Thursday, April 25, 2013

This Week in Luxembourg

The Grand Chamber (Judge Ilešič) slapped down the appeal by Laurent Gbagbo and his friends against the General Court’s order dismissing their action for annulment of their asset freeze as manifestly unfounded. The problem was that the applicants were out of time. The Court discussed the general theory of binding time limits before concluding that Gbagbo c.s. hadn’t argued anything specific to substantiate their appeal to force majeure. Gbagbo et al. v. Council

The Romanian gay rights NGO Accept decided to go for broke. They brought an employment discrimination case under Directive 2000/78 against Steaua Bucharest. If that wasn’t ambitious enough, the factual basis of the claim was an interview given by Gigi Becali, the main shareholder and “patron” of the club, which may or may not be enough to consider his statements as representing the club. All three answers of the court favour the applicant. Asociaţia ACCEPT v. Consiliul Naţional pentru Combaterea Discriminării Cf. Journal du Marché Intérieur Blog
In Jyske Bank Gibraltar v. Administración del Estado, Gibraltar banking secrecy clashed with Spanish and EU legislation to combat money laundering (specifically, Directive 2005/60) and the war on money laundering won.

Because Galileo is organised, under Regulation 876/2002 as a Joint Undertaking, i.e. an EU-level PPE, the EU’s staff regulations do not apply. Bark v. Galileo Joint Undertaking

(There was also a whole stack of boring and easy infringement cases.)

In the General Court, the baby seals won in the suit brought by the Canadian fur producers. The ban on seal products was lawfully enacted on the basis of art. 95 EC. The General Court rejected art. 133 EC as an additional legal basis by relying on the titanium dioxide case. The Court also spent some time on proportionality & subsidiarity, on art. 1 P1 and on the United Nations Declaration on the Rights of Indigenous Peoples, but to no avail. Inuit Tapiriit Kanatami v. Commission Cf. European Law Blog

Sunday, April 21, 2013

Wifi (2)

Last month, those b*stards at the Dutch Supreme Court ruined one of my favourite legal stories - the one I discussed in this blog post from 2011. What had happened is that the Court of Appeals in The Hague had decided that you couldn't - legally - hack a router, because a router was not sufficiently similar to a computer to qualify under the relevant provision of the Penal Code. The fact that the router at issue in the case was protected with a password was irrelevant. The defendant was (partially) acquitted.

Obviously, this is awesome on countlessly many levels, which is why I've used this case from time to time pour encourager les autres. Especially in criminal law, where the principle of nulla poena sine lege reigns surpreme, the law must be interpreted carefully, based first and foremost on the (plain) meaning of the words used by the legislator. Just because something is, in the words of the Court of Appeals, "socially undesireable", that doesn't mean the Court should therefore stretch the meaning of the law beyond its plain meaning.

For what follows, however, it should be noted that the statute in question is not exactly a paragon of clarity. For the Dutch readers, it says the following:
Art. 138a lid 1 (oud) Sr
Met gevangenisstraf van ten hoogste een jaar of geldboete van de vierde categorie wordt, als schuldig aan computervredebreuk, gestraft hij die opzettelijk en wederrechtelijk binnendringt in een geautomatiseerd werk of in een deel daarvan. Van binnendringen is in ieder geval sprake indien de toegang tot het werk wordt verworven:
a. door het doorbreken van een beveiliging,
b. door een technische ingreep,
c. met behulp van valse signalen of een valse sleutel, of
d. door het aannemen van een valse hoedanigheid.
Art. 80sexies Sr
Onder geautomatiseerd werk wordt verstaan een inrichting die bestemd is om langs elektronische weg gegevens op te slaan, te verwerken en over te dragen.
I'm not even going to attempt a translation. Suffice it to say that the list in the end is clear enough. It lists four things that at least qualify as hacking. The problem, both generally and for this case, is the term "geautomatiseerd werk", which I've never heard or seen anywhere else before or since, and which is clearly meant to refer to computers and such like without actually using the word computer or any other word that might be meaningful to the average Dutch speaker. Hence the problem of working out whether a router qualifies. Even the definition of art. 80sexies, which says that a "geautomatiseerd werk" is a tool used for storing, processing and transferring data electronically doesn't help as much as one might like.

