Who is steering the jurisprudence of the European Court of Justice? The influence of Member State submissions on copyright law

The complex and elusive structure we call ‘Europe’ is the result of multiple cultural, economic, social and political conditions under which Law, and in particular the jurisprudence of the Court of Justice of the European Union (CJEU), is performing a unique integrating role. The juridification of the European policy process is increasingly fragile, and little understood. This study extends a novel methodology developed in our previous work (MLR 2016) that revealed the workings of the Court through empirical analysis of the disciplinary background of judges and topoi engaged in their reasoning. Our new study turns the attention to the influence of Member States on the rulings of the Court. We focus again on the domain of copyright law in which we have seen a dramatic escalation of preliminary references to the Court, suggesting a normative void. Examining 170 documents relating to 42 cases registered between 1998 and 2015, we measure empirically the impact of submissions by Member States and the European Commission on the legal interpretation of copyright concepts. The findings show that France is the most influential country by some distance, both in terms of the number of interventions (an ‘investment’ in policy) and in terms of persuasive power (France’s arguments – 69% in favour of rightholders – are more often adopted by the Court than any other country’s). Surprisingly, three smaller countries are in the next group of strong influencers: Finland (slightly skewed pro-rightholders), Portugal (strongly user friendly) and Czech Republic (moderately rightholder friendly). Other countries appear to have very specific interests, intervening less often and in a targeted manner. They may be influential despite lower participation. One of the most effective governments arguing for the interests of copyright users is the United Kingdom. Our evidence suggests that the departure of the UK from EU litigation has the potential to disturb the delicate balance of European copyright jurisprudence.


Introduction
The Court of Justice of the European Union (CJEU) plays a role quite unlike any other international court in what has been called "a notable juridification of the European policy process" 1 .In a time of potential disintegration, it has become more important than ever to understand how the Court performs its integrating role in a contested political environment.
This requires an analysis that does not rely on the assumption of legal autonomy.We need to understand the forces that shape the behaviour of the Court of Justice beyond an analysis of the case law and the Court's evolving jurisprudence.
Political science treats the possibility of autonomous action of the Court with scepticism, often suggesting that the Court's rulings are determined by the anticipated reactions of national governments 2 .In this study, we develop an empirical approach that uses the window of Member States' written observations to investigate the influence of governments on the jurisprudence of the Court in one specific subject domain: copyright law.
In her pioneering research on the behaviour of Governments before the Court, Marie-Pierre Granger 3 argues that carrying out a study to measure the impact of Governments' written observations was not feasible.It is certainly true that since written observations are 'confidential', and are no longer published in the Court proceedings, this task can appear daunting.However, in order to make a significant advance in understanding the role of national governments in relation to the Court of Justice, these obstacles need to be overcome.
Our core research question was initially anchored in the fast moving domain of copyright law because the CJEU seemed to fill a normative void, with a dramatic escalation of cases. 4We were concerned with the possibility that the jurisprudence of the Court was open to capture, and that national governments may be able to steer the Court to produce policy outcomes that were politically unachievable.
Even though there are certain limitations to capturing all Governments' written observations verbatim, we show that it is possible to assemble sufficient data to establish trends and construct a number of possible scenarios.With the assistance of supportive Agents before the Court, the Court Registry, and Freedom of Information requests (both under Regulation 1049/2001 and national legislation), we have been able to produce a near comprehensive database of interventions and outcomes in 42 copyright cases before the Court of Justice.This allows a considerable advance in our knowledge (both methodologically and in substance) about the functioning of jurisprudence in the complex and elusive structure we call 'Europe'.For the first time, it is possible to see how Member States and the Commission are pushing the Court in a specific subject domain.Drawing on this picture, we can also simulate different scenarios prompted by Europe's unique geopolitical context, including the departure of the United Kingdom from the European Union.
Our earlier study tried to lift the lid on how the CJEU works as a social body making copyright law.The focus of this new research moves from the judiciary to the role of governments in shaping the copyright jurisprudence of the Court.
There are considerable resources available from the political science literature how to approach this question.In particular, empirical socio-legal research has suggested that while various organisations can use strategic litigation (preliminary references) to steer judicial policy 6 , Member States can use their intervention via written observations for the same purpose 7 .According to some research, governments use participation in litigation to try to obtain favourable decisions 8 because they understand that they can influence EU jurisprudence through well-written and well-argued observations 9 .A fortiori, governments have realised that with successful interventions before the Court they can obtain the reversal of European policies that they opposed without success in the EU Council. 10 However, other empirical research has suggested that the European judiciary is constrained 5 Fa va le, Kretschmer, Torremans, "Is there a EU Copyright Jurisprudence?An empirical a nalysis of the workings of the European Court of Justice" 79(1) MLR (2016) 31-75. 6Ra s mussen, On Law and Policy in the European Court of Justice (Ma rti nus Nijoff, 1986) a t 270. 7See Granger 2004, s upra note 3, a t 1. 8 Wei ler, "The Transformation of Europe" (1991) 100 Yale Law Journal 2403. 9Col l ins, "Representation of a Member State before the Court of Justice of the European Communities: practice in the Uni ted Ki ngdom" 27 E.L.Rev.(2002) 359. 10 Gra nger, 2004, s upra note 3 a t 9. in its rulings by fear of overturning or disapplication from the referring Member State, and it acts accordingly. 11ile the present study focusses on copyright case law, its methodological approach can be applied to any area of law before the CJEU.Indeed, the systematic analysis of briefs and submissions before a court is a promising and fast developing field of enquiry, opening a new perspective on the workings of any judiciary.
After a brief overview of the copyright literature discussing the normative role of the CJEU and the way it addresses recurrent litigation on particularly controversial topics, the article empirically analyses the Court's copyright jurisprudence, to provide measurable answers to a number of instrumental sub-questions: 1. What are the most contested legal concepts in EU Copyright Law? 2. Who are the Governments interested in shaping copyright jurisprudence and, specifically, on which legal concepts are they intervening?3. What interests are Governments supporting (e.g.rightholders' or users')? 4. To what extent are Governments successful in steering the Court towards their interpretation of legal concepts?
The article proceeds as follows.Having set the context (Section 1), our methodological approach will be explained and justified (Section 2).Section 3 will provide the conceptual framework for analysing the normative role of the Court, using the example of the concept of 'communication to the public' to illustrate the implications.Section 4 will identify the legal concepts that recur most frequently in copyright litigation; this we assume is outlining a normative void.Section 5 identifies the countries that are most active in copyright litigation.Section 6 presents and discusses our central empirical findings, quantifying the influence of each country before the Court and their relative position in favour of rightholders or users.Finally, Section 7 offers a conclusion, interpreting the findings for the 'production of Europe'.

Methodology
This article extends methods used in our previous work and in previous social science research.Marie-Pierre Granger12 was first to study the litigation strategies of Governments intending to have an impact on the jurisprudence of the Court of Justice.Her dataset includes the number of governments' written observations, the areas of intervention, and the number of interventions in preliminary references filed by a court of the same government in the years between 1995 and 1999.She used qualitative methods including structured questionnaires and interviews of Agents acting for Governments.
Carrubba measured the impact of governments' observations on ECJ rulings involving the government as a litigator.In particular, the work measured how the threat of noncompliance (in national courts) or overriding (in the Council) influences the Court's decisions.
To this end, he employed text analysis (coding) of legal documents (cases between 1987 and 1997) and statistical analysis, including regression modelling of binary response variables (probit model).The document analysis involved coding the ruling and the written observations by capturing the preference of the government on each individual legal issue. 13amér and colleagues 14 also use content analysis (coding) of governments' written observations and preliminary reports 15 before the CJEU between 1997 and 2008, and descriptive statistics.Unlike other research 16 they also break down the observations into legal arguments.The most interesting finding of their work is that the litigation behaviour of Member States is consistent with their government's economic model and therefore with their mainstream policies.Cramér et al.'s research also codes the relative position of countries as 'pro EU Integration' or 'pro National Sovereignty'. 17 In our study, we use content analysis of documents (coding), a structured questionnaire recording Government's suggested answers (followed by unstructured feedback from agents and civil servants on our preliminary findings) and basic statistical computation of the data.
The research design progressed in the following steps: First, doctrinal analysis was used to identify groups of preliminary references demanding clarifications from the Court on the same copyright concept (for example, 'communication to the public').The dataset, consisting of all 78 copyright cases registered between 1998 and 2015 (and concluded before the Court) was clustered on the basis of the main legal concept examined by each preliminary reference.The results show that more than half of all preliminary copyright references were filed on only five concepts, which therefore received more attention from the CJEU, indicating a potentially contested policy direction.
Secondly, the study employed content analysis (expert coding) on these 42 selected cases dealing with 'recurrent' concepts, to identify the arguments exposed in the written observations from Member States and the EU Commission, and their acceptance or rejection in the final ruling of the Court.
Thirdly, statistical analysis was deployed to establish relationships between the outcome of the case and the written observations.The analysis captures a) the interests of governments 13 Ca rrubba, 2008, s upra note 2 a t 440. in a particular legal concept; and b) the correlation of their submissions on the interpretation of individual copyright concepts with the outcome of the Court's decision in each case.Together they produce a measure of influence for each Member State.
To explore possible explanations for the observed patterns, finally, these findings were circulated among Court Agents and Civil Servants for feedback.This qualitative element allows a more subtle understanding of behaviour, and shapes the interpretation offered in our Conclusions.

