In particular, the CJEU was asked to balance several fundamental rights in actions against online intermediaries for copyright infringements committed by third parties. In SABAM v. Scarlet (case C-70/10; see also SABAM v. Netlog, case C-360/10), the CJEU defined the outer limits for “injunctions against intermediaries whose services are used by a third party to infringe a copyright or related right” (Art. 8(3) of the 2001/29 Copyright Directive). On 27 March 2014, the Court of Justice in UPC Telekabel (case C- 314/12) went further in delineating the requirements for copyright injunctions against online access providers and the implementation of filtering measures.

In the last years, the ECtHR ruled three times on the proportionality of some interferences with the freedom of expression online resulting from the protection and enforcement of copyright. One case (Fredrik Neij and Peter Sunde Kolmisoppi v. Sweden, 18 Febr. 2013) involves the operators of the infringing website The Pirate Bay. In Ashby Donald v. France (10 Jan. 2013), the question was whether the sanctions imposed by the French courts on the managers of a website containing photographs of fashion shows were passing the test of proportionality for restricting freedom of expression. In a March 11, 2014 decision (Yaman Akdeniz v. Turkey), the ECtHR considers as inadmissible the request made by the user of musical websites to which access has been blocked.

Here are a list of questions for you to consider at the light of (at least) the two decisions UPC Telekabel and Fredrik Neij and Peter Sunde Kolmisoppi v. Sweden:

• What are the fundamental rights that are most often referred to by the Court in Luxembourg and by the Court in Strasbourg?
• Is Article 10 of the ECHR applicable to the Internet in the same way as it is applicable to other means of communication?
• Does the protection of property as a fundamental right play a role in the balancing made by the CJEU and the ECtHR?
• Is it possible to rely on other fundamental rights to support the view that copyright is a fundamental/human right?
• Do you see a difference in the way the balancing is done by the CJEU and the ECtHR?
• What are the implications of a reasoning based on the balancing of fundamental right?
• As a result of those decisions, does copyright move from the sphere of private law to the realm of public law?

You do not necessarily have to respond to all those questions, you can select a few of them. What is important: your text should be concise, well-structured and to the point. Also it should flow well.

You do not have to be too … balanced in your response!

(You might as well take into account this additional question: Do you think the decision Akdeniz is completely in line with the previous case law of the ECtHR?).

Economic issues

The picture of the copyright industry that we outline in the third post of this series has many economic consequences. An important one deals with the economics of mediatisation and its impact on creation or diversity. Another with the role of copyright and related internalisation strategies.

Each copyright industry is now confronted with a change of its ecosystem, which affects both mediatisation and distribution. And which in turn, loops back onto production. This process is uneasy to address in a general way because each industry evolves within its own ecosystem. Moreover, each public figure – artist, author, columnist, sportsman, politician, etc. – behaves like an autonomous brand, generating its own topicality, building its own community on social networks.

intellectuallicensing.com

From a microeconomic standpoint, the utility of a copyrighted works can be termed as the sum of the expression and its associated mediatisation. When the mediatisation is subject to network effects, its utility grows with its coverage and with the consumption of the mediatised good. Copyright industries – from the serialized novels published in the newspapers to the Hollywood star-system (Caves, 2000) – have always been confronted with the necessity to generate network effects around brands and branded products. In the analogue world, such network effects were obtained through well-tested strategies combining a release investment calibrated to reach a critical mass and a media reverberation buzzing on the product. And when the buzz was feared to be negative, then the release investment was made to avoid it. (Cf. the case of the Hollywood “ten-ton turkey” release commented by Richard Caves (2000) p.166.)

An open question concerns the achieving of such network effects in digital ecosystems. According to a recent consulting study,

“total global artist income from recorded music in 2013 was $2.8 billion, down from $3.8 billion in 2000 but up slightly on 2012. Meanwhile artists’ share of total income grew from 14% in 2000 to 17% in 2013 (Mulligan, 2014).

But the top 1% artists grasped 77% of this revenue making digital recorded music a “superstar economy”. This suggests that network effects may be more massive but less accessible in digital ecosystems than in analogue ones. Such a trend would favour on the one hand, big incumbent artists whose long lasting presence grants a critical mass (a lower access cost to media), and, on the other, new artists showing extravagance enough to saturate news and social media (Rihanna, Psy, Lady Gaga, Miley Cyrus…). In other words, the competition for network effects in global digital media seems to induce more “winner-takes-all” behaviours than in the analogue walled ecosystems.

The current paradox of digital ecosystems is that works have become accessible through nearly infinite databases but they are less extensively consumed than in the analogue system. Some more empirical evidence is required to validate such hypothesis, but whether true or not it means that the “long tail” theory (Anderson, 2004) is seriously challenged by the economics of mediatisation. In effect, the diversity of creation – when related to effective consumption – relies more on mediatisation economics than on the productivity of search. A key issue both for economists and media is to design mediatisation models that could enhance the notice of niche products.

Mediatisation economics has long been associated with marketing or advertising whose external effects on sales are difficult to quantify. Simply because the utility of the meaningful complement relies on network effects that are as hazardous to predict as the success of creative works. However, more and more consulting firms are now measuring brands digital presence and their apparent valuation by consumers. Such data, if their collection costs are affordable, might be used to model the network effects associated with mediatisation procedures. (According to our experience, the collection of homogeneous data picked up from various digital media sources requires a dedicated task force with a high fixed cost. The specificity of each mediatisation scheme tends to inhibit such an investment.)

