British Academy of Songwriters, Composers And Authors & Ors, R (On the Application Of) v Secretary of State for Business, Innovation And Skills

Introduction, issues and conclusions 1-23

Parties 24-28

Legislative framework 29-46

The process of implementation in the United Kingdom 47-112

The independent economic analysis commissioned by the

Issue I: The relevant legal principles 126-168

Issue II: The meaning of “harm” 169-207

Issue III: The alleged irrationality and/or inapplicability

Issue IV: The submission that the decision was flawed

Issue V: Whether the Secretary of State predetermined

Issue VI: Does the introduction of Section 28B constitute

Conclusion: Next steps 317-319

Introduction, issues and conclusions

This application for judicial review concerns the decision of the Secretary of State for Business, Innovation and Skills (the “Defendant” or the “Secretary of State”) to introduce a new section (“Section 28B”) into the Copyright, Designs and Patents Act 1988 (“CDPA 1988”) which has the effect of creating an exception to copyright based upon personal private use. Henceforward, any person who legitimately acquires content (music, film, books) can copy that work for his or her own private use without infringing copyright. It can be copied onto other formats or stored in the cloud provided it is for private, non-commercial use. It can now be reproduced without even the theoretical risk of that person being pursued as an infringer of copyright. That purchaser cannot, however, under the exception, copy that content and give it to a family member, friend or colleague and cannot copy it for any commercial use.

This arises because the Government decided to exercise a discretion that it possessed under Directive 2001/29 of the European Parliament and the Council of 22^nd May 2001 on the harmonisation of certain aspects of copyright and related rights in the information society (“the Directive”) which permitted Member States to create exceptions to copyright in certain defined cases. One such case – set out in Article 5(2)(b) - was where a purchaser of content wished to copy it for private use.

The power under the Directive allowed Member States to introduce exceptions of variable scope; for instance in principle it would allow Member States to create an exception to copyright covering a purchaser of (say) a digital download who copied it to a friend or a family member, though, again, provided this was not for commercial use.

In the United Kingdom the Secretary of State decided to implement the Directive by introducing a narrow exception from copyright only for those purchasers of the content who wished to copy for their own private use. The Government decided against any broader exception which would, for example, have permitted copying by family, friends or others within a social circle.

A condition that was attached to the right of Member States to exercise the discretion under the Directive was that if the now permitted use caused more than de minimis harm to the copyright owner then compensation had to be payable.

A report prepared for the European Commission in May 2014 by the economic consultants CRA, “Assessing the economic impact of adopting certain limitations and exceptions to copyright and related rights in the EU – analysis of specific policy options” (“the CRA Report”), pointed out that different solutions had been adopted in different Member States. As of the date of that report (which was a few months before the coming into effect of the new section 28B in the UK) the authors pointed out that no exceptions for private copying existed in the UK and Ireland; and that in Malta, Cyprus and Luxembourg private copying was treated as de minimis. In those Member States where an exception for private copying had been introduced the exception generally stretched beyond personal private copying and covered, for instance, copying to family and friends and by way of gifts, i.e. a wider variety of non-commercial copying. The CRA Report recorded that 21 out of the 28 EU Member States had chosen to introduce private copying exceptions coupled to schemes for compensation funded through levies (ibid, page 87, fn136). In such States government backed or imposed levy schemes had in the main been introduced by way of the imposition of a levy charged on consumers of blank media (CDs, DVDs, Blu-Ray discs etc) and equipment (MP3 players, printers, PCs, etc). The levies were structured differently but in the main, followed two principal routes. First, the imposition of a charge based upon a percentage of the manufacturing or import price implying that the harm sustained by rightholders was a function of the price of the media or device. Secondly, a charge which increased with the memory capacity of the products upon the assumption that the greater the memory or storage capacity the greater the potential for private copying. Revenues collected were then distributed to rightholders via collecting societies albeit that the manner of re-distribution varied greatly.

In the UK pressure to introduce a personal use exception started to increase. In 2010 the Government launched a review (under the auspices of Professor Ian Hargreaves) into intellectual property generally; one recommendation of the resultant report was the introduction of a relatively broad based exception for private own use copying. However, there was also a strong view expressed that levy systems were bureaucratic and not necessarily conducive to efficiency or even innovation. This was of course a controversial view and one that (as many commentators recognised) tended to split along predictable and partisan lines with content creators disagreeing and content acquirers and copying device sellers agreeing.