Therefore, the Court of Appeals, like a good Continental court, looked at the legislative history and found that the intention of the legislature was particularly to protect "those who, by using security measures, have indicated that they wished to shield their data from nosy intruders". Given that a router doesn't hold data, the court argued that the legislative history supported the view that they don't count.

Last December, Advocate-General Vellinga had fairly little difficulty supporting this view. After summing up the history of the case and the law, he simply noted that courts are allowed to base their interpretation of the law on statements made in parliament at the time the law was enacted and, citing the same passage from the parliamentary transcripts as the Court of Appeals, he concluded that that court had been correct.

(Interesting side-note: the Attorney-General's appeal brief and the Advocate-General's opinion on this point were both formulated in terms of the Rechtsgut that was allegedly protected by art. 138a Sr. That's a topic I encountered only a few days ago talking about the possibility of a criminal prosecution of Thilo Sarrazin in Germany on the Volokh Conspiracy.)

The High Council (= Supreme Court), however, quoted more generously from the legislative history, focusing particularly on the word "inrichting", which I translated loosly as "tool" earlier. The Court deduced from the transcripts it quoted that the law was intended to protect not only separate devices, but also networks and parts of networks. In other words, while the three activities of storing, processing and transferring data are cumulative, they do not have to be met by each and every single potentially hackable device, only by the totality of the network or group of devices that the device in question belongs to. It follows that routers, too, can be hacked.

While this is a plausible reading, I still think that the Court of Appeals' approach is less forced. After all, the High Council did not address the Rechtsgut problem: Which interests, exactly, was art. 138a Sr enacted to defend? Just because the High Council quoted legislative history does not mean that they engaged in teleological interpretation. They didn't, because they didn't say a word about the intentions of the lawmaker. In fact, if I had to categorise their approach with one of the classic models of legislative interpretation, the only thing I can think of is plain textualist interpretation, except that they used the legislative history instead of a dictionary to discover the meaning of "inrichting". That, I would think, is not something we would want to encourage. If we are going to focus on the meaning of the words in the statute, we should rely on the meaning that ordinary Dutch speakers give those words, or the meaning that can be found in a dictionary. If the legislature wants to deviate from the normal meaning of words, they should write a definition into the law. The fact that they did so here, but not very well, should be no concern of the courts.

Moreover, I think this result violates the principle of in dubio pro reo. Given that there were two plausible interpretations of the statute, the Court should have preferred the one that favoured the defendant. When in doubt, no one goes to jail.

P.S. To clarify: Given that using your neighbour's unsecured wifi does not involve doing any of the four things specifically listed in art. 138a (oud) Sr, nor anything comparable, I would imagine it is still legal. That said, you should consult your lawyer before engaging in anything that is potentially a crime, just to be sure.

Thursday, April 18, 2013

This Week in Luxembourg

The Grand Chamber handed down two big cases this week:
On the one hand, it decided not to shoot down the European Patent Court a second time. (Cf. Opinion 1/09) Curiously, Spain and Italy tried to use art. 3(1)(b) TFEU to argue that the EPC deals with an exclusive Union competence and that the EPC somehow didn’t promote integration because it excluded two Member States. More sensibly, but also unsuccessfully, they relied on the last resort requirement and the reasoning of Opinion 1/09 (par. 89-94). Joined Cases Spain and Italy v. Council Cf. IPKat (1, 2) and Verfassungsblog

On the other hand, it crushed the tentative Belgian language peace by forbidding a Flemish rule that required all employment contracts to be drafted in Dutch. Las v. PSA Antwerp Cf. the blog Journal du marché intérieur and Euractiv Cf. Eutopia Law Blog and Außenwirtschaftslupe

Following their fairly poor results in the first batch of railway infringement judgements in February (cf. Commission v. Germany here), the Commission will be pleased with its (partial) win against France. The Court (Judge Borg Barthet, of course) found against France with regard to capacity allocation and the absence of mandatory performance elements in the access charges. Commission v. France

Two appearances of the principle of effectiveness this week: In Irimie v. Administraţia Finanţelor Publice Sibiu the Court (Judge Ilešič) insisted that Romania should not let procedure get in the way of recovering all of an unlawfully levied tax, and in L v. M the Court (Judge Bay Larsen) protected the Environmental Impact Assessment Directive 2001/42 against some German procedural problems.