Copyright policy implications

Last, but nor least, the growing importance of mediatisation changes the traditional vision of copyright exploitation. While copyright was for long identified as a way to internalise creation efforts, it is also a way to internalise mediatisation strategies. In a media ecosystem, a copyrighted good can be given for free if its consumption increases its media component. This means first, as noticed by various authors, that piracy or counterfeiting can prove socially useful. Not because it provides access to a lower segment of consumers (Watt, 2000). But because it can raise the utility of the good (the media component) and then, generate a higher willingness-to-pay that is collectable where excludability can apply (as already explained by Paul Belleflamme on this blog).

It also means that the previous consent of copyright holders may raise transaction costs preventing them to benefit from mediatisation opportunities. The “notice and take down” procedure applying on digital platforms corresponds to an ex post consent that allows some right holders to better internalise many spontaneous mediatisation efforts. At the cost, of course, of a careful watch of their product mediatisation. The case of the countless parodies posted on YouTube that can be withdrawn on right holders’ request illustrates this trend. Parodies are pieces of the informational complement that add value to the original copyrighted expressions.

The previous consent rule should then apply but with lenience. A too stringent behaviour would prevent right holders from valuable media externalities. Conversely, take down notices corresponding to an effective internalisation decision of the right holder should involve the responsibility of the noticed intermediary. This is the point where copyright enforcement now meets tort law (Lefort, 2013).

More generally, the dissemination of copyrighted works is anything but natural. Following the B&M premise, the utility of copyrighted expressions relies on a media complement produced through publishing and progressive dissemination. Such complement carries network effects and, thanks to its contextual dimension, allows right holders to gain from economies of scale in publishing other works. This point is clearly underestimated by the economic literature. It explains why the Copyright Term Extension Act (CTEA, 1998) has been mocked by so many conventional economists while supported only by a few copyright experts (Liebowitz and Margolis, 2004): the loss of property rights on the Mickey Mouse cartoons would have, no doubt, spoilt the Disney brand and its ability to mediatise new works. Hence, the economics of mediatisation deserve more research and quantitative analysis. It should also promote a better coordination between trademark and copyright law.

Media ecosystems

The term ecosystem is primarily used in biology do designate a community of living organisms (plants, animals and microbes) in conjunction with the non-living components of their environment (things like air, water and mineral soil), interacting as a system (Wikipedia). These biotic and abiotic components are regarded as linked together through nutrient cycles and energy flows. Willis (1997) traces back the history of the concept and its adaptation to the description of natural systems. Ecosystems are defined by the network of interactions among organisms, and among organisms and their environment. They can be of any size but usually encompass specific, limited spaces.

By analogy, a business ecosystem applies to a community of economic agents interacting for the balance of all, in a framework favourable to externalities (Teece, 2007).

A business ecosystem finds its roots in the idea of value networks and can be seen as a group of companies, which simultaneously create value by combining their skills and assets. Business ecosystems create value for a participant only when the participant is not capable to of commercializing a product or service relying on its own competences. (Clarysse, Wright, Bruneel, Mahajan, 2014, p.1164)

The concept of ecosystem differs from the one of industry, market or sector, because it relies on the prevalence of external effects. Quite frequently, the externalities come from private or informal information exchanges – on individuals, organizations or institutions – that decrease transaction costs and favour various forms of business clustering. Such kinds of ecosystems could be called “transactional” since most of the externalities come from the impact on transaction costs of circulating knowledge, informal relations and institutional information.

Similarly, one can consider that a communicating society runs like a biotope in which public messages are received and re-emitted by media. Media interaction participates in society coordination although media are not systematically linked together by formal or contractual relations. As such ecosystem contributes to the production of a “favourable notice”, the participants rely on it to commercialize their products. Quite often, “the skills and assets” that are shared by media deal with trademarks or talent names equivalent to “brands” that provide a meaningful context to goods or expressions (Bomsel, 2013). The mediatisation of people, firms or goods proceeds from ecosystems dedicated to specific products. One can distinguish the ecosystem of books, the one of the news, of scientific papers, of plastic arts, of music, of cinema, of sports, etc. But there is also a media ecosystem for fashion, luxury goods, cosmetics, automobile, food, drugs, etc. While transactional ecosystems decrease transaction costs between the participants, media ecosystems produce the public image – the informational complement – of business or institutional agents and their related products.

Digital technologies are bringing new media in all ecosystems: online newspapers and magazines, blogs, science portals, audio and video platforms, social networks… The echoing of the news, for instance, can follow many new routes and reach consumers faster and wider than could the pre-digital media. Moreover, the connection between advertising or mediatising a product and its actual purchase by a customer can be internalised through online procedures: an html link will transform a viewer into a purchaser, an audio or video platform teases for free until the viewer decides to pay for a service. As a consequence, the borders between mediatisation and sales tend to blur. This trend affects the economic means used to internalise the externalities of the media ecosystem. This is particularly sensitive in all the copyright industries that produce expressions to be distributed in a dematerialised form.

The mediatisation of copyrighted expressions

Copyrighted expressions are informational goods that have to be mediatised. An inspired poem or a scientific theorem posted on an anonymous blog will remain unnoticed. The release of a novel, of a scientific article, of a music tune, of a movie or a TV series, of a sports event, follows a symbolic route constituting a publishing protocol. The public image of a movie disclosed at the Festival de Cannes will differ from the one of a blockbuster presented in a Hollywood showy opening, and even more from a B movie released straight on DVD. The mediatisation of copyrighted expressions adds meaning and value to the works. It is achieved through a media ecosystem providing a direct and indirect echo to the works.

Copyrighted expressions differ from other goods by the fact that, being non-rival, they can be mediatised through free consumption. Thanks to external effects, the supply of free copyrighted works can be financed by ads paid by other firms, by authors (gold open access reviews), or by other contents bundled on a publishing platform. Digitization, which extends the non-rivalness thanks to the dematerialisation of the supports, increases the potential of mediatisation through free consumption. Hence, the coexistence of free and pay content – news, music, video, scientific articles, sports, etc. – on the same digital platforms, tends to narrow the gap between mediatisation and distribution.