Sample construction
Our sample includes only preliminary references, in line with all cited previous research, because these are central to the question of juridification of the European copyright policy process. 18Within our initial population of all 78 copyright cases registered between 1998 and 2015, we identified 42 cases in which there was a 'recurrence' of the same legal issue (which we defined as five or more references to the CJEU on the same concept).The justification for reducing the sample to cases that dealt with concepts that were subject of repeat references is twofold.Pragmatically, it produced a manageable data set.More pertinently, the prevalence of certain legal issues within copyright law points to a normative void, i.e. a regulatory space within the acquis communautaire that stakeholders and Governments are trying to fill with alternative policy paths.This could be seen as an alternative to legislative action, or as an opportunity to correct earlier legislation or case law. 19r these 42 'strategic' cases, 170 case documents were compiled, consisting of the published court documents (Advocate General's Opinion and the Court's Ruling) and the written submissions by the European Commission and Member States that chose to intervene.Access to the content of 'written observations' was sought and obtained through contacting and interviewing Court Agents and Civil Servants, where possible.See note *, a cknowledgements. 21The l ack of transparency a bout Member States' a nd the Commission's i nterventions before the Court is troublesome, and wi l l be the subject of a separate a rticle.There should not be a need for freedom of information requests nor s ubterfuge (for

The normative role of the CJEU in the copyright literature
There is consensus in the literature that Court of Justice is playing an increasingly central role in shaping not only the nuances but the contours of copyright law in Europe.Most academic analysis responds to specific rulings of the Court, often by means of individual case comments focussing on decisions that may be seen as seminal, or affect wider sectoral interests.With the dramatic increase in preliminary references since about 2008, the copyright rulings of the Court also become traceable as an emerging body of jurisprudence, inviting transversal and evolutionary treatment across a range of cases.Competition practices have been analysed by Preck 22 and Stuik 23 .Charter rights perspectives have been developed by Oliver and Stothers 24 .The exclusive right of communication to the public 25 has been extensively discussed.Copyright exceptions are explored by a number of authors, in particular, the reproduction for private copy and the related levies to provide fair compensation 26 , the exception for parody 27 , the exception for libraries 28 .
Some scholars have taken issue with the so-called 'normative' role of the CJEU, stepping in when the law is not sufficiently clear or exhaustive. 29In our previous work 30 we investigated the claims by this literature that the Court is pursuing a harmonising agenda, 31 overstepping its mandate by an excessively teleological interpretation of the law.
These systematic studies of the emerging European Copyright jurisprudence are central to our discourse because they define the doctrinal landscape of copyright which our empirical study seeks to explain.Mireille van Eechoud, Andreas Rahmatian and Eleonora Rosati for example, in separate works, focussed on the originality requirement as defined by the Court in Infopaq and as confirmed in the following jurisprudence. 32Jonathan Griffiths and Martin Husovec analysed the role of fundamental rights in key cases addressing the liability of internet intermediaries. 33The most ambitious early analysis of the CJEU's copyright jurisprudence as a whole is Matthias Leistner's review of 2014. 34istner aims to locate the key issues of law discussed before the court in the full range of copyright cases, which together give shape to a new body of European copyright law.He examines the originality requirement, 35 the reproduction right, 36 the communication right, 37 the distribution right, 38 digital exhaustion, 39 copyright exceptions, 40 fair compensation, 41 as well as the specific concept of communication to the public. 42He recognises both the horizontal range of the autonomous interpretation of copyright concepts and the vertical depth with which these concepts are discussed.The normative role of the Court, in his view, is a necessary complement of a legislative machine that has become more and more cumbersome. 43The activist role of the Courts with respect to copyright law should not be discouraged, but even supported by the use of preliminary references by Member States. 44e significance of the conceptual developments identified by Leistner is confirmed by our empirical analysis which displays a disproportionately high concentration of cases on these topics.Most prominence is given to the concept of 'communication to the public', whose evolution through the jurisprudence of the CJEU provides a telling example of the Court's normative role.

The concept of 'Communication to the public'
The question "who steers the development of law" presupposes that the law is indeterminate or offers sufficient flexibility for jurisprudence to take different directions.For the purposes of this study, it is useful to illustrate this claim with a specific example.
The main EU provision regarding the exclusive right of communication to the public is found in Article 3 of Directive 2001/29. 45Three elements are included in the definition of a communication to the public: (i) it can take place by wire or through wireless means and (ii) on demand services and (iii) communications to the public are included.But all other elements and questions surrounding the concept of communication to the public and how it applies in a modern setting are left undefined.
The international instruments, such as the Berne Convention46 and the WIPO Copyright Treaty 199647 (which Article 3 is implementing) do not add much detail either.That leaves plenty to play for when it comes to the interpretation of the concept of communication to the public.
In a digital on-line environment, there is scope for rebroadcasting of many sorts.Hotels (and spas, clinics, etc.) have for years made centrally received broadcasts available in individual hotel rooms and lobbies, but in the same vein it is now possible to pick up a signal and stream it on-line, in real time or in delayed and edited format.Do these activities involve a new act of communication to the public and is there a new authorisation that needs to be paid for?Or is there, in all or in some of these circumstances, merely one original communication to the public?Broadcasters and hotels, to name just these, have opposite interests on this point.A narrow definition of communication to the public would favour the hotels and with them the national government that wants primarily to encourage the tourism industry.Those in favour of the content providing industry would on the other hand favour a broad interpretation.And whatever the answer was the Court eventually gave, a slight change in (technological) circumstances would mean that parties had an incentive to go back to the Court in the hope of obtaining a slightly different answer that favoured their interests.
Similarly, how do we deal with hypertext links?Does the provision of such a link involve a new communication to the public of the copyright work that is found at the location to which the link directs the user?Parties with very different interests have an incentive to return to the Court to obtain a slightly different answer when new technologies emerge.
So how did the 'Communication to the public' case law of the Court take shape?In the view of the Grand Chamber, who dealt with this topic in Reha Training 48 in the aftermath of various judgments touching upon the topic, there are several complementary criteria, which are not autonomous and which are interdependent.Depending on the circumstances of the case these factors may be present to widely varying degrees.As we will see, these criteria appeared at different stages in the development of the Court's case-law and favour the position of different parties.The Grand Chamber is merely trying to bring them together in an overall framework, pointing to the Phonographic Performance Ireland 49 case, where the Court ruled in favour of a (new) communication to the public in the hotel and therefore in favour of the rightholders by using certain of the criteria, whilst on the very same day favouring other criteria in the SCF 50 case: no communication to the public in the dentist's waiting room and hence favouring the position of the users.
The Grand Chamber then takes as a starting point that there has to be both an act of communication and the communication of the work has to be to a public.This is illustrated with the SBS Belgium 51 case where the scenario of direct injection of broadcasts yielded an act of communication on behalf of the broadcasting stations, but no public.There was also an act of communication (and communication to the public overall) when add-ons that gave access to works without permission were pre-installed on a multimedia player. 52 act of communication refers to any transmission of protected works, irrespective of the technical means or process used.And any transmission or retransmission which uses a specific technical means must separately be authorised.Here the Court favours the interests of the rightholders.ITV v TV Catchup 53 is a clear example of a case where parties returned to the Court to try and include, or exclude depending on the party concerned, on-line streaming of a television broadcast from the concept of an act of communication.
Retransmission by internet streaming was eventually held to amount to a communication to the public in the later case. 54condly, the work must be communicated to the public.From the older cases, such as SGAE 55 , the Grand Chamber takes the basic idea that the concept of a public involves an indeterminate number of potential recipients and a fairly large number of persons, only to qualify that immediately with the teaching of SCF 56 that the work must be made available in any appropriate manner to 'persons in general'.Individuals belonging to a private group can 48 Ca s e C-117/15 Reha Training Gesellschaft für Sport-und Unfallrehabilitation mbH v Gesellschaft für musikalische be excluded, as can be small or insignificant groups.The size of the audience is determined in the light of the potential cumulative effect of there being various audiences, either a large group at the same time or individuals or small groups in succession. 57e Grand Chamber then also refers to the requirement of there being a new public, i.e. a public not taken into account by the rightholders when they authorised the original communication to the public.This is where the line of cases involving the operators of cafes, restaurants, hotels and spas (and now also rehabilitation clinics) comes in.In each of these cases the Court held that there was a new public, but there was no public in the SCF case in dentist Del Corso's waiting room.And neither was there a new public in the leading case in the sub-clusters of internet hypertext link cases, Svensson 58 .The work linked had already been communicated to all internet users and the link, whilst involving an act of communication, could not reach a new public.Other cases then followed to see e.g.whether the same logic applied to framing scenarios. 59That was held to be the case, but on the other hand the posting on one website of a photograph previously posted on another website, without any restriction preventing it from being downloaded and with the consent of the copyright holder, did amount to a communication to the public and was distinguished from the hyperlinking scenario. 60rther criteria highlighted by the Grand Chamber include the profit-making nature of the broadcast or rebroadcast involved and the receptivity of the public.What becomes clear from the Grand Chamber's attempt to propose a comprehensive approach is that in different cases different criteria will play a vital role. 61The emphasis will not only be placed on different points, there is also clear evidence that the Court brought each factor, as its case law developed.These criteria and their weighing will determine whether in a particular case there is a communication to the public or not and whose interests will eventually prevail.That point was again brought home by the Court's second chamber in the GS Media 62 case, when the for-profit argument and the knowledge point were emphasised and it was stated that "whether the fact of posting, on a website, hyperlinks to protected works, which are freely available on another website without the consent of the copyright holder, constitutes a 'communication to the public' within the meaning of that provision, it is to be determined whether those links are provided without the pursuit of financial gain by a person who did not know or could not reasonably have known the illegal nature of the publication of those works on that other website or whether, on the contrary, those links are provided for such a purpose, a situation in which that knowledge must be presumed."This element was also important when the Court held that the operation of a sharing platform that through indexing data gave users access to copyright works amounted to a communication to the public. 63e crucial point for this study is that we aim to determine whether the observations submitted by the parties (including their repeat submissions if cases and follow-on cases are brought deliberately) have influenced the development of the Court's case law.This may provide indications of a deliberate strategy to bring cases to steer or alter the evolution of the case law of the Court.