The Secretary of State decided upon a wide ranging consultation. This process is criticised in these proceedings by the Claimants. Nonetheless, it set out to evaluate a series of options for the introduction of a limited exception to copyright based upon personal use.

It is right to record that the Government had a strong predisposition which it set out clearly in the consultation document. It was that levy schemes were generally a bad thing. Consumers did not wish to pay a levy which was akin to a tax on an activity which they strongly believed was legal or ought to be legal; and, therefore, any new exception to copyright should not be accompanied by a levy or compensation scheme.

However, given the framework for analysis under the Directive, which conditioned any exception to copyright for personal use which caused harm to rightholders upon payment of compensation, the Government’s room for manoeuvre was significantly constrained. The Government therefore placed its cards upon the table and stated that, to avoid having to introduce a compensation scheme, it intended only to create an exception which either caused no (zero) harm to rightholders or de minimis harm and this meant that the exception would be strictly limited to private own, non-commercial, use by legitimate bona fide purchasers of content. In this way the Government sought to reconcile its desire to introduce a personal use exception with its parallel desire to avoid having to introduce a burdensome compensation scheme.

This issue in law thus became crystallised: Would the introduction of a limited private use exception cause no more than minimal or zero harm? The Government’s provisional view was that if a strictly limited exception of this kind was introduced then there would in fact be zero or de minimis harm. This conclusion was however based upon two economic premises which are at the heart of a number of the Grounds of challenge advanced in this litigation.

The first premise was that the only relevant “harm” that would, in principle, need to be compensated for was the risk to rightholders of lost, duplicate sales i.e. the only harm for which compensation would be due were the sales that were lost as a consequence of legitimising what were hitherto unlawful acts. As to this the Secretary of State concluded that whilst consumers considered that they should be entitled to copy content they had acquired they were not, to any material degree, prepared to go out and purchase duplicate copies in order to be able to achieve multiple use. In other words there was no automatic correlation between the desire to copy and lost sales; if the former was constrained the latter would not necessarily occur. And as such if a limited exception to copyright was introduced this would not impact upon duplicate sales.

The second premise was that either fully, or very substantially, sellers of content had already priced-in to the initial sale price the fact that consumers treated content that they purchased as fair game when it came to copying for personal use; in other words when a purchaser acquired content the seller had already factored into the price an assumption that the consumer would in fact copy that work even if it was, in law, unlawful so to do. This has become known as the “pricing-in” principle.

The Government did not simply assume that these economic theses were correct; they put them out to consultation and asked for economic and other evidence proving or disproving them. The Defendant instructed the Intellectual Property Office (“IPO”) (an Executive Agency within the Defendant) to conduct a series of detailed impact assessments. The Defendant accepted that, ultimately, the question whether any material “harm” of the type referred to would flow from the introduction of the new exception to copyright would turn upon “empirical evidence”. The IPO instructed external consultants to conduct empirical analysis into the extent to which post-sale copying was already factored into initial pricing. Research was conducted in the music, film and book sectors with a control conducted in the software sector (it being a control because software is outside the scope of the Directive).

At the end of the process, and based upon reasoning set out in an “Updated Impact Assessment” of March 2014, the decision was taken by the Secretary of State to introduce a new exception for limited private use, but without a concomitant compensation scheme. The explicit justification was that the value of the copying that would occur under the new exception had been and would in the future be built into the initial price. As such there was no “harm” to rightholders for which a compensation scheme would be required.

In this judicial review the Claimants challenge the assumptions adopted by the Defendant as both legally and factually incorrect and flawed, and they also challenge the inferences and conclusions drawn from the evidence actually collected. They also attack the procedure adopted in the consultation submitting that the Secretary of State was so determined to introduce the copyright exception without any compensation scheme that he predetermined the consultation and closed his mind to the evidence and generally acted unfairly in that consultative process. Finally, the Intervener submits that the economic effect of the introduction of section 28B (which is to confer a quantifiable advantage upon the “tech-industry”, and in particular cloud service providers) is contrary to Articles 107 and 108(3) TFEU which prohibit certain types of State aid which have not been notified to the EU Commission. I have grouped the various arguments together under 6 Issues which I set out at paragraph [18] below.