The Court (Judge Tizzano) followed AG Cruz Villalón and held that the dispute between Systran and the Commission is contractual, rather than tort. As such, it belongs in Luxembourg court. Commission v. Systran

Levi’s is trying to prevent another jeans manufacturer from attaching red labels to its jeans by using a number of its trademarks. As far as I can tell, they won. Colloseum Holding v. Levi Strauss & Co (Judge Juhász) Cf. IPKat

The somewhat creative Dutch Essent privatisation litigation has, predictably, ended up in Luxembourg. AG Jääskinen argued that the ban on privatisation is shielded under art. 345 TFEU and therefore OK. As to the ban on TSOs being part of wider energy conglomerates, the AG concluded that this was a justified limitation of the free movement of capital. The Netherlands v. Essent, Eneco and Delta (NL, DE, FR)

AG Wahl has a nice case where the place where the place where the work is habitually carried out is not the place most closely connected to the contract. (Cf. art. 2 and 6 Rome Convention) Schlecker v. Boedeker

In Martin y Paz Diffusion v. Depuydt and Fabriek van Maroquinerie Gauquie AG Cruz Villalón discussed the case of a trademark owner who was naughty in a way that may or may not have been illegal. The plaintiff withdrew his consent for the defendants to use his trademark in a way that the AG concluded was (probably) OK; consent cannot be irrevocable and abuse of right law cannot provide a permanent remedy here. Cf. art. 5(1) of Directive 89/104.

Even though the Commission is already litigating as a victim in Belgian court, the appeal of one of the elevator cartel cases is still pending. AG Kokott recommended this week that Schindler should lose its appeal. (Its complaint was about the way the Commission and the General Court had treated it like a single entity for competition law purposes.) Schindler Holding v. Commission (DE, FR)

AG Kokott is also potentially saying something interesting about public procurement law & structural funds in a case about a Club Med village on Martinique. The Commission had taken the tough approach, and the AG now proposes annulling the General Court’s decision that sided with the Commission. Now let’s see what the Court does. France v. Commission (NL, DE, FR)

Germany v. Puid is one of those cases that give lawyers a bad name. According to AG Jääskinen, an asylum seeker does not have an enforceable claim for a Member State to use its authority under the first sentence of art. 3(2) of Regulation 343/2003 to examine a claim it doesn’t have to. However, when “a national court cannot be unaware” of the deficiencies of the asylum system of the country that is supposed to examine the claim, it is still required to suspend the transfer of asylum seekers to that country under N.S. and M.E. Sounds right to me…

Friday, April 12, 2013


In a recent judgment, the Supreme Court of the Netherlands held against “The Board of the District Court of [A]”. I would humbly submit that that was taking anonymisation several steps too far.

Discussing the question of anonymity in court documents with my friends on the Volokh Conspiracy, I have generally defended the way the balance is struck in the Netherlands between privacy and transparency. Free speech is not an issue: there is no law forbidding anyone from publishing the names of people convicted of a crime. There is merely an agreed upon custom among the press that people’s surnames are not mentioned. Courts, in their judgments, also don’t mention the names of the individuals before them – although they do mention company plaintiffs and defendants by name – but that is not a free speech problem. Free speech does not require that the government give you information to talk about.

So that leaves transparency. The logic here is essentially the same as for Access to Documents legislation. As the EU legislature put it in recital 2 of Regulation 1049/2001:

Openness enables citizens to participate more closely in the decision-making process and guarantees that the administration enjoys greater legitimacy and is more effective and more accountable to the citizen in a democratic system. Openness contributes to strengthening the principles of democracy and respect for fundamental rights as laid down in Article 6 of the EU Treaty and in the Charter of Fundamental Rights of the European Union.