In effect, a media ecosystem reflects a state of the technology. Let’s take the case of music. Under the phonographic period, music was taped on records and mediatised through record sleeves, radios, concerts, magazines, video clips and record shops. The radios had to notice the release of the records and to start the buzz. Competition between record producers gave even rise to the payola system – a bribe to disc jockeys in order to get aired – that was depicted by Ronald Coase (1979) as a hidden publicity. Concert tours were keeping the topicality of the artists. Magazines were feeding fans with stories related to music performers. At the end of the chain, record shops were the ultimate media displaying the sleeves and their assortment in racks. In this ecosystem, most of the relations between copyright holders and media were internalised by formal arrangements. The production and sales of records was shaping the ecosystem. Broadcasters were paying license fees while getting paid through publicity. Showcase and tour operators were hiring performers as to sell tickets and drinks. For the most popular artists, record producers invested in video clips and TV ads. Distributors were rewarded on record sales. Only a few players (magazines, talk-shows, variety-shows) benefitted from informal cross-externalities.

With digitization, the phonographic ecosystem gets supplemented both on the media and the distribution side. On the media side, streaming platforms and social networks can create and amplify a buzz that was formerly produced by showcases and radios. The online press will link a music review to a playlist that is just one-click from the reader. TV promotion is re-focussed on ads and talent shows whose viewers argue on social networks. On the distribution side, music downloads and, increasingly, streaming services substitute to CD sales.

A key pattern of this trend is the convergence of mediatisation and distribution. Streaming platforms deliver both mediatisation – through free access service and discovery or thematic apps – and distribution. On such platforms, a subscriber can share his playlists with his social network’s friends. Thus, a listener can become a disc jockey, now called a curator, getting followers for his selection service. On video platforms, amateurs can post video parodies that will increase the popularity of the original. Right holders can either accept or refuse such parodies and even, for a commercial use, license performers their music publishing rights. In other words, with the continuous surge and rollout of new media, the music ecosystem gets denser and denser making the internalisation strategies more open and flexible than in the past.

What happens with music also happens with books, scientific articles, news, sports, cinema, TV series… Each copyright industry – characterised by the production and the publishing of specific expressions – is confronted with the surge of new media that change its former ecosystem. For a while, this change has been perceived as a dematerialisation arousing competition or cannibalisation between distribution channels. This view was suggested by the vertical design of the industry resulting from the copyright transaction chain and the distribution of material supports.

Actually, the change is far more sophisticated. The newness of the digital ecosystems is that mediatisation gets blurred with distribution. In other words, digitization makes copyrighted works distribution a particular case of mediatisation. This already existed when, for instance, radio stations had to pay license fees to mediatise music. But with digital ecosystems, such practises are widespread and continuously expanding. This in turn, makes internalisation – copyright exploitation – an iterative process adapting to new media, new practises, requiring new contracts and new pricing schemes.

The last post of this series examines economic issues and policy implications related to media ecosystems.

Advertising as a complementary good

Since George Stigler’s paper on information costs (1961), the issue of advertising is widely discussed among economists. For many of them, advertising is a means to reduce consumer’s transaction costs and therefore, to increase demand by decreasing the real price of the product. However, as many advertisements do not provide objective information, some economists see them also as a means to change consumer’s preferences about the product. Therefore a distinction is made between informative and persuasive advertisement: informative ads are perceived as socially useful because they cut transaction costs, while persuasive ads generate artificial, if not lousy, needs (Dixit and Norman, 1978). This approach raises two issues: first, it is nearly impossible to separate informative from persuasive advertising as most messages carry the two aspects; second, from a theoretical point of view, if the consumer has free will, his or her preferences cannot be manipulated by third parties (Stigler and Becker, 1977).

This is why Becker and Murphy (B&M, 1993) uphold that advertising should be seen as a complementary good being positively or negatively valued by the consumer. Advertising is then an informational complement, comparable to a quality surplus or deficit, that helps to better discriminate among consumers: consumers who value more advertising will get more of it attached to the product while others will buy the product according to their valuation of its plain utility. Since the meaning associated with information can have a social value, the information complement may carry network effects increasing the utility of the good with the number of consumers. Such analysis of a meaning value brought by advertising to the product is both in line with the free will consumer hypothesis, and with the change in product utility that publicity can bring.

In their approach, B&M define advertising as giving a “favorable notice” to the good advertised so that an increase in ads raises the marginal utility of this good. They underline that such “notice” can either be paid by the advertiser and given away to the consumer, or sold to him or her through various media channels.

For example sports columns in newspapers provide plenty of notice about local professional teams, even though sports sections are not free to readers, and team owners do not pay for the columns.” (p.946).

Such notice generate additional audience to sporting events, whatever their distribution channel.

Marketers commonly distinguish three kinds of channels: paid media that is used for traditional advertising (TV, radio, Press, etc.); owned media, that is the information delivered by the product itself, by an integrated distributor, by an owned website and all the channels related to it; earned media that is the information published by reviewers or consumers, more and more through social networks (Edelman and Salsberg, 2010). In other words, the information complement associated with the product is not only paid advertising it is the accrued meaning resulting from all media channels. This is why we prefer to call it “mediatisation”. Mediatisation economics consists in getting the most valuable complement at the cheapest cost.

thinkofthat.net

A good illustration is given by the historic launch of the Apple Macintosh by Steve Jobs in January 1984 (Isaacson, 2012, Chapter 15). At that time, Jobs had financed a commercial movie shot by Ridley Scott setting up the myth of the liberating machine. The ad was such a storybook that the machine did not even appear in it. Jobs decided to broadcast it during the Super Bowl, reaching at once 96 million Americans. The ad created a massive expectation enabling Jobs to disclose the product in a theatrical keynote echoed as a global event. In other words, while the paid ad was focussed on persuasion, the informative side was covered by the keynote, which got relayed for free all across the world. This “publishing protocol” became a routine that helped Apple to establish its brand as a powerful context adding value to all the new products marketed by the firm. The innovative genius of Steve Jobs did not only focus on products but also on their mediatisation that catalysed their commercial success.