Recurrent Concepts in Copyright Litigation
Having demonstrated for one central concept of EU copyright law how reasoning may be shaped, favouring one party or another, we now explain our selection of concepts for detailed empirical investigation.We reduced the complete population of 78 copyright cases (registered between 1998 and 2015) to 42 copyright cases that ruled on concepts that had been referred to the CJEU 5 times or more (which we term 'recurrent' concepts).We assume that these repeat references are an indication of a contested policy direction or a normative void, i.e. an opportunity by litigators and Member States to steer the evolving jurisprudence of the Court.This section identifies these cases and analyses which Member States appear to intervene on which underlying concepts.
To this end, we have compiled two datasets: one in which we have coded the main issue discussed in each case, and one in which we have captured suggested answers to all questions referred to the Court for a preliminary ruling in each case.The dataset in which each case is coded according to the main issue discussed allows us a computation of the number of written observations submitted by each country.This becomes a measure of the interest of Member States in specific legal issues.Combined with the dataset coding suggested answers, this will allow an analysis of the correlation of written observations with the interpretation given in the ruling.
Our empirical analysis of copyright litigation starts with the identification of the most common legal concepts submitted to the attention of the Court.Five clusters of legal issues emerge as dominant in the sample.These are: 1) Communication to the public, 2) Copyright exceptions, 3) Levies (fair compensation for private use), 4) Distribution Rights, 5) Liability of Intermediaries.All other legal issues have four or less occurrences within the sample of 78 cases.It is reassuring that the copyright literature reviewed in Section 3.1.above confirms that these are indeed the topics of greatest doctrinal interest.
The following tables offer descriptive statistics identifying the most common legal concepts discussed before the Court.The cases were coded according to all copyright legal issues brought within each case.Cases often discuss multiple concepts.This identifies a total of 51 occurrences of the five 'recurrent' concepts within the 42 cases.

Figure 1 -Recurrent Legal Concepts as the Main Issue in CJEU Copyright Preliminary References Figure 2 -Recurrent Legal Concepts in CJEU Copyright Preliminary References
The of the fair compensation for private reproduction (Art.5.2(b) InfoSoc Directive) follows with 9 cases and 11 total occurrences.
However, if we consider the total number of submissions (i.e.attempts to influence the outcome) in cases in which the above two are the main legal concept discussed, we see that Fair compensation is leading, followed by Communication to the public.It is also interesting to compare the sum of the submissions elicited by these two legal issues (Communication to the public, and Levies), which outruns the sum of the submissions elicited by the other three legal issues.This indicates that parties to the litigation think most is to be gained from steering the Court on these issues.Moreover, it is notable that while litigation on Copyright Exceptions and Distribution rights has occurred more often than preliminary references on the potential liability of Internet Intermediaries (e.g.filtering or monitoring obligations for service providers) the total number of submission for the latter outnumbers the former by a ratio of 30:17.The stakes appear to be higher.
We need to consider that Member States do not have a say on what cases will be sent before the CJEU, as this is decided by national judiciaries. 64However, it is entirely at the Governments' discretion to submit written observations in a given case. 65Therefore we assume that the total number of submissions on a given legal concept can serve as one indicator of the interest of Member States in that issue.This figure however needs to be adjusted to other indicators, as the next section shows.
We now turn to examine more in depth who are the most active governments in submitting legal observations before the Court, and to explore their interest in specific copyright issues.

Repeat Players in Copyright
According to previous research, it is clear that governments use participation in litigation to try to obtain favourable decisions because they understand that they can influence EU jurisprudence through good written observations. 66Moreover, governments have realised 64 However, i t has been theorised that some ca ses a re willingly 'pushed' before the court i n order to foster discussion on certa i n subject-matters.that with a successful written observation before the Court they can obtain the reversal of European acts that they have opposed without success in the EU Council. 67 first sight, the number of written observations in preliminary references points to the Member States most invested in copyright litigation before the Court.We have computed these written observations by country, obtaining a ranking of these countries.France and Italy seem the most invested, followed by the UK and then Spain, Poland, Germany, Austria, Finland, Poland, Czech Republic, Belgium and the Netherlands.However, previous research 68 suggests that in order to identify the most strategic countries we need to control for the number of written observations in their own preliminary references.In fact, some national litigation offices are mandated by the central government to intervene in all preliminary references submitted from their own countries.It is therefore important to understand to what extent Goverments' interventions derive from preliminary references originating from their own country.The following table captures the relationship between the number of written observations responding to a country's own preliminary references and the overall number of this country's observations.Arguably this indicates which countries behave most strategically.The comparison between a) the number of written observations submitted by one country, b) the number of preliminary references originating from that country, and c) the number of submission of this country within preliminary references originating from the same country, produces insights about the propensity to litigation of the various Member States.When for example the number of preliminary references is equal to the number of submissions in their own preliminary references and higher than the submissions in other cases, there is a propensity to litigation which is directly proportional to the number of 'external' submissions.This indicates a likely mandate by this Government to intervene in their own country's preliminary references, but it is also invested in litigation irrespective of this mandate.When however, other variables being equal, the number of submissions responding to preliminary references originating from their own country is lower, this indicates that this Government is conscious of the effects of preliminary references in its own country, and therefore tends to intervene selectively even if not mandated.We argue that this indicates an investment in litigation according to policy interest, pointing to more strategic behaviour.The more a Government intervenes in preliminary reference originating from other countries, the more strategic the litigation behaviour.

Control: Total CJEU litigation for the same period
The analysis above needs to be controlled for a number of factors, such as GDP and overall propensity to litigate before the Court of Justice.Below the number of interventions in all copyright cases (population of 78 cases registered 1998-2015) is displayed next to the GDP of each country.We can observe that the litigation activity of Governments is not always proportional to their country's GDP.For example, Germany and the Netherlands display a low number of interventions despite their relatively high wealth.For other countries however the relation between propensity to litigation and size of economy is more evident, for example for France, Italy and the UK.

Figure 4 -Ranking of Member States by GDP and by number of interventions in Copyright litigation
The overall propensity to litigation on the acquis communautaire of copyright needs to be understood within the context of the overall investment in litigation of each member state before the CJEU.The table below compares the overall number of submissions in concluded preliminary references by EU Member States for the same timeframe as the population of copyright cases (78 cases registered between 1998 and 2015).It can be observed that some countries are active litigators before the CJEU but do not seem to be focussing specifically on copyright law.This is the case, for example, of France, Italy, and Germany.Conversely, countries with a high ratio between overall litigation and copyright litigation seem to invest specifically on this policy issue.These countries include the Czech Republic, Poland and Finland. 69The figure below illustrates the pattern by calculating for each Member State the ratio of copyright interventions as a share of overall litigation before the Court of Justice. 69The l ower the number of interventions, the l ess reliable the figures.See for example Malta.When we consider the country submissions for each of the clusters identified in Section 2, we have another interesting picture.Some countries, despite their overall involvement in EU litigation and their extremely high involvement in copyright litigation, are in fact interested only in a limited number of legal issues.It might be suggested that they intervene only on topics that are relevant for their domestic policy and where they hope to make an impact.Below are the figures on the copyright submissions from each country broken down by main legal concept in the case.The brief analysis above suggests that Governments' interventions in European case law isat least in most cases -strategic; countries tend to invest most in topics relevant for their domestic policies.However, this picture needs to be read by taking into account the routine intervention of some countries in cases where officials are mandated to intervene in preliminary references originating from their own country.