Specifically in the context of court judgments, it is claimed that transparency allows citizens to know which of their fellow citizens have done wrong, and that it allows them to scrutinise the work of the court better. Against this we put the privacy rights of criminals, and more generally of parties to litigation. We have abolished the pillory as a form of punishment, and it does not do to recreate it in the press.

Based on these considerations, I have generally argued that the right balance is struck by giving the press and the general public free access to the court room – absent special considerations requiring confidentiality of proceedings – and thus allowing them to see and report trials in a non-anonymous way. If the public interest requires that the names of litigants are reported, the press will know these names, and will be free to make that decision, but I can’t just Google my neighbour to see if he’s ever been convicted of anything just because I’m nosy.

The reason why I was thinking of this is the rather curious case of Karel Knip. (His alleged misconduct is at several removes from the case I’m interested in, so there’s no reason not to mention his name.) Karel Knip wrote about the Trafigura case, and did so in a way that noticeably favoured the company in the middle of what was generally highly bad press. While this might lead the neutral observer to question his integrity, the District Court in Amsterdam, in handing down its judgment in the Trafigura case, went well beyond that and took the highly unusual step of mentioning this journalist – who, mind, was not a party to the case – in its judgment. They wrote:

6.10. (...) Trafigura toont daarmee dat zij geen enkel vertrouwen heeft in de rol van de media en de oprechtheid van de journalisten, zij het dat zij één uitzondering maakt en dat is voor de journalist [naam 46]. Hij is de enige die het begrijpt en die - naar de rechtbank moet aannemen met behulp van informatie van Trafigura - tijdens deze strafzaak en nà het pleidooi van Trafigura nog eens komt met een artikel waarin de relatieve onschuld van de slops wordt beschreven.

My translation:

6.10. (…) Trafigura shows in this way that it has no confidence whatsoever in the role of the media and the sincerity of journalists, be it that it makes one exception, for the journalist [name 46]. He is the only one who understands and who – as the court is supposed to assume with the aid of information provided by Trafigura – during this criminal trial and after the concluding remarks of Trafigura brings an article that describes the relative innocence of the slops in question.

(The brackets are the court’s, of course. A slop is technical jargon for a chemical spill.)

The daily newspaper Trouw noticed this, and reported about what the court had said. Paraphrasing the judgment, the newspaper said that the court had questioned Mr. Knip’s integrity. Predictably, he did not like this characterisation, so he took action against the paper before the Press Council, an independent body tasked by the Dutch press with enforcing journalistic best practice. (Note that it is emphatically not a public law body. Even in the Netherlands we like to avoid litigating press stories.)

This is where we get to the second unusual event in this saga. The Council, in examining the complaint, got in touch with the court, asking it whether the newspaper’s interpretation of the judgment was correct. The clerk of the court answered that it was a reasonable interpretation. Upon receiving this answer, Mr. Knip abandoned his case before the Press Council.

He did wonder, as well he might, why on earth the clerk of the court replied to this question. So he made a complaint with the President of the District Court, who backed the clerk, and with the Council for the Judiciary, the governing body of the Dutch court system. The latter replied that this was not a question of judicial misconduct, but one of alleged misconduct in the management of the courts, and therefore someone else’s problem.

Given that this is not a very satisfactory answer, Mr. Knip decided to bring out the big guns by bringing a complaint under art. 13a of the Judiciary Act (Wet op de Rechterlijke Organisatie), a new procedure that was only created in 2011. This involves asking the Procureur-General at the Supreme Court, a magistrate whose independence is guaranteed in the same way as the independence of judges, and whose job includes, in theory, bringing prosecutions against members of the government for abuse of office, to petition the Supreme Court asking that Court to investigate the matter. (Since there has never been a prosecution against a Minister, the Procureur-General normally spends his days as the boss of the Advocates-General at the Supreme Court, joining them in advising the Supreme Court about the best resolution of the cases before it.)

The Procureur-General granted the request and brought his petition, and the Supreme Court heard the Amsterdam court at a hearing in January. (This being a civil law country, there were no spectacular Law & Order theatrics. Instead, the Amsterdam court made a statement via counsel, and Mr. Knip declined to make an appearance.)