The multiplicity of media channels and the differences in their pricing schemes underline the importance of external effects in providing the media complement. The quest of such effects does not concern only marketers, but all the talents involved in the product, as well as media professionals and consumers. In the Apple example, the media earned by Jobs’ keynotes are external effects of the initial ad investment. The heuristics of marketing aim at generating the most valuable mediatisation as possible. It is frequently related to the maximisation of the externalities associated with the release investment.

Mediatisation results from a publishing protocol

www.dulwichclassics.com

Schematically, two components contribute to the media complement value: one is the myth, the story conveyed by the media; the other is the reach, the social coverage of the message. The myth, which starts with the design or the packaging, the visible appearance of the product, relies on a creative work. It adds a universe that can be more or less valued by the consumer. (For instance, Alfred P. Sloan (1965, pp. 272-273) explains how the curvaceous design of the 1929 Buick has been mocked as « the poor Buick enceinte » leading to a pitiful commercial flop.) The social coverage can either be obtained trough paid, owned or earned media. It may generate network or snob effects according to the positive or negative valuation of the number of consumers. In some cases, the buzz corresponding to earned media can get hi-jacked by angry consumers willing to spoil the image of the product.

However, these components are not independent. For instance, the disclosure of the product relates both to the story and to the coverage: the product’s entry into the public sphere, its arrival to the world takes part in its mythology. The delivery of the Macintosh coming after the Super Bowl announcement and brilliantly staged by its creator, is fully part of Apple’s myth. The disclosure sets up a before and an after which has the dimension of an event. If spectacular, such event can be covered and echoed freely through the news. A resounding release establishes both the newness and the coverage of a product whatever its objective innovation.

In other words, mediatisation proceeds from a “publishing protocol” which provides the product – whether material or not – its meaningful complement (Bomsel & al., 2013). We use this expression by analogy with a correspondence protocol, which connects together two identified agents who agree to start an information exchange. A publishing protocol connects an identified emitter to indistinct receivers. It announces the disclosure of a message and establishes its public status. By doing so, it creates a context adding meaning to the published expression. Hence, such protocol has the economic dimension of a creative work: nobody knows in advance what its utility will be and how much it will add to the product. Although a mediatisation is needed to give the product its public status, it can be perceived either as a good or as a bad.

In this process, the information complement (the myth) is embedded in a set of names: the emitter of the expression who can be a single artist, a sportsman, a politician, or a cast, a team, provides a context which adds sense to the expression. It is the role of the publishing protocol to make the emitter identified by the public and to provide his name – in fact, his brand – with a contextual value. A publishing brand can be seen as an informational asset providing the branded expression with a context valued by the consumer. Such context has cumulative properties: its meaning value changes at every occurrence. A football team, where each player gets mediatised at every game, is a contextual asset portfolio (Bomsel & al., 2013).

Once the utility of the myth is tested, many techniques can be used to maximise its social coverage and the sales revenue. Because a “publishing protocol” deals with disclosure, it is also a piece of news. Hence, it can benefit from the externalities of the news echo. In effect, the news is not restricted by copyright and fuel media competition: a sensational event is instantly relayed on all the news channels including social networks (Facebook, Twitter, Instagram, Weibo, Weixin, QQ, etc.). Moreover, the brand value of popular artists attracts audiences on every platform that takes up the news. The importance of such externalities in producing the media complement deserves a specific focus. It will refer to the concept of ecosystem, which we explore in the next post of this series.

The objective of this series of posts is to show that in the case of copyright, the value added to proprietary expressions is produced through media ecosystems. The prevalence of external effects associated with the media coverage – the mediatisation of the works – reinforces the internalisation function of copyright. But in a way that differs from the traditional vertical chain of contracts. And that makes copyright exploitation a highly empirical activity.

Copyrighted expressions are usually created to address audiences, e.g. receivers that are unknown from the emitter. They are then brought to the attention of the public through media.

Media provide to the expression its public status. Conversely to business talks, or private correspondence, mediatised expressions are messages formally addressed to the public.

Of course, this is not only true for expressions. It applies to any kind of goods addressing indistinct consumers on competitive markets. Consumers are noticed through media of the availability and characteristics of tradable goods and services.

However, expressions are specific because they consist of information, a non-rival commodity subjected to peculiar property rights. As long as copyrighted expressions were published on material supports, the symbolic dimension of the mediatisation was hidden by the economics of the support. The support was bringing rivalness to the expressions, which were then considered as the informational component of a delivery media (Watt, 2000). In a digital environment, the publishing of copyrighted expressions adds information to information.

This series of posts deals with mediatisation that is the operation carried by media so to give material or immaterial goods a public status. Mediatisation indeed covers signalling or advertising, but also all the kinds of information – design, brands, names, reviews, buzz, word-of mouth, social image – attached to a public or tradable item.

Our thesis is that mediatisation is an information complement attached to the mediatised item. It is a complementary good that alters – increases or decreases – the utility of the plain product. The complement is produced through a set of media packed with information externalities. In the analogue world, such ecosystems used to be rather stable. So were the relations between producers, media and distributors, allowing fixed arrangements to internalise the externalities. In the digital environment, media ecosystems get continuously denser which makes internalisation strategies a moving target.