The Performance of Government submissions and their support of Rightholders or Users
After the identification of Repeat Players in copyright and their propensity and specific interest in copyright litigation we move to the core of this study.We now aim to analyse the success rate of the written observations of each country and the proximity of their arguments to the positions of rightholders or users.To this end, we have extracted all copyright-related questions asked within preliminary references and the suggested answers proposed by the submissions of the intervening parties. 70r sample of 42 cases dealing with the five most contested ('recurrent') copyright issues has been coded by each legal question asked in the preliminary reference to the Court and each suggested answer by the parties, the AG and the intervening third parties.147 questions were identified, to which a variable number of governments suggested an answer.Most of these answers were coded with a Yes or No.When the suggested answer could not be reduced to a Yes/No answer, it was coded with an X.Only the answers coded with a Yes or No where considered for statistical computation.
Overall, from the sources discussed in detail in the methodology section (AG Opinions, Court Registry, Agents, Freedom of Information requests) we retrieved a total of 584 suggested answers within the sample of 42 cases and excluding the countries that made only one submission (such as Romania, Norway, Sweden, Latvia).In addition to these, the total number of questions not answered with a Yes or No (and therefore coded with an X) is 136. 71The next figure provides an example of the coding of a case.
Case C-145/10 Eva-Maria Painer v Standard Verlags GmbH was a preliminary reference from Austria (decided by the CJEU in 2012). 72A number of German and Austrian newspapers, magazines and websites had published photographs of a girl that had escaped after being held in captivity for eight years.The photos were taken by Ms Painer (a photographer) while the girl was at nursery school before she was abducted in 1998 (aged 10).She was held captive until she escaped in 2006.This was a huge media story. 70The questions are extracted from the Opinion of the Advocate General (AG) in a way requiring a Yes or No a nswer.We have chosen to elaborate the questions from the Opinion of the AG and not from the Application because the structure of the questions and answers in the ruling mostly follows the pattern of questions as reorganised by the AG.Also, the s uggested a nswers i n the written observations ci ted by the AG, which forms the basis for this part of the analysis, follows the s ame structure. 71There are 30 (=5.8%) missing data points where neither freedom of information requests nor reconstruction from other s ources produced sufficient information to code a country's s uggested a nswer to a question. 72Ca s e C-145/10 Eva-Maria Painer v Standard Verlags GmbH, ECLI:EU:C:2013:138.

Figure 7 -Coding example
As a next step, the answers were coded for 'match with ruling', i.e. if the suggested answer from the country corresponded to the answer given by the Court of Justice in the ruling.This allows the assessments of correlations between the interventions of each third party with the outcome of cases.In addition, the suggested answers and the rulings by the Court were coded for 'match with Rightholder' and 'match with User' (which is widely understood to include Intermediaries).This is useful to assess the position of the intervening third party towards these interests.
In the following table, these matches have been ordered by legal concept under discussion.The table gives the overview of the number of 'matches' with the Ruling, with the Rightholder and with the User/Intermediary.The percentages in the table give the ratio between the 'matches with ruling' and the number of suggested answers.In other words, the ratio also reflects the suggested answers that did not match with the ruling.Therefore, in order to assess the performance of each government the percentages are the key indicators, not the absolute numbers. 7373 Where there are low numbers of interventions or a l ack of data points, percentages need to be treated with caution (see for exa mple the ca se of Hungary).Since the numbers i n the dataset a re too l ow to test for s tatistical significance, we address this issue by constructing a Cartesian diagram.

Analysis
The following analysis focusses on general trends that emerge for the juridification of the European policy process regarding the five 'recurrent' legal issues (which mark a normative void in copyright law).A fuller doctrinal context for each the five legal issues, and of the interests at the heart of individual Government submissions is given in Appendix III.
The issue of Communication to the public is particularly contentious.Consensus among Member States' observations is rare, and it is limited to particularly simple features of this concept (acte clair).For example, it is uncontroversial that installation of TV sets in hotel rooms does not amount to Communication to the public, and that ISP are intermediaries in the light of Article 8 of the InfoSoc Directive.On most nuances surrounding the interpretation of the concept of Communication to the public, the written observations of governments and EU Commission disagree.SGAE, Phonographic Performance and GS Media are among the most notable example of this disagreement.The position of France, among the submitting governments, is to be noted as being often in disagreement with the others.However, this is not as to say that the other countries agree among themselves on every point.The impact of governments' observation on the rulings on this specific topic is also erratic and difficult to pin down.Sometimes the Court follows the opinion of the majority (as for example in SCF); in other occasions the ruling follows the minority (as for example in GS Media).This data indicates that the influence of a Member State does not depend on the number of its allies.The number of submissions from a given governments on a given topic, combined with its success rate (number of matches between its suggested answers and the ruling) gives a more reliable indicator.
In the case of Communication to the public, the figures computed above indicate France as a rather unchallenged protagonist.Other countries, such as Italy, the UK, Ireland, Spain, and Germany, have produced each about half of the number of submissions of France.Often, they take the same stance among each other before the Court, but they tend to succeed mostly when they reach critical mass (e.g. when they coordinate their efforts: see SCF) or when the question is not particularly controversial (and France does not intervene).
The legal issue of Fair compensation (levies) is the one that received the highest attention from member states, with the highest number of submissions in our sample.Finland, Spain and Austria, among these countries, are the most active.Their success before the Court is not consistent.Generally, Spain pushes for broader country self-determination and more flexibility in choosing the features of the levy systems, whereas Finland vouches for a more narrow interpretation of the directive, for example by supporting compensation only for 'personal' copies.Overall the Court favours this latter interpretation.Other countries, like Germany and the UK, rarely intervene on this delicate legal concept, but when they do so their suggested answers are normally accepted by the Court.France keeps an intermediate stance between Spain and Finland: it normally pushes for a strong protection of the rightholder, and it often succeeds in convincing the Court of its arguments.
On Distribution rights, Article 4 of the InfoSoc Directive is broadly interpreted by the Court, which protects the exclusive right of the European owner against infringement coming from non-EU countries.There seem to be a consensus on the point that international exhaustion does not apply and works sold or offered on foreign web sites are infringing, as well as works manufactured in countries with temporary exemption from copyright protection. 74With the only exception of Poland, which seldom manages to convince the Court, the other submitting countries suggest answers that are in line both with the Ruling and the Opinion.
On Exceptions, we can detect a rather balanced approach of the Court (as confirmed by our previous research 75 ) with a number of broad interpretations against an equivalent number of narrow interpretations.The data on the submitting countries is insufficient to determine with a degree of reliability what country is in favour of broad interpretation and what country is against.The available data suggests that the majority of the countries are for a narrow interpretation of copyright exceptions, but their degree of success in front of the court is erratic.
The trends identified can be translated into a Cartesian diagram, in order to give an overview of the relative level of influence of each Member State (and European Commission) and their respective stance in support of rightholders or users.For constructing the graph, the Yes/No coding has been translated into numerical values.Each 'match with ruling' was assigned a value of 1, whereas to each suggested answer not matched in the ruling the value of -1 was assigned.Likewise, to each 'match with the rightholder' the value of 1 was assigned and to each position against the rightholder (for user or intermediary) the value of -1 was assigned.These latter figures differ slightly from the previous table because the cases in which rightholders and users held the same position (very few occurrences) where not computed.The Cartesian diagram give a better idea of the relative position of the countries to each other, and of their performance even in the case of low number of observations.The 'investment' in copyright litigation is conveyed by proxy with the number of written observations for each country, displayed in the graph by the size of the circles. 74See the case of overlapping design and copyri ght protection as i n C-168/09 -Flos ECLI:EU:C:2011:29.See generally Bentl y, "The return of i ndustrial copyri ght?", 34(10) EIPR (2012) 654-672. 75Fa va le, Kretschmer, Torremans, 2016, s upra note 5 a t 58.France stands out for its superior level of influence and for its support for rightholders.Finland is also very effective, but more balanced in its choice of support, as is Germany.Austria and Greece appear to be the least successful countries.Interesting is the effective performance of some 'smaller' players such as the Czech Republic and Portugal.As for the balance between rightholders and users, the evidence here suggests a prevalence of rightholder support, counterbalanced by the strong presence of the EU Commission in the users' playfield.Few countries, among them Portugal, Ireland and the United Kingdom, regularly advance the users' (and intermediaries') perspective.
This analysis is particularly relevant in the current exceptional European context, which might exclude from this picture the participation of the United Kingdom at some point in the future. 77While the above scenario, as it stands, shows an overall -yet delicate -balance, the disappearance of the United Kingdom's voice before the Court is likely to have consequences for the recognition of interests of copyright users and intermediaries.