A precursor to all of these disputed issues arises from the Claimants’ submissions that I must adopt either a merits (right/wrong) review or an intense judicial review to decide whether the Secretary of State acted lawfully. The Secretary of State submits, to the contrary, that this is a traditional judicial review and he enjoys the widest possible margin of appreciation and that the Court simply has no jurisdictional competence to adjudicate upon many of the issues arising. This disagreement raises some important points of principle about the approach that I should adopt to the issues, not least because they reflect different types of dispute: some are pure points of law; others are by their nature hard-edged questions of fact; and others have a far greater element of evaluative judgments attached to them. The issue is complicated further in that some are issues of EU law whilst others are matters of domestic law.

The six issues that are between the parties are as follows:-

In relation to the substantive issues (i.e. II-VI) I have decided Issues II, III, V, and VI in favour of the Secretary of State.

However, I have decided Issue IV in favour of the Claimants: see section I below at paragraphs [232] – [273]. This is because the conclusions and inferences which have been drawn from the evidence the Secretary of State has relied upon are simply not warranted or justified by that evidence. This is a conclusion I would arrive at on the basis of any test for judicial review; it is not as such dependent upon the review being intensive or merits based.

Issue IV is material to the legality of the decision of the Secretary of State to adopt section 28B. In my judgment it is sufficient therefore to result in the decision being rendered unlawful. However, this has potentially complex implications for section 28B. It does not necessarily result in that section being struck down. It is, in theory, possible for the Secretary of State to re-investigate the issue in order to address the evidential gap which now prevails. If he does this then one possible outcome would be that the gap that I have identified is plugged and the present decision becomes justified. Another outcome might be that following further investigation the gap in the evidence remains un-plugged in which case the Secretary of State could either repeal section 28B or introduce a compensation scheme. A third possibility is that the Secretary of State simply decides to introduce a compensation scheme without more.

A second issue arising is that although I have come to conclusions on the central legal issues arising there can be no doubt but that the questions of law are of wide significance both in this jurisdiction and in the EU and elsewhere. I therefore should consider the possibility that questions should be referred to the Court of Justice before making any final orders in this case.

The net effect of my conclusions is that in principle the claim for judicial review succeeds and that the next stage is for the parties to make submissions as to next steps.

Parties

The First Claimant is the British Academy of Songwriters, Composers and Authors Limited (“BASCA”), which is a company limited by guarantee representing the interests of music writers of all genres, ranging from emerging talent to established composers. BASCA exists to support and protect the artistic, professional, commercial and copyright interests of music writers. As authors, its members may own copyrights themselves. They may also have assigned those rights in exchange for contractual rights which give them a continuing economic interest in the value of the copyrights.

The Second Claimant is the Musicians’ Union (“MU”), which is an independent trade union. It represents over 30,000 musicians (mostly UK residents) ranging from major artists to self-employed and student musicians. MU assists its members with various issues including the provision of legal advice, understanding their rights and career advice. A large number of its members are either holders of performers’ property rights, or benefit from an income stream resulting from having assigned rights in their performances, or hope to do so in the future. Many of its members are also songwriters.

The Third Claimant is UK Music 2009 Limited (“UK Music”), an umbrella group which represents the collective interests of the UK commercial music industry. It represents all parts of the music industry including songwriters, publishers, musicians, managers, record labels, producers, collecting societies and the live music industry. UK Music lobbies for the rights of its members and promotes their interests through campaigns and events.

The Intervener is the Incorporated Society of Musicians (“ISM”). This is a company limited by guarantee. It was established in 1882 as an association in the United Kingdom representing professional musicians. Its objects include “to promote and support the Art of Music” and “to promote and support the interests of persons working as professionals within the music profession”. Its membership includes in excess of 6,600 individual musicians and 148 organisations who are corporate members of ISM. Its membership has traditionally been drawn from the classical side of the profession though in recent years its membership has expanded and now includes members from a wider range of genres including jazz, musical theatre and rock and pop.