Mediatisation adds meaning to functional goods or services. In the case of copyrighted expressions, this meaning is added to products whose utility is already meaning. Besides, as we will show, the dematerialisation of media tones down the borders between mediatisation and distribution. Hence, mediatisation economics are even more difficult to assess for copyrighted goods than for material ones.

The rest of this series of posts (to be released in the coming weeks) is structured as follows:

• The second post recalls the economic literature presenting advertising as a complementary good instead of a transaction cost. This approach will be extended to unpaid advertising or mediatisation. This post also examines the mediatisation content and its production through what we call “publishing protocols”. Publishing protocols will be used to describe the information externalities that shape the media ecosystems.
• The third post describes the specificities of the mediatisation of copyrighted goods and the changes brought by digitization.
• The fourth post underlines the importance of mediatisation economics and its impact on creation diversity. It also presents our concluding comments, regarding the internalisation function of copyright in media ecosystems.

Pour contourner l’obstacle juridique de Capital Records, ReDigi a annoncé début 2014 (communiqué de presse disponible ici) qu’elle avait désormais recours à une nouvelle technologie brevetée. Le brevet américain n° 8,627,500 B2 concerne une méthode de gestion et de revente de contenus numériques qui ne nécessite aucune copie sur la plateforme de revente. Cette technologie assure une “transaction atomique” selon les termes utilisés par ReDigi: il s’agit d’un mécanisme de transfert de fichier basé sur le cloud qui transfère automatiquement un bien “original” d’un propriétaire vers un nouveau propriétaire, sans faire de copie intermédiaire. Le système garantit aussi que la copie de départ, ainsi que les copies personnelles qui auraient été faites par le vendeur du fichier d’occasion, ont été retirées de son ordinateur. (Mais l’effectivité de cette technique reste à démontrer, notamment compte tenu des autres copies que le propriétaire originaire a pu faire sur d’autres machines). La technologie de ReDigi se distingue donc de celles dites de “copy and delete” qui font une copie tout en s’assurant que la copie originale est détruite. Ces dernières technologies déjà connues (et brevetées par Amazon ou Apple) ne permettraient pas de contourner l’écueil mis en évidence dans l’arrêt Capital Records v. ReDigi. Cette innovation technologique illustre bien le jeu permanent de cache-cache qui se joue entre technologie et droit.

La question demeure de savoir si la règle de l’épuisement du droit de distribution connue en droit européen est applicable dans l’univers du numérique. L’arrêt Oracle c. UsedSoft du 3 juillet 2012 de la Cour de justice de l’UE (C-128/11) qui porte sur l’épuisement du droit de distribuer des logiciels reste la référence en Europe.

Trois questions à examiner.

Faut-il considérer l’arrêt UsedSoft comme définissant une lex specialis applicable aux seuls logiciels ou, au contraire, considérer que la même règle de l’épuisement est applicable à d’autres fichiers numériques (par ex. des MP3 musicaux ou des ebooks)?

Ensuite, si l’on conclut que la jurisprudence UsedSoft est applicable aux seuls logiciels, cela vaut-il pour des jeux-vidéos ou des ‘apps’ distribués en ligne? La réponse oblige à s’interroger sur la notion de logiciel et à revenir sur un autre arrêt de la Cour de justice de l’UE, l’arrêt dit BSA ou BeSoft (22 déc. 2010, C-393/09).

Enfin, pensez-vous que la nouvelle technologie proposée par ReDigi permet de toute manière de considérer comme licite la revente en ligne de fichiers audio ou texte licitement acquis?

On February 13, 2014, the CJEU handed down its response (here). The decision contains only a few paragraphs. For the Court,

“the provision on a website of clickable links to works freely available on another website does not constitute an ‘act of communication to the public‘ ” (operative part of the decision)

Therefore, the person placing such a link must not ask for a prior authorization as hyperlinking does not fall under the acts covered by copyright.

This is a positive outcome, not only for the functioning of the web, but also for preserving copyright’s relevance on the web. Everyone knows that hyperlinks are the glue with which the Web is made. Tim Berners-Lee, one of the founders of the Internet, saw the potential of using hyperlinks to link any information to any other information (see wikipedia article on hyperlink). Here is another telling presentation :

“Why are you here? It’s not an existential question – it’s a navigational one. And it’s as critical to the future of the Internet as it has been to its past. If you are viewing this article online, chances are that you have reached this destination because you followed a humble hyperlink. And if you’re above the age of, say, four, you know how it works. You click on a hyperlink – and you are magically beamed from one page to another. The hyperlink is the building block of the World Wide Web.”

To limit the possibility to establish hyperlinks could thus be considered as a brake on the engine of free expression that the web definitely supports and expands. Of course, some constraints to the freedom of expression and information, which arguably includes the freedom to place hyperlinks to other online sources as a way to expand the online conversation, are justified. Those limitations must respect the constraints put forward by Article 10(2) of the European Convention on Human Rights. Basically, the limitation by a law (such as copyright law) must be legitimate and proportionate. An across the board obligation to ask for a prior authorization before linking arguably would not pass the proportionality test. Therefore, concluding, as the CJEU did, that inserting clickable links does not require an authorization is right – and positive for the online sharing and discussion. It is also positive for the preservation of copyright online: indeed, would the CJEU have decided differently, the effectiveness and relevance of copyright would have been severely under question. Indeed, the burden required by the application of the basic principle of copyright, that authorization must be sought before committing an act falling under copyright’s scope, would mean that a copyright law encompassing the simple establishment of hyperlinks would not be respected at all in practice. This, in turn, would risk to further undermine the relevance and value of copyright in the online world.