Conclusion: The making of a transnational rule of law
Our series of studies on the CJEU's copyright jurisprudence was initially motivated by a specific normative void that seemed to be filled by judicial policy making.But when we lifted the lid in our underpinning study 78 revealing the workings of the Court, it became clear that the empirical approach would have much wider theoretical implications.
Never before had a dataset of been produced that allowed to isolate the influence of Member States on the evolution of jurisprudence in a specific subject domain.We were observing the making of transnational law from a new perspective.Our findings suggest a potentially transformative advance in the understanding of the 'production of Europe'.
Our method makes it possible to construct influence rankings, quantifying the jurisprudential relationship between Member States, the European Commission and the Court of Justice in the judification of policy.This is extendable to other areas of law.
In the specific area of copyright law, the study allowed us to target the forces moving jurisprudence.We found that copyright interventions revolve mainly around the concepts of Communication to the public and Copyright levies (private copying), with Exceptions, Distribution rights and Intermediaries also attracting considerable attention.The analysis of recurrent legal issues before the Court of Justice confirms that there are areas of copyright law vulnerable to strategic litigation, arguably because of a normative void.
We found that through written observations, Governments aim to steer the Court towards rulings in line with their respective policies.Member States most invested in copyright law are the dominant litigators of the acquis communautaire in general (France: 15 out of 42 cases in our sample, Italy: 15 cases, the UK: 12 cases), although there are other Member States that invest disproportionately in shaping the evolution of copyright law (Czech Republic: 7, The Netherlands: 6, Portugal: 5, Belgium: 5).
Some governments' interventions are highly skewed, addressing repeatedly very specific copyright issues, which are presumably relevant for their own domestic policies.its investment in litigation across all copyright policy areas.Germany intervenes rarely (only in 6 cases in our sample) but is often successful when it does.
The impact of a written observation on the ruling of the Court is not only determined by critical mass, such as the number of submissions or the size of the country.More important appears to be the quality of the legal reasoning, as the relative success of smaller Repeat Players suggests (for example, Finland and Portugal).In this context, it is interesting to note the relative lack of success of certain governments that specifically invested in copyright litigation but who are seldom rewarded by the Court, such as Austria, Spain and Poland.Poland's written observations are often ignored both by the Advocate General and the Court for unclear reasons.
Finally, it is interesting to read the above findings in the light of the political scenario predicting the loss of the United Kingdom as a strong litigator.Our evidence suggests that the UK carries weight in advancing the perspective of copyright users before the Court of Justice.There are not sufficient data points at this stage to test for statistical significance, but the pattern we identify points to a potential disturbance in the delicate balance between the interests of rightholders and users before the Court.
The findings in the domain of copyright law are striking.The extraordinary closeness of the Commission to the Court of Justice stands out, as does the apparent desire by the Court to maintain legitimacy by moving with Member States' interventions (collectively understood).The Court is not free 'to produce Europe'.
during the proceeding, not after.Moreover, not all FOIAs mention judicial proceedings among the exclusions. 84e EU has its own version of a FOIA in the form of Regulation 1049/02, which allows access to the documents issued by the three main EU institutions: The Parliament, the Council and the Commission.Since the CJEU is formally excluded by this Regulation, it has issued at the end of 2012 a Decision85 that largely borrowed from the norms and principles of the Regulation.However, with this Decision the Court grants access to the documents it holds "in the exercise of its administrative function".This principle would exclude the documents held for the purpose of its judiciary function.
Also these European FOIAs include a number of exceptions 86 to the freedom to access administrative documents.The Regulation for example excludes a number of cases where the disclosure would undermine the protection of commercial interest, court proceedings or legal advice. 87These exceptions however are subject to limits.In particular, they do not apply if there is an overriding public interest to the disclosure. 88Even more interestingly for our study, limits to the access to the Court documents are limited to the time frame during which this prohibition is justified. 89portantly, our application to access Government submissions only relates to cases already concluded; and otherwise our access to these documents would not affect any of the specific interests mentioned above, either by the national or European regulations.In fact, the exclusion of court proceedings in the above-mentioned regulations neither affects those that protect a sensitive interest (public interest, personal or sensitive data, trade secrets), or those whose disclosure would affect the proceeding itself.
The CJEU in its own jurisprudence considered the excludability of court pleadings from disclosure in Sweden v. API, and concluded that judicial deeds can be excluded per se, 84 Ita ly, for example, does not i nclude judicial proceedings per s e among the exceptions l isted in Art. 24, L. n. 241/1990 ("Nuove norme i n materia di procedimento amministrativo e di diritto di a ccesso ai documenti a mministrativi") Gazzetta Uffi ciale del 18 a gosto 1990 n. -commercial interests of a natural or l egal person, i ncluding i ntellectual property, -court proceedings a nd l egal a dvice, -the purpose of i nspections, i nvestigations and audits".
because of the need to preserve the 'serenity of judgements' 90 and the 'equality of arms'. 91his case is interesting because the Court argues the rationale of the EU legislation (from the TFEU 92 to the Regulation) for the exclusion of judiciary deeds ('pleadings') from the freedom of access.The CJEU specifies that its documents are rightly excluded from the Regulation 'while those proceedings remain pending'. 93reover, Government submissions are not documents produced by the parties of the judicial controversy.They are mere legal opinions on points of law raised before the Court by the referring national court.One may wonder whether they could not be considered as the "third party documents" mentioned by Article 9 of the above-mentioned Court Decision.
According to this article, when the Court is required to disclose documents received by a third party (which can include a Member State) it will have to consult said third party before granting access to the applicant. 94ny CJEU commentators and practitioners have called for reform of the Statutes regulating the Court towards greater transparency. 95Some argue that the very case law of the Court in 90 See joined ca ses C-528/07 P a nd C-532/07 P Sweden a nd Others v API a nd Commission [2010] ECR I-8533 para 93, where the CJEU s ta tes 'exposing judicial activities to external pressure, albeit only i n the perception of the public, a nd would di s turb the serenity of the proceedings'.It needs to be s pecified that the case at hand regards a disclosure of Commission's documents under Reg1049/02 not access to Governments written observation 91 Ibi d, para 87 'In addition, such a situation could well upset the vital balance between the parties to a dispute before those Courts -the s tate of balance which is a t the basis of the principle of equality of arms -s ince only the institution concerned by a n a pplication for a ccess to i ts documents, and not all the parties to the proceedings, would be bound by the obligation of di sclosure.'In this ruling however the Court considers the disclosure of submissions from the EU commission; it a rgues tha t the principle would be trumped because only one party (the Commission) would be bound to disclosure, a nd not the others .We however required the disclosure of all submissions, hence respecting the principle of 'equality of a rms'.Contra s t AG Ma dura's Opinion (at 37), a rguing that "once a final judgment has been delivered, the parties' s ubmissions s hould be available to the public unless exceptional reasons demand that secrecy be maintained i n a particular ca se.However, such reasons cannot be presumed to exist in a ll ca ses.Given the weighty reasons militating i n favour of making thi s information public, such exceptions s hould be limited." 92Arti cl e 255 EC. 93 Ibi d, para 94 'It i s therefore appropriate to allow a general presumption that disclosure of the plead ings l odged by one of the institutions in court proceedings would undermine the protection of those proceedings, for the purposes of the s econd indent of Article 4(2) of Regulation No 1049/2001, while those proceedings remain pending.' 94 Art.4.4 Regulation and Art. 9 Decision, quoted «Third party documents 1.The Court of Justice of the European Union s hall not grant access to third party documents in its possession until it has received the consent of the third party concerned.2. For the purposes of the present article, 'third party' s hall mean any natural or l egal person or body external to the Court of Jus tice of the European Union, including the Member States, the other i nstitutions, bodies, offices and agencies of the European Union and non-member States.3. When the Court of Justice of the European Union receives an a pplication for access to a third party document, the competent a uthority shall consult the third party concerned i n order to a s certain whether the third party opposes release of that document, unless i t decides of i ts own motion to refuse to rel ease the document on the basis of one of the exceptions s et out i n Arti cle 3". 95 terms of access to documents of the European Institutions calls for a revision of Regulation 1049/200196.
Among EU institutions, the Council of Europe is actively pursuing a policy of transparency for EU public institutions.For example, the Council issued an international convention that needs yet two ratifications in order to enter in force. 97Moreover, in a Recommendation on access to administrative documents, 98 the Council lists ten possible limitations to this right to access. 99One of these applies to our case: the equality of parties concerning court proceedings.This limitation, according to the guidance of the Council, would "allow a public authority to refuse access to its own documents so as not to weaken its position during proceedings to which it is a party". 100Now, in our sample of preliminary references the Government is never a party.The intervention of a Government in our preliminary references, not featuring said Government as a party, can be assimilated to an amicus curie: a legal opinion from a third party.Accessing these legal opinions, after the ruling has been issue, can hardly weaken the position of a party in that proceeding.
Finally, these legal opinions from Governments, as this paper confirms, are strategically used by Member States to shape the legislation through jurisprudence.They could be compared in nature more to proposals in a legislative process than to pleadings in litigation between parties; and the European legislative process aims to be of the highest standard of transparency. 99Ibi d, a t 27.Possible limitations a re for: i .national s ecurity, defence a nd international relations; ii.public safety; i ii. the prevention, investigation a nd prosecution of cri minal activities; iv.pri vacy a nd other l egitimate private i nterests; v. commercial a nd other economic interests, be they pri vate or public; vi .the equality of parties concerning court proceedings; vii.nature; vi ii.i nspection, control a nd s upervision by public a uthorities; i x. the economic, monetary a nd excha nge ra te policies of the state; x. the confidentiality of deliberations within or between public a uthorities during the i nternal preparation of a matter. 100Ibid, at 15.