The Defendant is the Secretary of State for Business, Innovation and Skills. He has responsibility for, inter alia, the reform of intellectual property. The Intellectual Property Office (“IPO”) (an operating name of the Patent Office) is an Executive Agency of the Defendant and was responsible, in practical terms, for conducting the consultative exercise which led to the introduction of section 28B CDPA 1988.

Legislative provisions

The origin of the personal private use exception to copyright emanates from Article 5(2)(b) of the Directive. To put that provision into context it is necessary to set out certain other recitals and provisions of the Directive. The Directive was adopted in the context of the provisions of the Treaty providing for the establishment of an internal market and the institution of a system ensuring that competition in that internal market is not distorted. It is recognised that harmonisation of the laws of the Member States on copyright and related rights contributes to the achievement of such objectives.

The Directive purports to harmonise the legal protection of copyright in only specific defined categories of activity. Article 1(2) sets out the areas where the Directive does not apply. These include: the legal protection of computer programs; rental lending and analogous rights; rights applicable to broadcasting of programmes by satellite cable retransmission; the term of protection of copyright; and the legal protection of databases.

Article 2 is entitled “Reproduction right” and lays down the scope of the exclusive right to authorise or prohibit direct or indirect temporary or permanent reproduction:

“Article 2

Reproduction right

Member States shall provide for the exclusive right to authorise or prohibit direct or indirect, temporary or permanent reproduction by any means and in any form, in whole or in part:

(a)

for authors, of their works;

(b)

for performers, of fixations of their performances;

(c)

for phonogram producers, of their phonograms;

(d)

for the producers of the first fixations of films, in respect of the original and copies of their films;

(e)

for broadcasting organisations, of fixations of their broadcasts, whether those broadcasts are transmitted by wire or over the air, including by cable or satellite”.

Article 5 is entitled “Exceptions and limitations”. It confers a power upon Member States, inter alia, to introduce exceptions to copyright. Article 5(2)(b) is the provision at the heart of this judicial review and is in the following form:

“Member States may provide for exceptions or limitations for the reproduction right provided for in Article 2 in the following cases:

…

(b)

in respect of reproductions on any medium made by a natural person for private use and for ends that are neither directly nor indirectly commercial, on condition that the rightholders receive fair compensation which takes account of the application or non-application of technological measures referred to in Article 6 to the work or subject-matter concerned;”.

Article 5(5) stipulates that such exceptions or limitations may not “unreasonably prejudice” the legitimate rights of the rightholder:

“5.

The exceptions and limitations provided for in paragraphs 1, 2, 3 and 4 shall only be applied in certain special cases which do not conflict with a normal exploitation of the work or other subject-matter and do not unreasonably prejudice the legitimate interests of the rightholder”.

There are, in broad terms, two types of exception provided for under Article 5(2). First, there are those where the exception is conditioned upon a right on the part of the copyright holder to receive fair compensation. This covers Article 5(2)(a), (b) and (e), which concern, in addition to private use, photocopying and reproductions of broadcasts made by social institutions pursuing non-commercial purposes such as hospitals or prisons.

However, Article 5(2)(c) and (d) provide for exceptions to copyright for which no right to compensation is explicitly mandated. These refer respectively to reproduction made by publicly accessible libraries, educational establishments, museums and archives which are not for direct or indirect economic or commercial advantage, and ephemeral recordings of works made by broadcasting organisations by means of their own facilities and for their own broadcasts.

In those cases where exceptions are conditioned upon the right on the part of the rightholder to receive fair compensation there is no express explanation or elaboration as to how this is to be achieved. However, some broad principles governing compensation are set out in the recitals to the Directive. It is not in dispute that the Directive must be construed purposively and that the purpose is, in the first instance, to be found in the recitals.

Recitals 31-39 address this issue. I set them out in full below:

“(31)

A fair balance of rights and interests between the different categories of rightholders, as well as between the different categories of rightholders and users of protected subject-matter must be safeguarded. The existing exceptions and limitations to the rights as set out by the Member States have to be reassessed in the light of the new electronic environment. Existing differences in the exceptions and limitations to certain restricted acts have direct negative effects on the functioning of the internal market of copyright and related rights. Such differences could well become more pronounced in view of the further development of trans-border exploitation of works and cross-border activities. In order to ensure the proper functioning of the internal market, such exceptions and limitations should be defined more harmoniously. The degree of their harmonisation should be based on their impact on the smooth functioning of the internal market.

