañas v. Commission (FR)

In Impacto Azul Lda v. BPSA 9 et al., the Court (Judge Lõhmus) held that art. 49 TFEU allows national legislation that “excludes the application of the principle of the joint and several liability of parent companies vis-à-vis the creditors of their subsidiaries to parent companies having their seat in the territory of another Member State”, because the parent can easily contract around this.

The Court (Judge Rosas) agreed with AG Mengozzi that Luxembourg had discriminated impermissibly against foreign students in its system for financial aid. However, the way it got there was quite different. Giersch et al. v. Luxembourg Cf. Eutopia Law blog

Now that all the easy cases on mutual recognition of professional qualifications are dealt with, it’s time to move on to more difficult situations. In Nasiopoulos v. Ipourgos Igias kai Pronoias we have a German-trained Greek medical masseur-hydrotherapist (‘Masseur und medizinischer Bademeister’) who wants to work as a physiotherapist in Greece. While the Court (Judge Levits) agrees that that is a bit of a stretch, it thinks he should at least be allowed to practice that part of the profession that he is actually qualified for.

In the joined cases VG Wort v. Kyocera et al. and Fujitsu and HP v. VG Wort, the Court (Judge Malenovský) gave some guidance on art. 5(2)(b) and 6 of Directive 2001/29, the copyrights directive. As it turns out, printer manufacturers can be sued for some of the total “fair compensation” owed for all those naughty internet users printing off books in their attics, but not all of it.

The Court (Judge Jarašiūnas) signed off on a Maltese excise duty on mobile telephone use, concluding that neither art. 12 nor art. 13 of the Authorisation Directive applied to such a “consumption tax”. Vodafone Malta et al. v. Avukat Ġenerali et al.

AG Jääskinen, quoting pre-Supreme Court Louis Brandeis, argued that there is no general “right to be forgotten” under existing EU data protection law. The defendants wanted an allegedly incorrect search result deleted from Google. Google v. Agencia Española de Protección de Datos (AEPD) and Mario Costeja González Cf. UK Human Rights Blog and European Law Blog

AG Mengozzi has a state aid case in national court, where Deutsche Lufthansa complained about alleged state aid from Frankfurt-Hahn airport to Ryanair. As a result of this litigation, the Commission decided to get interested, with the result that the standstill clause of art. 108(3) TFEU came into effect. Given that the German court doesn’t necessarily agree that there is unlawful state aid here – the case was initially rejected by the Landgericht – the question is what the distribution of responsibilities and obligations is between the Commission, the national court and, potentially, the ECJ. Deutsche Lufthansa v. Flughafen Frankfurt-Hahn (NL, DE, FR)

For whatever reason, they let AG Kokott near one of those classic legal basis & common commercial policy cases. (Cf. my LL.M. thesis, long ago, here.) The case is about this Council of Europe convention. The Commission wants the EU to ratify it based on the normal rules of the common commercial policy under art. 207 TFEU, while the Council prefers a mixed agreement based on art. 114 TFEU. Curiously, the AG argues – correctly – that art. 3(2) TFEU codifies the ERTA doctrine, but then uses that to conclude that the EU’s competence in this area is not only exclusive, but also based on art. 207 TFEU. Commission v. Council

The General Court rejected two action for annulment in the Aluminium Fluoride cartel case. Most creatively, one of the applicants – ICF from Tunisia – tried to plead art. 36 of the Euro-Mediterranean agreement between the EU and Tunisia from 1998 as a grounds of invalidity. Unsurprisingly, the Court did not go for that one. ICF v. Commission (FR) and Fluorsid SpA and Minmet Financing v. Commission