APPENDIX II: Written Observations, and the Rules of procedure of the CJEU
In order to understand the function of the intervention of Governments by submitting written observations, and at what point in time they get to the attention of the Court, a few notes on the rules of procedure are helpful.
The procedure of preliminary ruling is prompted by a reference from a national court of a Member State.The national court requires from the Court the interpretation of a point of European Law, which might indirectly involve the infringement of EU law from the Member State. 101The reference is lodged with the Registrar's Office, which deals with the translation of the full text of the reference into the EU official languages.The Registrar also defines the category within which the case should be classified (e.g. the legal subject-matter, as for example 'social security' or 'citizenship' etc.). 102The case is then transferred to the President of the Court and to the First Advocate General.
The President of the Court assigns the case to a chamber, and appoints a Reporting Judge within this chamber, whereas the First Advocate General assigns the case to an Advocate General.According to the procedural rules of the Court, no particular order, as for example a rotation order, is provided for the assignment of a case to a particular chamber.The only rule established by the procedure in terms of case assignment provides that particularly important cases can be entrusted to the Court in its formation as Grand Chamber. 103 The case assignment to the Reporting Judge and to the AG is made again without specific procedural rules (e.g.there is no rotation among Court Members).However, it is frequent that the same judges and the same AGs are responsible for cases relating to similar subjectmatter. 104Usually the AG to whom a case is assigned does not belong to the Member State involved in the litigation, or to the Member State of one of the parties in the case of a preliminary ruling. 105However, this unwritten rule is not always followed. 106nce a case has been assigned, the 'interested parties' have two months, from the notification of the order from reference, to submit their written observations.These 'interested parties' are: a) the litigants before the national court, b) the Member States, c) the EU commission, EU Parliament, the Council, the European Central Bank, d) in some cases, the other EEA countries,107 the EFTA108 supervising authority or a non-member state which is party to an agreement providing its participation within determined subject-matters. 109ritten submissions of such statements of case, defences, and observations by the parties and by any intervener are received by the Registrar and circulated among the panellist judges and the AG.The submissions are accompanied, where possible, by the relevant documentary evidence.
After the parties have filed their submissions, the Reporting Judge writes a preliminary report, to be presented at the Court general meeting, 110 in which a summary of the case is made.The summary includes the main factual situation, the norms potentially applicable, previous referable case law, the questions referred by the national courts and the suggested answers of the parties and of the intervening Governments. 111The Reporting Judge might also indicate points of national law that need further clarifications or even details of the fact situation that would need further measures of inquiry.He or she can suggest to which formation of the Court the case should be assigned and whether to dispense with the hearing or the opinion of the AG. 112ter the conclusion of the written part, a date for the hearing, if a hearing has to be held, is fixed.At the hearing, the pleadings of the parties and any observation by interveners are heard by the panel of judges that forms the chamber.If a supplement of inquiry is necessary, also witnesses and experts are heard by the Court.In the procedure for preliminary ruling the parties do not have the chance to reply in writing to the introductory statements of their adversary, therefore they respond to their argument at the hearing. 113ter the parties have exposed their arguments, at a request of the president of the chamber, the Reporting Judge and the AG can ask for clarifications of the speakers.Finally, the Advocate General can issue his or her Opinion, but more often a date is announced for an opinion to be issued, around a month later. 114In the Opinion, the AG exposes thoroughly any doctrinal and jurisprudential arguments underpinning her proposed solution of the case.Alternative draft opinions can also be submitted by the AG to the Court, supporting different arguments. 115After the reading of the conclusions of the Opinion of the AG, the oral procedure is officially closed.However, the hearing can be reopened, after the submission of the AG, if further measures of inquiry are required. 116fore the deliberation, the Reporting Judge circulates a note to the other judges to suggest a draft ruling for the case.In this note she will state her agreement or disagreement with the opinion of the Advocate general.Any dissenting judge will normally circulate another note to express her views.If there is clear disagreement on a case within the Court, a meeting would be normally set to discuss the matter. 117Dissenting opinions of any member of the panel are not cited in the judgement.Normally, the Opinion of the AG is not cited in detail within the judgement either.If the ruling disagrees with the opinion of the AG, it elaborates its own legal reasoning.In alternative, when the judges agree with the Opinion, they refer to it in full or in part, often without restating the arguments expressed therein. 118e AG, in her Opinion, analyses previous Court case law on the same issue and, if necessary, points out discrepancies and inconsistencies of the Court's previous legal reasoning.A comparative analysis of Member States legislation can also be included in the Opinion in order to support the arguments of the AG. 119Overall, the opinion of the Advocate General provides the most detailed and exhaustive analysis of the case, and it is often much more voluminous than the ruling.Interestingly, some Opinions offer an overview of the written observations submitted by the 'interested parties' before proceeding to analyse the case, some don't.Even when they do, however, they do not necessarily cite all the written observations.Also the Preliminary Report from the Judge Rapporteur, which routinely displays an overview of the submissions before the Court, sometimes overlooks some written observations.It might be inferred that this is a preliminary indication of the relative impact that some submitting governments and EU institutions have on the jurisprudence of the Court. 116Ibid a t 28. 117Ibid. 118Ibid. 119Tri di mas, 2007, supra note 104 a t 1360.
requirements to be considered a communication ex Art. 3 of the InfoSoc Directive.France, The United Kingdom, Spain, Italy, and Czech Republic submitted observations.The opinion of the Advocate General suggested a negative answer, along with Spain and the EU Commission.The other countries did not give a yes/no answer, except for France, which was for an affirmative answer.The Court agreed with the latter.
Bezpečnostní softwarová asociace (BSA),124 another copyright-related case, examined the question of communication to the public, as a secondary issue. 125The question was whether a TV broadcasting of a graphic user interface was communication to the public ex Art. 3 InfoSoc Directive.The answer of the Court was 'No', in agreement with the AG's opinion.Czech Republic, Finland, and the EU Commission had suggested an affirmative answer.In Airfield,126 a Satellite broadcaster was ruled not to need the authorization of the right holder neither for the direct or the indirect retransmission of a broadcasting signal, since this did not reach a public.The norm in question was Article 2 of the Satellite Broadcasting Directive. 127Only Finland submitted observations and suggested the same interpretation of the Court, together with the AG, whereas the EU Commission suggested an affirmative answer.
SCF128 was the following case, interestingly assessing communication to the public both from Article 3 of the InfoSoc directive and Article 8 of the Rental and Lending Rights Directive. 129rance, Greece, Ireland, and Italy submitted observations.The Advocate General suggested that while broadcasting of music in the waiting room was a communication to the public ex Article 8 of the Rental and Lending Right Directive, was not so ex Article 3 of the InfoSoc Directive.The EU Commission, Ireland, and Italy were against a communication to the public in the case at hand, both ex Article 3 and Article 8, whereas France was in favour, on the basis of both directives.The Court eventually agreed with the majority.
In Phonographic Performance, 130 the issue of broadcasting in hotel rooms was again examined, this time in the light of Article 8 of the Rental and Lending Right Directive.A positive answer would have determined that the hotel has to pay equitable remuneration to the rightholder.The referring court also asked whether a hotel could be considered a user, thus benefiting from the exception for private use ex Article 5(2) of the InfoSoc Directive.France, Greece, and Ireland submitted observations.The ruling was in favour of this being a communication to the public, in accordance with France and the EU Commission, and in disagreement with Greece and Ireland.Moreover, according to the AG and France, although the hotel can be considered a user, it cannot be exempted by Article 5(2) of the InfoSoc Directive.Ireland, Greece and the EU Commission disagreed.According to the Commission's observations it is up to Member States to decide whether a hotel can be exempted from paying equitable remuneration.The AG suggested that the hotel is not a user, but if the Court should decide differently, Member States are free to decide whether to exempt them ex Article 5(2) InfoSoc Directive.Eventually the ruling agreed with France on each point.
In Circus Globul Bucaresti, 131 a decision had to be made on whether communication to the public was possible when the public is present.Romania and Spain submitted observations.There was no Opinion of the AG.The Court responded negatively, in disagreement from Spain.In ITV broadcasting 132 there was a question on whether communication to the public ex Article 3 of the InfoSoc Directive included terrestrial TV broadcast, and whether factors such as activity for profit and direct competition with the original broadcaster were relevant.The UK, France, Italy, Poland, and Portugal submitted observations.The ruling responded affirmatively to the question, in agreement with all submitted observations, and negatively to the relevance of the two factors, as suggested by France and Portugal, in disagreement with Italy and the UK.Poland and the EU commission finally were split of the relevance of the two factors.In OSA 133 the issue of communication to the public ex Article 3 of the InfoSoc Directive was transferred from hotels to spa premises, with the same result.The ruling was in favour of a communication to the public, as was the AG, the EU Commission and Czech Republic.Also Germany, Austria, and Hungary had submitted observations, but their position is not known or clear.
Svensson 134 was one of the most hotly debated cases in copyright literature.It discussed hyperlinking as a communication to the public ex Article 3 of the InfoSoc Directive.There was no Opinion; and France, Italy, and the UK submitted observations.France suggested an affirmative answer, whereas the Court eventually ruled in the negative, in agreement with the UK and the EU Commission.To be exact, the Court ruled that hyperlinking was not communication to the public if the linked website was freely accessible, because in this case there was no new public.The decision was controversial because hyperlinking was considered in principle capable of making a communication, although there was no transmission of the content. 135In C More Entertainment, 136 there was a question of extending the rights of broadcasting organizations to sport fixtures live on the internet, and the answer of the Court was 'no', against the opinion of Finland and the EU Commission.
There was no Opinion of the AG.In UPC Telekabel 137 it was asked whether the ISP of the person accessing a work infringing Article 3(2) of the InfoSoc Directive can be regarded as an In SBS Belgium138 the Court ruled that point-to-point broadcasting, without the involvement of the public, was not a communication to the public, in the absence of an Opinion and in contrast to France, Poland and the EU Commission.More recently, in Reha Training, 139 the communication to the public was again examined both from the InfoSoc Directive and the Rental and Lending rights Directive.This time the broadcasting of content had taken place in a rehabilitation centre.The Court was required to clarify whether the interpretation of communication to the public ex Article 3 of the InfoSoc Directive and Article 8 of the Rental and Lending rights Directive had to be made according to the same criteria.The answer was 'yes' to all the questions, in agreement with France, Germany and the EU Commission.
Finally, in GS Media, hyperlinking to unauthorised content was discussed, and it was concluded that this was indeed a communication to the public.More in detail, the Court ruled that hyperlinking to an infringing website is a communication to the public, unless the content has previously been made available with the consent of the right holder.Awareness of the lack of right holder's consent, which is presumed in the case of for profit activity, is relevant.In contrast, the opinion of the AG had suggested a negative answer to the main question, and also that all suggested factors were not relevant, in agreement with Germany, Portugal and the EU Commission.For France, the consent of the right holder was not relevant whereas a key factor is the easiness to find the website.According to the observations of France, there is communication to the public if the link acts as a facilitator.Slovakia generally agreed with the others, except on the knowledge of the lack of authorization, which they deemed relevant.In this ruling, in essence the Court did not seem to accept most of the suggested answers.