(32)

This Directive provides for an exhaustive enumeration of exceptions and limitations to the reproduction right and the right of communication to the public. Some exceptions or limitations only apply to the reproduction right, where appropriate. This list takes due account of the different legal traditions in Member States, while, at the same time, aiming to ensure a functioning internal market. Member States should arrive at a coherent application of these exceptions and limitations, which will be assessed when reviewing implementing legislation in the future.

(33)

The exclusive right of reproduction should be subject to an exception to allow certain acts of temporary reproduction, which are transient or incidental reproductions, forming an integral and essential part of a technological process and carried out for the sole purpose of enabling either efficient transmission in a network between third parties by an intermediary, or a lawful use of a work or other subject-matter to be made. The acts of reproduction concerned should have no separate economic value on their own. To the extent that they meet these conditions, this exception should include acts which enable browsing as well as acts of caching to take place, including those which enable transmission systems to function efficiently, provided that the intermediary does not modify the information and does not interfere with the lawful use of technology, widely recognised and used by industry, to obtain data on the use of the information. A use should be considered lawful where it is authorised by the rightholder or not restricted by law.

(34)

Member States should be given the option of providing for certain exceptions or limitations for cases such as educational and scientific purposes, for the benefit of public institutions such as libraries and archives, for purposes of news reporting, for quotations, for use by people with disabilities, for public security uses and for uses in administrative and judicial proceedings.

(35)

In certain cases of exceptions or limitations, rightholders should receive fair compensation to compensate them adequately for the use made of their protected works or other subject-matter. When determining the form, detailed arrangements and possible level of such fair compensation, account should be taken of the particular circumstances of each case. When evaluating these circumstances, a valuable criterion would be the possible harm to the rightholders resulting from the act in question. In cases where rightholders have already received payment in some other form, for instance as part of a licence fee, no specific or separate payment may be due. The level of fair compensation should take full account of the degree of use of technological protection measures referred to in this Directive. In certain situations where the prejudice to the rightholder would be minimal, no obligation for payment may arise.

(36)

The Member States may provide for fair compensation for rightholders also when applying the optional provisions on exceptions or limitations which do not require such compensation.

(37)

Existing national schemes on reprography, where they exist, do not create major barriers to the internal market. Member States should be allowed to provide for an exception or limitation in respect of reprography.

(38)

Member States should be allowed to provide for an exception or limitation to the reproduction right for certain types of reproduction of audio, visual and audiovisual material for private use, accompanied by fair compensation. This may include the introduction or continuation of remuneration schemes to compensate for the prejudice to rightholders. Although differences between those remuneration schemes affect the functioning of the internal market, those differences, with respect to analogue private reproduction, should not have a significant impact on the development of the information society. Digital private copying is likely to be more widespread and have a greater economic impact. Due account should therefore be taken of the differences between digital and analogue private copying and a distinction should be made in certain respects between them.

(39)

When applying the exception or limitation on private copying, Member States should take due account of technological and economic developments, in particular with respect to digital private copying and remuneration schemes, when effective technological protection measures are available. Such exceptions or limitations should not inhibit the use of technological measures or their enforcement against circumvention”.

Recital 44 concerns the exercise of the discretion to introduce exceptions and limitations contained within Article 5(2), to the effect that they should be exercised in accordance with obligations imposed under international law. It provides as follows:

“(44)

When applying the exceptions and limitations provided for in this Directive, they should be exercised in accordance with international obligations. Such exceptions and limitations may not be applied in a way which prejudices the legitimate interests of the rightholder or which conflicts with the normal exploitation of his work or other subject-matter. The provision of such exceptions or limitations by Member States should, in particular, duly reflect the increased economic impact that such exceptions or limitations may have in the context of the new electronic environment. Therefore, the scope of certain exceptions or limitations may have to be even more limited when it comes to certain new uses of copyright works and other subject-matter”.

Finally, Recital 58 articulates an important, and well-established, principle of EU law to the effect that Member States should provide for effective enforcement of the rights and obligations set out in the Directive:

“(58)

Member States should provide for effective sanctions and remedies for infringements of rights and obligations as set out in this Directive. They should take all the measures necessary to ensure that those sanctions and remedies are applied. The sanctions thus provided for should be effective, proportionate and dissuasive and should include the possibility of seeking damages and/or injunctive relief and, where appropriate, of applying for seizure of infringing material”.