Last month, the Supreme Court handed down its judgment in this case, and it was a doozy. The Amsterdam court was reprimanded severely for having allowed the clerk to respond to the Press Council’s question, and for supporting the clerk’s decision afterwards. Judges speak through their judgments, and if any commentary is necessary there is a judge-spokesperson for every court. And the Supreme Court stressed that even that person, who is himself a judge, should be careful not to create the impression that his statements are in any way authoritative. The way the Amsterdam District Court handled this matter could not have been more wrong.

There is just one odd thing about the Supreme Court’s judgment. At the top, it is styled: “Judgment in the matter of the complaint brought by [complainant] against the Board of the District Court of [A]”, and that same style is maintained throughout. The fact that Mr. Knip’s name was removed is entirely normal, and in line with Dutch court practice. The fact that the reference to Amsterdam was also removed is less normal, and also much less defensible. Not only would the custom with regard to companies suggest differently, but so would the underlying first principles. If ever there is a strong case for transparency, it is when the judiciary is trying to clean its own house.

As it happens, there is only one District Court that starts with an A, making this case of anonymisation particularly silly. (There used to be more, but several District Courts were merged in recent years.) While the board of the court is strictly speaking a group of individuals, the same goes for the board of a corporation, such as the board of Trafigura who ended up with their company name printed for all to see in the jurisprudence. As long as the judges in question don’t end up with their names printed in the judgment, I don’t see how the normal rules for anonymisation require that the name of the court be removed from the judgment.

Even if, applying the normal rules, one were to say that this is a borderline case, surely the public interest in observing how the judiciary polices itself is greater than the public interest in transparency normally would be? In other words, the Supreme Court should have erred on the side of naming the board of the Amsterdam District Court.

In order to remedy that omission, I will take the liberty of naming the current members of the board. Board members for courts are appointed for six year terms, so I cannot be sure that these judges were already on the board at the relevant time, but it will have to do. They are:
  • Mrs. C.T.E. Eradus, President of the Court and chairwoman of the Board;
  • Mrs. E. de Rooij, judge-member.
  • Mr. Hans Janssen, non-judge member appointed as of 1 April 2013.

This blog post is based on this Dutch-language story by Hugo Arlman on The opinions expressed here are, of course, my own.

Thursday, April 11, 2013

Today in Luxembourg

The case of Mindo v. Commission (judge Arabadijev) gave the Court the opportunity to explore the consequences of joint and several liability for competition fines when one of the persons liable is bankrupt. It held that the General Court was wrong to ask Mindo to prove that the other creditor, AOI, actually intended to recover part of the fine from Mindo. More embarrassingly for the General Court, the main focus of the judgment was the failure of the GC to state adequate reasons.

In the Aarhus case of R. (Edwards and Pallikaropoulos) v. Environment Agency et al., the Court (judge Bonichot) tried to give some usable guidance as to the rule that judicial proceedings in this area of the law should not be “prohibitively expensive”. It concluded that the applicant’s ability to pay is not the only criterion. It also listed:
  • The reasonableness of the costs in isolation;
  • “The situation of the parties concerned”;
  • The claimant’s “prospect of success”;
  • “The importance of what is at stake for the claimant and for the protection of the environment”;
  • “The complexity of the relevant law and procedure”;
  • “The potentially frivolous nature of the claim at its various stages”;  and
  • “The existence of a national legal aid scheme or a costs protection regime”.
It turns out that the fact that the claimant has not, in fact, been deterred is not dispositive. Also, the criteria have be consistently applied regardless of the stage of proceedings at which the question is considered. Cf. UK Human Rights Blog and GAVC Law Blog

For prof. Heldeweg, if he still has time for such things, the Court (judge Arestis) has a case on the precautionary principle and the Habitats Directive. Sweetman et al. v. An Bord Pleanála

The Court (judge Toader) held that Regulation 44/2001 is applicable to a claim for recovery of a sum unduly paid by a public entity in the context of Nazi-era reparations. The Court also devoted some attention to the question of whether it is necessary for various co-defendants to live outside Germany. Land Berlin v. Sapir et al.