Fair compensation (Articles 5(2) InfoSoc Dir.)
The first case on fair compensation stemming from the exception to exclusive rights introduced by Article 5 (2) of the InfoSoc Directive was Padawan. 140The ruling made clear: a) that Member states are not free to choose the system they prefer in the matter of fair compensation; b) that a fair balance of rights has to be respected when calculating fair compensation ex Article 5(2) InfoSoc Directive; c) that the levy must have a connection with the presumed use of the device; d) and that levies cannot be applied to professional use, because in this case no 'private' copy is possible.As a consequence, the Spanish legislation on this matter was declared incompatible with the InfoSoc Directive.Several countries and Collective Management Organizations (CMOs) submitted written observations to this case.
The Opinion aligned with the ruling, on all points.On the freedom to decide by Member States, the observations of the UK, Germany, and Finland were in line with the Opinion, as well as the EU Commission and the CMOs EGEDA and AIE.Spain, France and Greece were instead in favour of self-determination of national governments, as well as the CMOs CEDRO and AGEDI.On the fair balance of rights that has to be taken into account when calculating fair compensation the position of Greece, the UK, EGEDA, and AGEDI aligned with the Opinion; whereas France, Spain, Germany and CEDRO differed.On the matter of the connection between the levy and the presumed use of the device France, the UK, and AGEDI agreed with the Opinion, whereas Spain, Greece, Finland, CEDRO, and EGEDA disagreed.The Opinion also stated that levies cannot be imposed on professionals, because in that case there is no private copy, followed by France, the UK, and Germany.On the opposite position were Spain, Greece, the EU Commission, and all the CMOs.Finally, the legislation of Spain was considered incompatible with the InfoSoc Directive in the Opinion, against the observations of France, Spain, and most CMOs.
The second case on fair compensation in our sample is Stichting de Thuiskopie. 141It was a question of whether distant sellers have to pay levies and the ruling confirmed that this was the case, in agreement with the Opinion and with Spain, Lithuania, Austria, and Finland, and in disagreement with Belgium and the EU Commission.In WG Wort142 the Court ruled that the InfoSoc Directive does not apply before its entry into force; that levies can be paid by an intermediary manufacturer in the case of linked devices; and that the possibility to implement technological protection measures (TPM) on the protected content and the authorization of the right holder are irrelevant for the applicability of the levies.To the prodromal applicability of the Directive the observations and the Opinion were unanimous to say that it had to be excluded, except for Spain.According to the observations of Czech Republic, Austria, and the UK levies can be paid by intermediaries, whereas all the others disagreed.Finally, the possible implementation of TPM and the authorization of the right holder are irrelevant for Germany, Poland and the EU Commission, relevant for most of the others.
The ruling in Amazon.com 143decided that equitable remuneration can be levied only through a CMO.This levy applies also to intermediaries, but national law can a priori allow for exceptions to payment.However, equitable remuneration applies only to recording media sold to natural persons; and it can be paid either in direct or indirect form.Finally, the CMO can require a payment in a Member State also if a levy has already been paid in another Member State.Austria, France, Portugal and Finland submitted observations.The Opinion was followed point by point by the ruling, as was largely the submission of the EU commission.France and Finland broadly agreed, except on the point that the levy can be applied only on a recording media sold to a natural person.The submission of Portugal did not take a yes/no position on the matter.
Copydan Båndkopi144 further elaborated the questions of VG Wort by asking whether the authorization of the right holder and the implementation of TPMs (the actual implementation, not only 'possible' as in VG Wort) could be a consideration when deciding on the applicability of levies.In addition, Copydan also included the harm to right holder, as a further consideration.The referring court also asked about the application of levies to multifunction equipment, such as mobile phones and the collection from producers and importers, if the cost can be passed onto the final user.The Court confirmed that the authorization of the right holder and the implementation of TPMs is irrelevant, but not the harm to the owner.Levies can be applied to mobile phones and can be collected from producers and importers.The Opinion agreed with the ruling on the lack of relevance of TPMs and the authorization of the right holder, but disagreed with the relevance of the harm for the owner and the levies on mobile phones.As for the governments' submissions, France, the Netherlands, the UK, and the EU Commission were for the relevance of the authorization of the right holder while Finland disagreed.On the relevance of TPMs, France, Finland and the EU Commission were in favour, while the Netherlands was against, and the UK di not express a yes/no position.On the possibility of levying also when harm to right holder is minimal, they all agree that this was possible, and also on levies on mobile phones and on producers and importers.
In ACI Adams,145 the important question of whether to implement levies to compensate right owners against unlawful copies was raised, and the answer was negative.A small number of corollary questions were also posed by the referring court.The Court did not answer to whether the three steps test could expand copyright exceptions, and most written observations din not address this point; except for Spain, which suggested a negative answer.The CJEU ruled that levies are not payable to compensate for unlawful private copy and they cannot be paid regardless of the presence or the absence of rightholders' authorization.This position was shared entirely by the Opinion, Spain, Italy, Lithuania, and the EU Commission, whereas Austria and the Netherlands differed.Everybody agreed (except for the plaintiff) that the Enforcement Directive is not applicable to levies.
Other details on the interpretation and definition of the boundaries of fair compensation for private copy were discussed in HP Belgium. 146A few governments submitted observations: Belgium, Czech Republic, Ireland, Austria, Poland, Portugal, and Finland.Here the Court ruled that fair compensation is different for any user and for a private user without commercial ends, in agreement with Ireland, Portugal and the EU Commission, and in disagreement with Belgium, Austria and Poland.The stance of Czech Republic and Finland was not a yes/no answer.The Court ruled also that levies cannot consist of a lump sum and a proportional payment as well, in agreement with Ireland, Portugal and the EU commission, and in disagreement with Belgium, Austria and Poland.Finally, the Court ruled that fair compensation cannot be shared between the author and the publisher;147 and that there cannot be levies for reprography of sheet music, in agreement with Portugal, Finland and the EU Commission.The Opinion agreed with the ruling except for the sharing of fair compensation with publishers.
Finally, in EGEDA (2014), 148 there was a discussion on whether fair compensation for private copy can be financed with the General State Budget, and be determined without taking into account the harm to the owner.Spain, Greece, France, Finland, and Norway submitted observations along with a substantial number of collective management organizations (EGEDA, Dama, AISGE, CEDRO, AGEDI, AIE, and SGAE).The Opinion disagreed with the Ruling and stated that fair compensation can be financed by the general state budget.In line with the Opinion were Spain, Finland, Norway, and the EU Commission.Greece, France, and all the CMOs strongly disagreed.Also the Opinion, France, Finland, and all the CMOs declared that Article 5(2) of the InfoSoc Directive precludes a determination of fair compensation within budgetary limits without taking into account the harm to the right holder.Only Spain and the defendant (Ametic) responded negatively.The ruling did not address this point.