Pursuant to Article 12 the EU Commission is required to submit to the European Parliament, the Council and the Economic and Social Committee a report on the implementation of the Directive which, inter alia, examines the operation of the exceptions provided for in Article 5 in the light of the development of the digital market. To the extent necessary the Commission is required to submit proposals for amendments to the Directive.

In the first instance, Member States were required to implement the Directive by 22^nd December 2002. However, this did not preclude Member States exercising discretions under the Directive to introduce exceptions to copyright (whether with or without compensation provisions) subsequent to that date. There was, in this regard, no time limit imposed upon the exercise of the powers conferred by Article 5(2).

In 2010 the Government decided to consult over the possible implementation of exceptions in Article 5(2) and it set in train a consultation process leading ultimately to a decision to introduce Regulations creating an exception to copyright for a defined category of reproduction for private use. The Copyright and Rights in Performance (Personal Copies for Private Use) Regulations 2014 (SI 2014/2361) (“the Regulations”) were made on 26^th August 2014 and came into force on 1^st October 2014. The Regulations were adopted pursuant to section 2(2) of the European Communities Act 1972 and introduced, inter alia, a new section 28B CDPA 1988.

The new section 28B provides as follows:

“28B Personal copies for private use

(1)

The making of a copy of a work, other than a computer program, by an individual does not infringe copyright in the work provided that the copy—

(a)

is a copy of—

(i)

the individual’s own copy of the work, or

(ii)

a personal copy of the work made by the individual,

(b)

is made for the individual’s private use, and

(c)

is made for ends which are neither directly nor indirectly commercial.

(2)

In this section “the individual’s own copy” is a copy which—

(a)

has been lawfully acquired by the individual on a permanent basis,

(b)

is not an infringing copy, and

(c)

has not been made under any provision of this Chapter which permits the making of a copy without infringing copyright.

(3)

In this section a “personal copy” means a copy made under this section.

(4)

For the purposes of subsection (2)(a), a copy “lawfully acquired on a permanent basis”—

(a)

includes a copy which has been purchased, obtained by way of a gift, or acquired by means of a download resulting from a purchase or a gift (other than a download of a kind mentioned in paragraph (b)); and

(b)

does not include a copy which has been borrowed, rented, broadcast or streamed, or a copy which has been obtained by means of a download enabling no more than temporary access to the copy.

(5)

In subsection (1)(b) “private use” includes private use facilitated by the making of a copy—

(a)

as a back up copy,

(b)

for the purposes of format-shifting, or

(c)

for the purposes of storage, including in an electronic storage area accessed by means of the internet or similar means which is accessible only by the individual (and the person responsible for the storage area).

(6)

Copyright in a work is infringed if an individual transfers a personal copy of the work to another person (otherwise than on a private and temporary basis), except where the transfer is authorised by the copyright owner.

(7)

If copyright is infringed as set out in subsection (6), a personal copy which has been transferred is for all purposes subsequently treated as an infringing copy.

(8)

Copyright in a work is also infringed if an individual, having made a personal copy of the work, transfers the individual’s own copy of the work to another person (otherwise than on a private and temporary basis) and, after that transfer and without the licence of the copyright owner, retains any personal copy.

(9)

If copyright is infringed as set out in subsection (8), any retained personal copy is for all purposes subsequently treated as an infringing copy.

(10)

To the extent that a term of a contract purports to prevent or restrict the making of a copy which, by virtue of this section, would not infringe copyright, that term is unenforceable”.

Regulation 5 of the Regulations stipulated, in addition, that a copy made prior to the introduction of section 28B, but which would have been lawful under that section, is “…to be treated as a personal copy of the work made under section 28B”. Accordingly, that new copy would not therefore be an infringing copy for the purpose of the related provisions of the CDPA 1988. Nothing in this provision however deprived copyright owners of accrued rights. It served to legitimise future copying but it had no retrospective, cleansing, effect.