While the outcome of the Svensson case appears clear, it is not easy to understand the whole reasoning of the Court of Justice (We also lack an explanatory background as no Advocate General delivered an opinion in this case). There is  an unclear zone or, one could say, a missing link in the Court’s reasoning. Obviously, the CJEU considers that providing clickable links to protected works is “an act of communication”, but that the second condition for a communication to the public, that the communication is made to a “new public” (as an indeterminate and fairly large number of potential recipients of the communication), is not met. I can only agree with this reasoning. In an article published in 2001 (Liability with Regard to Hyperlinking, 24 Columbia Journal of Law and the Arts 403), Nicolas Ide and I had written with a reference to the international rules (the 2001/29 Directive was not yet adopted) the following:

“Another question is whether the provision of a hyperlink is not equivalent to an act of communication to the public, within the meaning accepted in the WIPO treaties of 20 December 1996, namely ‘the making available to the public of works in such a way that members of the public may access these works from a place and at a time individually chosen by them.’ We do not think this to be the case: as the work is already available to the entire Internet community at the linked site’s web address, we cannot be dealing with a new act of making it available to the public. The link does not extend the work’s audience: surfers who access the work by activating the link can also consult the page directly (as long as they know its URL).”

In Svensson, the issue turns around clickable links posted on the site of a media monitoring agency (Retriever) to the articles of journalists (including Mr Svensson) published on their newspaper’s website. Retriever’s customer who has paid a subscription fee can access a search function on Retriever’s site. In response to a query based on search terms, the customer receives a list of clickable links. If the customer clicks a link, the linked-to site article is displayed in a pop-up window. The customer is not clearly informed that, by clicking on the link, he has been transferred to the website of the press publisher. The process thus can give the impression that the content originates from Retriever, but the risk of confusion is in principle not relevant for the copyright issue under consideration. However the possible confusion could be relevant for an action based on a claim of unfair competition. Links which blur the distinction between the linking and the linked-to sites can confuse the public or appear unfair.

What about framing links as in the CJEU pending case Bestwater (C-348/13; see the preliminary question here). This case brought forward by the German Federal Supreme Court (Bundesgerichtshof) regards the legitimacy of framing video content. YouTube’s option to embed video clips on other websites is at the core of the Bestwater case (see on the kluwercopyrightblog a first presentation by B. Schuetze, here).

Now what about the next paragraph of the Svensson decision?

On the other hand, where a clickable link makes it possible for users of the site on which that link appears to circumvent restrictions put in place by the site on which the protected work appears in order to restrict public access to that work to the latter site’s subscribers only, and the link accordingly constitutes an intervention without which those users would not be able to access the works transmitted, all those users must be deemed to be a new public, which was not taken into account by the copyright holders when they authorised the initial communication, and accordingly the holders’ authorisation is required for such a communication to the public. This is the case, in particular, where the work is no longer available to the public on the site on which it was initially communicated or where it is henceforth available on that site only to a restricted public, while being accessible on another Internet site without the copyright holders’ authorisation.” (para 31)

By adding this paragraph, the Court very probably wants to keep the door open for a copyright claim. Thus, in certain circumstances, posting clickable links might fall under the notion of “communication to the public”. But when exactly? When some access restrictions are placed on the linked-to site and the link allows to by-pass those restrictions, the public that has access through the link can differ from the public that can access the work on the linked-to site. The Court is right to consider that one should be able to oppose this circumvention of the access restrictions. The CJEU probably has in mind a paywall system. Another case pending before the CJEU, C More Entertainment v. Linus Sandberg (C-279/13, here) seems to be about a paywall system (A link to streamed ice-hockey matches is offered by C More Entertainment for visitors who pay for it. The defendant discovered the linked-to page and put the link on its own webpage, thus allowing to bypass the access restriction put in place by C More Entertainment).

While there might be good reasons to prohibit those practices, is it the best option to consider that a communication to the public takes place in those circumstances? Are there not other ways to address such situations?

How do you interpret the Svensson decision and what do you expect the CJEU responses to be in Bestwater and C More Entertainment?

Also, what do you think of the last sentence of the paragraph 31 in the Svensson decision? Could this impact the way search engines function?

(The post is in English, but you are free to respond in French if you so prefer!)

Source: bgr.com

You probably know “Game of Thrones”, the most popular TV series of the moment. The first two seasons were a huge hit and the third season premiered on March 31, 2013 on HBO (the American premium cable network that produces the series). There are about 4.5 million subscribers who follow the series. But overall, the audience is much broader as one should also count several million (the exact figure is difficult to estimate) of “illegitimate” viewers. For example, the first episode of season 3 was downloaded over one million times in the space of 24 hours.

Contrary to what one might think, these record numbers of illegal downloading have not panicked HBO officials. Quite the opposite! HBO programming president (Michael Lombardo) told Entertainment Weekly that he viewed this rather positively:

I probably should not say this, but this is a kind of compliment. The demand is there. And it clearly did not have a negative effect on DVD sales.

One of the directors of the series (David Petrarca) went even one step further by stating that in fact, the series benefits from piracy because it feeds the “cultural buzz” that allows this kind of program to “survive”.

Beyond the specific case of this series, what do economists think of all this? From a theoretical perspective, numerous studies have demonstrated the positive effects that piracy may have on the profits of content producers. I sketch here three mechanisms (for a thorough description, see Belleflamme and Peitz, 2010).

• First, illegal copies of a content good can play a sampling role by attracting consumers and driving them to purchase a legitimate copy later.
• Second, many content goods generate network effects; that is, the attraction of the good increases with the number of consumers of that good, regardless of the consumed copies being legitimate or not.
• Finally, in the same vein, piracy can increase demand for goods that are complementary to the pirated content; the producer is then able to capture indirectly the value that consumers attach to the pirated good.