Distribution Rights (Article 4 InfoSoc Directive)
On the Distribution Rights, the first case of Copyright ECJ Jurisprudence was Laserdisken. 149ere, essentially, the Court was asked whether Article 4 of the InfoSoc Directive precluded International exhaustion in EU Member States.This implies that once a copyright work has been lawfully put into commerce in a non-EU country, it can lawfully be imported and sold within the EU without the consent of the right holder.Poland was the only Government to submit observations.On the other hand, not only the EU Commission but also the EU Parliament and the Council of Europe submitted written observations.All the parties and intervening institutions agreed that international exhaustion does not apply in Europe, against the pleadings of the plaintiff and the written observations of Poland.
Peek & Cloppenburg150 was the following case in which the Court was asked to establish whether a work can be considered as distributed ex Article 4 InfoSoc Directive when: a) it is made temporarily available to third parties; b) it is displayed in a shop window; c) and although the work is not copyright-protected in its country of manufacture.Again, only Poland submitted observations, in favour of the application of Article 4; but the Court, in agreement with the AG and the EU Commission, ruled differently.
facilities" was discussed. 157The exception was interpreted in a broad sense, by including among "own facilities" also the facilities of third parties acting on behalf of the broadcaster.This was in line with the AG's opinion and in disarray with the only submitting country, Spain.
In Deckmyn, 158 the concept of the copyright exception for parody was simplified by the court, which required for the definition of this concepts less restraints compared to some national legislation, with the full accord of the AG and the only submitting country, Belgium.
In Meltwater, 159 the national court essentially asked whether on screen website copies met the requirements for Article 5(1) of the InfoSoc Directive, and the answer was affirmative, in agreement with the UK, Poland, and the EU commission.Finally, in Technische Universität Darmstadt 160 the question of dedicated terminals in public libraries, normally protected by the exception for libraries ex Article 5(c) InfoSoc Directive, 161 was discussed.In particular, the possibility for library patrons to print out or save on USB memory supports the library material was questioned; and received by the Court a negative answer.On the possibility of printing, the submitting member states (Germany, Italy, Poland, and Finland) disagreed.Germany and the EU Commission were against, whereas Italy and Finland were in favour.All submissions agreed on the impossibility of saving the documents on USB memory stick.

Responsibility of Intermediaries (Article 8 Enforcement Directive)
Promusicae 162 was the first case discussing the responsibility of intermediaries with relation to online copyright infringement.The CJEU was asked whether the e-commerce directive, the InfoSoc Directive and the Enforcement Directive 163 require an obligation for the provider to communicate personal data of users in order to identify copyright infringements in civil proceedings, and the answer was 'no'.Slovenia and the UK were against, in line with the ruling, with the AG and the Commission, while Italy and Finland were in favour.
First, automatic filtering device were scrutinised by the Court in Scarlet, 164 where the only permissive positions were held by the rightholders and the intervening collecting societies (BEA Video, BEA Music, ISPA), whereas the unanimous opinion of the AG, the EU Commission and the submitting countries (Belgium, Czech Republic, Italy, the Netherlands, Poland, and Finland) was against.Subsequently, in Bonnier, the Court had to rule on whether data protection directives preclude asking information on users ex Article 8 of the Enforcement Directive in order to identify copyright infringers in civil proceedings, and the answer was 'no', in line with the Opinion, the EU Commission and all submitting countries (Sweden, Czech Republic, Italy, and Latvia).
Pre-emptive filtering devices implemented by providers were again considered inadmissible by the Court with the unanimous agreement of all submitting countries and against the sole opinion of the EU Commission in SABAM. 165Interestingly, in UPC Telekabel166 mandating ISPs to prevent their customers from accessing pirate web sites was declared inadmissible by the Court, against the unanimous opinion of the AG, the EU commission, and all the submitting countries (Austria, Italy, the Netherlands, the UK and Ireland).However, very recently in Mc Fadden167 an injunction against internet service providers to monitor users' identity or to restrict access in order to prevent anonymity was considered not incompatible with article 12 of the E-Commerce directive.Poland was the only government to submit written observations and it aligned with the ruling of the Court.
The position of the Court on the Responsibility of the Intermediaries is consistently in favour of the balance of rights (in particular, the fundamental rights to privacy and data protection) even when the rest of the players are against.The intervening Member States instead agree that preventive indiscriminate filtering devices go too far vis-à-vis these fundamental rights, but they are favourable to specific measures to counteract infringement, once an infringement has been detected.

Figure 5 -
Figure 5 -Control: Overall litigation rates compared to copyright litigation rates

Figure 6 -
Figure 6 -Repeat Players on individual Legal Concepts

Figure 9 -
Figure 9 -Cartesian diagram illustrating Member State/Commission influence and promotion of Rightholder or User/Intermediary interest France for example seems to be particularly interested in the concept of Communication to the public.Italy appears particularly interested in Exceptions and Intermediaries, while the UK allocates 77 Arnol d, Bently, Derclaye, Dinwoodie, "The Legal Consequences of Brexit Through the Lens of IP Law", 101(2) Judicature (forthcoming), Oxford Legal Studies Research Pa per No. 15/2017, Uni versity of Ca mbridge Faculty of La w Research Pa per No. 21/2017. 78Fa va le, Kretschmer, Torremans, 2016, s upra note 5.
Gaps were filled by Freedom of Information requests to the European Commission and various Member States.The Court Registry also provided a number of older 'Reports for the Hearing' which are compiled by the Reporting Judge in most cases.Each of these documents may have numerous data points or 'observations', suggesting answers to multiple questions posed to the Court for a preliminary ruling.The final data set therefore contains a total of 584 observations (for an overview, see Figure8below).21 18Brea ch proceedings brought by the Commission a gainst allegedly infringing Member States were excluded.See Favale, Krets chmer, Torremans, supra note 5 a t 38.19See generally on this point Griffiths, "Constitutionalising or harmonising?-the Court of Justice, the ri ght to property a nd European copyri ght law" 38 ELR (2013) 65-78; a nd va n Eechoud, "Al ong the Road to Uniformity -Diverse Readings of the Court of Jus tice Judgments on Copyright Works" 3 (2012) JIPITEC 60. Soulier and Doke v Ministère de la Culture et de la Communication ECLI :EU :C :2016 :878 a t pa ra 36. 59the order in case C-348/13 BestWater International GmbH v Michael Mebes and Stefan Potsch ECLI:EU:C:2014:2315. 60 As i t wa s, amongst other reasons, not required for the sound operation of the Internet.Ca se C-161/17 Land Nordrhein -Westfalen v Dirk Renckhoff ECLI:EU:C:2018:634.61Theneed for an individual assessment was confirmed in ca se C-527/15 Stichting Brein v Wullems ECLI:EU:C:2017:300.62Cas e C-160/15 GS Media BV v Sanoma Media Netherlands BV and others ECLI:EU:C:2016:644.
192, a s modified by L. 15/2005.The l aw i s available online at http://www.altalex.com/documents/codici-altalex/2014/12/09/legge-sul-procedimento-amministrativo.Other countries, such as the Scandinavians, grant full access to all public documents, including Court proceedings, unless a s uperior interest i s endangered.85DECISIONOF THE COURT OF JUSTICE OF THE EUROPEAN UNION of December 2012 concerning public a ccess to documents held by the Court of Justice of the European Union in the exercise of its administrative functions (2013/C 38/02) (hereinafter "Decision").Arti cle 1. Scope 1. "The present decision s hall a pply to a ll documents held by the Court of Justice of the European Union, that is to say, documents drawn up or received by it and in its possession, as part of the exercise of i ts a dministrative functions". 86Deci sion, Exceptions 2. "The Court of Justice of the European Union s hall refuse access to a document where disclosure would undermine the protection of: As a n example of a cademic concerns, see Al berto Al emanno 'Unpacking the Pri nciple of Openness in EU Law: concerns, see generally the intervention of Michael Pa tchett-Joyce, Barrister a t the Hearing on the Reform of the Court of Jus tice of the European Union, 24 Apri l 2013 Brussels, a vailable a t< http://www.europarl.europa.eu/document/activities/cont/201305/20130521ATT66428/20130521ATT66428EN.pdf>(a ccessed 20-8-18).Similar a rguments have been made i n the context of Trilogues under the EU legislative processes.See Curti n a nd Leino, "In s earch of transparency for EU l aw-making: trilogies on the cusp of dawn", 54 Common Market Law Revi ew (2017) 1673-1712.
(as per Article 8 InfoSoc Directive) whose services are being 'used' by the copyright infringer.The Court answer was affirmative, in unanimous agreement with all submitting governments (Austria, Italy, the Netherlands, the UK, and Ireland) and the Opinion.