There is some disagreement between the parties as to the exact scope and effect of the Regulations and section 28B. There is no need for me to express any conclusions as to this dispute. Nothing in this judgment is therefore to be taken as expressing any definitive view as to the exact scope of section 28B. The new exception provided for in section 28B will have, at the least, the following effects. First, individuals who have acquired recordings in the past which were sold without the benefit of a licence to make personal copies, and copies thereof which were therefore at the time (and remain) infringing copies, will henceforth be entitled to make personal copies of such recordings without restriction or further payment. Secondly, there is now and will in the future be a right to make copies including “for the purposes of storage” in a remote storage facility such as the cloud for which a third party is “responsible”. Thirdly, nothing prevents content owners using technology (insofar as it exists) to control usage, for instance by preventing copying beyond (say) five times.

In the “Explanatory Note” (which of course is not part of the formal Regulations) a summary, in the following terms, of the scope of the new section 28B is provided:

“New section 28B provides that an individual may make personal copies of a copyright work (other than a computer program) which is lawfully owned by that individual, provided the copies are made for that individual’s private use, without infringing copyright in the work. Any personal copies must be destroyed if the individual transfers the work from which they were made to another person, unless the copyright owner authorises the transfer of the personal copies to that person. Any personal copy which is not so destroyed or is transferred to another person without the authorisation of the copyright owner shall be treated as an infringing copy for the purposes of the Act. Sub-section (10) provides that any term of a contract which prevents or restricts the making of a personal copy in accordance with section 28B is unenforceable”.

The process of implementation in the United Kingdom

I turn now to consider the process of implementation as it occurred within the UK. This serves to highlight the basic economic and legal premises upon which the Defendant acted, and the evidence relied upon by the Defendant to justify the ultimate conclusions arrived at which is that the exception would cause “minimal or zero” harm to rightholders.

In the text below I set out, chronologically, the evidence collecting process which led, ultimately, to the adoption of the Regulations introducing the new section 28B CDPA 1988.

In November 2010 the Government commissioned Professor Ian Hargreaves (assisted by a review team) to conduct an independent review of intellectual property and growth in the digital market. The review team was tasked to express a conclusion as to whether the current intellectual property legal framework was sufficiently well designed to promote innovation and growth in the UK economy. Professor Hargreaves concluded that the relevant framework was falling behind that which was required and he was of the view that the UK could not afford to permit an outdated legal framework to impede emerging business sectors. Chapter 5 of the Report (“the Hargreaves Report”) was entitled “Copyright: Exceptions for the Digital Age”. In paragraphs [5.27ff] Professor Hargreaves addressed the question of private copying and format shifting. He observed that EU law permitted Member States to introduce exceptions for private copying provided that fair compensation was paid and he noted that in other EU Member States private copying exceptions were supported by levies on copying equipment. He observed:

“…the schemes vary greatly in terms of the size of levies, what they are charged on, and how the revenues are used”.

He cited in this regard a comparative study of copyright levies in Europe prepared in 2011 by Professor Martin Kretschmer “Private Copying and Fair Compensation: An Empirical Study of Copyright Levies in Europe”, (an ESRC report for ERE/IPO) (“the Kretschmer Report”).

Professor Hargreaves was in favour of the UK implementing the exception for private copying. This was predicated upon two principal conclusions both of which are relied upon by the Secretary of State but, in varying degrees, challenged by the Claimants. First, that private copying was “…part of the normal use of devices”; and secondly, that since rightholders were well aware of consumer behaviour that the benefit of unrestrained private copying was “…already factored into the price that rights-holders are charging”. Paragraphs [5.28] and [5.30], which incorporate these observations, are in the following terms:

“5.28

The UK has a thriving market for personal media devices which rely upon private copying. We see no economic argument for adding an extra charge to these devices in order to authorise reasonable private acts which are part of the normal use of devices. Indeed, without that copying, normal use of those devices would be largely restricted to playing music or films bought online. We are not aware of strong evidence of harm to rights-holders done by this kind of private copying in the normal course of using digital equipment to play works. There is considerable evidence of overall public benefit from consumer use.

…

5.30

The Review favours a limited private copying exception which corresponds to what consumers are already doing. As rights-holders are well aware of consumers’ behaviour in this respect, our view is that the benefit of being able to do this is already factored into the price that rights-holders are charging. A limited private copying exception which corresponds with the expectations of buyers and sellers of copyright content, and is therefore already priced into the purchase, will by definition not entail a loss for rights-holders”.