So here are three mechanisms that could calm the fears of content producers regarding piracy. But naysayers might say that these are nothing but vague theoretical possibilities. To answer this criticism, theories must be submitted to the test of facts. Fortunately enough, the empirical literature on piracy (mainly in the music industry) has grown considerably in recent years (for a recent survey, see Waldfogel, 2012). Numerous studies have converged on the conclusion, now widely accepted, that piracy has had a negative impact on physical music sales (that is to say, CDs). In the jargon of this literature, it is said that illegal downloading has “displaced” physical sales. However, the estimated “displacement rate” is slightly above zero and nowhere near unity, reflecting the observation that the vast majority of goods that were illegally consumed would not have been purchased in the absence of piracy (contrary to what the recording industry would like the general public to believe by counting any download as a lost sale).

Does this conclusion also apply to the online legal sales of digital music (through online stores such as iTunes or via “streaming” services such as Spotify)? This is what Aguiar and Martens (2013) try to establish by analyzing the behavior of digital music consumers on the Internet. The authors use direct observations of the online behavior of more than 16.000 Europeans, whose socio-economic characteristics (age, gender, education, occupation, household size and income, number of children, region of residence) are also known. The data comes from Nielsen, a company that specializes in audience measurement (on television or on the Internet).

What are the results of this analysis? Essentially that illegal downloading has no effect on legal consumption of digital music. At best, this effect is positive: a 10% increase in clicks on illegal download websites leads to an increase of 0.2% in clicks on legal purchase websites. Piracy does not induce any “displacement” of the legal music purchase in digital format; it might even slightly boost sales.

Of course, we must not forget that the households in the sample have willingly accepted to be observed by Nielsen. This introduces two potential biases: on the one hand, it is quite likely that the “heavy downloaders” have refused to be part of the sample; on the other hand, individuals in the sample may have changed their behavior knowing that they were observed. We must also remember that online music sales still account for a small fraction of the overall revenues of the music industry and that physical sales have been shown to suffer from piracy. This being said, the share of online music is increasing (5% in 2010 and 8% in 2011 according to IFPI).

The results of this study can therefore, at least, be interpreted as a sign that the legal purchase of music in digital format is becoming a serious competitor for illegal downloading. Piracy could even become an ally of copyright holders.

So, would the absence of piracy become even worse than piracy itself? I’d be happy to hear your views about that.

(This is a shortened version of a post that was originally published in French on this blog and republished later on Regards Economiques.)

Last Thursday Judge Denny Chin of the US District Court for the Southern District of New York [yep, the same one who rejected the proposed settlement agreement in 2011, holding that it was too unbalanced in favour of Google] issued his much-awaited ruling in the Google Books Library Project saga, which started back in 2005. It was then that the Authors’ Guild and the Association of American Publishers (AAP) sued Google for copyright infringement over non-authorised scanning of quite a few books.

Just to avoid any confusion among readers, Google Books is the broader project that includes the Library Project and the Partner Program (formerly known as Google Print). What is sometimes known as the Google Books case just involves the Library Project.

Background

Since 2004 Google has scanned more than 20 million books in their entirety [with approximately 93% of books being non-fiction, and the great majority of works being out-of-commerce], and delivered digital copies to participating libraries [the New York Public Library, the Library of Congress, and a number of university libraries can download a digital copy of each book scanned from their collections, but not copies from other libraries' collections], created an electronic database of books, and made text available for online searching through the use of snippets [users can search the full text of all the books in the corpus, although it is not possible to view a complete copy of a snippet-view book]. Some libraries have agreed to allow Google to scan only public domain works, but others have also permitted the scanning of in-copyright content. Overall, libraries have agreed to abide by the copyright laws with respect to the copies they make.

The AAP and Google concluded a settlement agreement last year (here and here), but this did not affect the still ongoing litigation between the Authors’ Guild and Google. In particular, the main question left on the table was whether Google could successfully sustain that its Library Project activities were protected as fair use under §107 of the US Copyright Act.

Under US law, the following factors must be considered in order to determine whether the use made of a copyright-protected work may be considered fair:

1. the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes [a key consideration here is whether the use is transformative, ie whether the new work merely supersedes or supplants the original creation or whether, instead, it adds something new, with a further purpose or different character. For recent, yet controversial, cases see here and here];
2. the nature of the copyrighted work;
3. the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
4. the effect of the use upon the potential market for or value of the copyrighted work.

Why Google Library Project is great

Before considering the four fair use factors, Judge Chin highlighted the benefits of the Library Project, including:

• Providing a new and efficient way for readers and researchers to find books (“Google Books has become an essential research tool, as it helps librarians identify and find research sources, it makes the process of interlibrary lending more efficient, and it facilitates finding and checking citations … Google Books has become such an important tool for researchers and librarians that it has been integrated into the educational system — it is taught as part of the information literacy curriculum to students at all levels.”)
• Promoting a type of research referred to as “data mining” or “text mining” [does this ring any bell to European readers?]
• Expanding access to books, eg by providing “print-disabled individuals with the potential to search for books and read them in a format that is compatible with text enlargement software, text-to-speech screen access software, and Braille devices.”
• Helping to preserve books and give them new life, eg in the case of out-of-commerce works.
• Helping authors and publishers (“When a user clicks on a search result and is directed to an “About the Book” page, the page will offer links to sellers of the book and/or libraries listing the book as part of their collections … Google Books will generate new audiences and create new sources of income“) [basically, this means that plaintiffs litigated this case for years against their own interests ...]

Fair use factors

(1) Purpose and character of the use – “Google’s use of the copyrighted works is highly transformative“, the Judge found, in that “Google digitizes books and transforms expressive text into a comprehensive word index that helps readers, scholars, researchers, and others find books.” Furthermore, “[t]he use of book text to facilitate search through the display of snippets is transformative“. In addition, “Google Books is also transformative in the sense that it has transformed book text into data for purposes of substantive research, including data mining and text mining in new areas“. Finally, “Google Books does not supersede or supplant books because it is not a tool to be used to read books.” Although “Google is a for-profit entity and Google Books is largely a commercial enterprise … even assuming Google’s principal motivation is profit, the fact is that Google serves several important educational purposes.”