Berger v. Bavaria dealt with food that was unfit for human consumption but not a health risk. The Court (judge Bay Larsen) interpreted art. 10 of Regulation 178/2002 to permit national legislation that allowed for intervention in these circumstances. Note the – somewhat hypothetical – question 2 posed by the national court, which the Court unfortunately ignored.

In Jeltes, Peeters and Arnold v. UWV, the Court (judge Fernlund) held that its judgement in Miethe v. Bundestanstalt für Arbeit was no longer good law in light of the enactment of Regulation 883/2004.  The plaintiffs were all atypical frontier workers in the sense of Miethe because they lived in Belgium or Germany while maintaining a professional and social life almost entirely in the Netherlands.

This week, the Czech Republic gives us a particularly egregious case of gender discrimination by having a law that lets women retire younger than men, and making their retirement age dependent on how many children they’ve raised. Soukupová v. Ministerstvo zemědělství

In his opinion in Commission v. the Netherlands (NL, DE, FR), AG Wathelet offered not one but two reasons why the Commission should lose. The boring one is that the case concerns a purchase of land, not an award of a concession. More interestingly, he discussed the applicability ratione temporis of Directive 2004/18 to this case, which concerned decisions taken at various moments before and after the enactment and the entry into force of that directive.

While Regulation 44/2001 does not apply to Denmark, there is a parallel international agreement between the EU and Denmark that brings it back in through the back door. AG Kokott did some legal interpretation ninja to explain why the Danish court is allowed to ask a prejudicial question notwithstanding the text of the agreement before arguing that the Regulation applied to the case at bar. Her Majesty’s Revenue & Customs v. Sunico et al.

AG Cruz Villalón argues that, for once, the Turkish plaintiff should lose in an EU-Turkey association agreement case. Specifically, he argues that the passive freedom to provide services is not covered by the Agreement, and in the alternative he argued that the plaintiff’s reliance on that freedom was too speculative, given that she only really wanted to get around the visa requirement in order to visit her family in Germany. Demirkan v. Germany Cf. European Law Blog

AG Mengozzi disentangled a case where Germany refused a visa on the grounds that the applicant was likely to overstay. Cf. art. 21(1) Visa Code. Koushkaki v. Germany (NL, DE, FR)

In other news, I came across the Commission’s notification letter concerning its decision to open an art. 108(2) procedures against a number of Dutch football clubs, including PSV.

Tuesday, April 09, 2013

The Plague Reaches EUObserver

Repeat after me: Not being able to win a referendum is not "undemocratic". The are many things that you'd never win a referendum on. Taxes. Piers Morgan. Quantum Physics. But that doesn't make any of these things undemocratic. Instead, we have democratically elected legislatures, the marketplace and scientific consensus to decide on these things instead, respectively. Can we please, please, please stop pandering to the electorate by pretending that referendums are the only form of true democracy, or even the best kind?

P.S. I should emphasise that the Ph.D. thesis by Jens Peter Paul that is discussed in this article does not suffer from the same problems, which is actually rather important given that it deals with the theory of democracy (and its practice). Without having read the whole thing, the statement of results (p. 248-252) seems to imply the following assumptions about a well-functioning democracy, which seem correct:
  1. In a healthy democracy, there is a (high-quality) public discourse about important decisions, in the form of newspaper articles, interviews in talkshows, etc.
  2. In a healthy democracy, members of the legislature do not feel obligated to rule out one or more alternatives.
  3. In a healthy democracy, parties position themselves in the debate based on their substantive preferences rather than power-seeking behaviour.
  4. In a democracy, election results can only be said to legitimate decisions taken in the previous legislative period to the extent that those decisions were prominent in the election campaign.
  5. -
  6. In a healthy democracy members of the legislature are not exposed to undue pressure.
Comparing these ideals to what actually happened, the qualification "undemocratic" might not be that far off the mark. However, that's not a conclusion you can draw based on what the article says, which focuses mainly on what Kohl said - being interviewed by the author of the thesis - about the unpopularity of the euro.

(Referendums don't play a prominent part at all in the thesis - the word on appears a handful of times in the main body. Funny how reporting on science works...)