(2) Nature of copyrighted works – Two considerations favoured a finding of fair use in respect of this factor: (1) most scanned works were non-fiction books, and (2) the books were published works.

(3) Amount and substantiality of portion used – Although Google limited the amount of text displayed in response to a search, the fact that Google scanned full-texts and offered full-text search of books was found to weigh slightly against a finding of fair use.

(4) Effect of use upon potential market or value – Google does not sell its scans and the scans do not replace the books. Above all, “a reasonable factfinder could only find that Google Books enhances the sales of books to the benefit of copyright holders … Google Books provides a way for authors’ works to become noticed, much like traditional in-store book displays.”

Conclusion

Considering all the factors above, the judge concluded that “Google Books provides significant public benefits” and granted Google’s motion for summary judgment.

While it is not said that this is the end to the Google Books saga, last Thursday’s ruling certainly represents an important victory for Google.

Looking at the decision through European lenses, two sudden questions arise:

1. Are orphan works a fake problem? At least under US law it would seem so, as there is no mention of them in the decision of Judge Chin.
2. Are text and data mining activities something which falls outwith the scope of copyright protection tout court? From what Judge Chin wrote, it would seem so: text and data mining would require neither a licence nor a specific exception. Although US open-ended fair use clause differs from InfoSoc Directive’s exhaustive list of exceptions and limitations, Judge Chin did not seem to consider that such activities could infringe exclusive rights of copyright owners. From an EU perspective, if one transferred the interpretation of “commercial” provided in the ruling to this context, it could be argued that most text and data mining activities would be already covered by Article 5(3)(a) of the InfoSoc Directive.

[This blog post was originally published on The IPKat on 17 November 2013]

As  readers will promptly recollect, this case concerned interpretation of Article 5(3) of the Brussels I Regulation in the context of alleged online copyright infringement cases.

The questions referred by the French Court of Cassation read as follows:

1.      Is Article 5(3) of … [the Regulation] to be interpreted as meaning that, in the event of an alleged infringement of copyright committed by means of content placed online on a website,

–      the person who considers that his rights have been infringed has the option of bringing an action to establish liability before the courts of each Member State in the territory of which content placed online is or has been accessible, in order to obtain compensation solely in respect of the damage suffered on the territory of the Member State before which the action is brought,

or

–      does that content also have to be, or to have been, directed at the public located in the territory of that Member State, or must some other clear connecting factor be present?

2.      Is the answer to Question 1 the same if the alleged infringement of copyright results, not from the placing of dematerialised content online, but, as in the present case, from the online sale of a material carrier medium which reproduces that content?

The CJEU replied that Article 5(3) must be interpreted as meaning that, in the context of an action for alleged copyright infringement brought following unauthorised reproduction and online distribution of copies of a work, the courts of a particular Member State have jurisdiction to establish liability and award the damages occurred therein, if the website through which the infringing copies are distributed is accessible from that territory.

As Advocate General (AG) Jääskinen acknowledged in his Opinion (here), Pinckney was the first case concerning the criteria for localising the place where the damage occurs in the event of an alleged online infringement on an unregistered IP right.

In reaching its decision, the CJEU departed significantly from the Opinion of the AG. In doing so, it also marked a discontinuity from recent case law (notably, Joined Cases C‑585/08 and C‑144/09 Peter Pammer, Case C-5/11 Titus Donner (here), and Case C-173/11 Sportradar (here and here) which employed the notion of ‘intention to target’ in the context of online infringements. The stance taken by the Court is not particularly shocking, as departure from the ‘intention to target’ approach may be justified on consideration that (as recalled by the Court itself) “Article 5(3) lays down, as the sole condition, that a harmful event has occurred or may occur.”

Following the decision in Pinckney, the jurisdictional treatment of alleged online copyright infringements is more akin to that of alleged breaches of personality rights, rather than registered IP rights. This conclusion re-states the approach taken following the 1994 decision in Case C-68/93 Fiona Shevill with the additional result that the principle of territoriality, which still governs (is supposed to govern?) copyright, appears significantly diluted, at least in relation to the special rule of jurisdiction laid down in Article 5(3) of the Brussels I Regulation.

The criterion of accessibility eventually adopted by the CJEU was rejected by the AG, on consideration that it “would encourage forum shopping, contrary to the case law of the Court, which has consistently attempted to curb that risk in its interpretation of Regulation No 44/2001.” This fear appears exaggerated, especially considering that a national court will be able to award only the damages occurred in the territory of that particular Member State.

In line with other recent decisions (eg the twin decisions on ISP filtering Case C-70/10 Scarlet and Case C-360/10 Netlog, on which see here, here, here), also in Pinckney the actual response of the Court was overly fact-specific:

While there is little doubt that its meaning ought to be intended broadly, it is questionable whether the last word on jurisdiction in cases of alleged online copyright infringement has yet been said.

Similarly to another reference for a preliminary ruling concerning filtering and currently pending before the CJEU (Case C-314/12 UPC Telekable Wien, on which see here), it is not excluded that new references on jurisdiction in online copyright infringement cases will flourish.

There is no need to think too far ahead in the future, though.Case C-441/13 Pez Edjduk (on which see here) is currently a reference for a preliminary ruling from Austria which seeks clarification on the interpretation of Article 5(3) in the context of a case of alleged online copyright infringement. The question referred by the national court is centred around the (now infamous) notion of ‘intention to target’ … Does that mean that ‘intention to target’ is dead, long live ‘intention to target’?