Professor Hargreaves was influenced by evidence provided to him by innovators that the state of copyright law, as it then was, created a barrier to entry and expansion. He cited by way of illustration the evidence of Mr Martin Brennan who invented the Brennan J7 Music Player which enabled consumers to store music from CDs which they had purchased onto its hard disk making them easily accessible for playing from one point. Professor Hargreaves stated that it was hard to see this product as an undesirable innovation. However, the Advertising Standards Authority had ruled that advertisements for the Brennan J7 should include a warning that using it involved copyright infringement. Mr Brennan made the following submission to the review team:

“My company is possibly one of the best examples of the sort of SME that can help lead this country out of the recession – 10,000% growth in 30 months during the recession – but out of date legislation and red tape may sabotage my growth. It is no exaggeration to say that this matter has caused me more sleepless nights and wasted days than any other in my company’s history…Aside from legal headaches I faced the cost of reassuring customers that record companies will not sue them. It’s daft because US companies Apple and Microsoft had been selling format shifting products in the UK for a decade”.

Professor Hargreaves was also taken with the solution proposed by Nokia who submitted that as a general matter of principle the exception (i.e. that contained within Article 5(2)(b)) should be drawn as broadly as possible to embrace all those acts of format shifting which everyone knew were happening as a matter of course and that most reasonable people believed should be permissible. However, the exception should be drafted on a sufficiently narrow and limited basis to ensure that it caused no significant harm to rights-holders and, in consequence, did not “…give rise to a requirement for payment of compensation”.

The conclusion of Professor Hargreaves was that the Government should introduce an exception to allow individuals to make copies for their own use and that of their immediate family on different media. Importantly, Professor Hargreaves thought that the question of compensation was wrapped up in the rightholders’ freedom to choose an appropriate price:

“Rights-holders will be free to pursue whatever compensation the market will provide by taking account of consumers’ freedom to act in this way and by setting prices accordingly”.

Professor Hargreaves was given only 5 months within which to compile the review and then report both to the Secretary of State and to the Chancellor of the Exchequer. Accordingly, the amount of time during which evidence could be collected was limited. Nonetheless, the conclusion arrived at was that the exception in Article 5(2)(b) of the Directive should be implemented but in a manner which limited the exception to those areas where copying was already taken account of in initial pricing decisions and that, as a concomitant of this, no compensatory mechanism or levy scheme should be introduced. This was an important conclusion and provided an analytical blueprint which was adopted by the Government subsequently and which was ultimately reflected, albeit in modified form, in the new section 28B CDPA 1988.

Professor Hargreaves published, along with his final Report, a “Supporting Document” entitled “Economic Impact of Recommendations”. In chapter 5 he expressed a broad recommendation about copyright exceptions for the digital age:

“Government should firmly resist over regulation of activities which do not prejudice the central objective of copyright, namely the provision of incentives to creators. Government should deliver copyright exceptions at national level to realise all the opportunities within the EU framework, including format shifting”.

In relation to format shifting, Professor Hargreaves recorded that format shifting of music by consumers was widespread and an “acceptable practice”. He recorded the evidence given by the British Phonographic Institute (“BPI”) which had publicly stated that it was in its members’ interests not to pursue, through legal proceedings, format shifters. He also observed that format shifting was explicitly permitted with the digital purchase of music from iTunes, Amazon and other retailers free of digital rights management (“DRM”). He observed:

“In addition 97 per cent of 18-24 year olds format shift legally purchased content and 87 per cent say that being able to do so is important to them; it increases the private value to derived from purchasing music”.

Professor Hargreaves, in a conclusion of direct relevance to this litigation, also expressed his view about three connected economic concepts which had been aired in the economic literature: First, the extent to which consumers would, absent an ability to copy and format shift, purchase duplicate or incremental copies of the content in question; secondly, the extent of pricing-in; and thirdly, the concept of economic value or “harm” to producers. He stated:

“Who is likely to lose and what are the risks?

Varian (developing Liebowitz’ concept of “indirect appropriability”) shows that if sharing takes place – a concern from the producers’ side of format shifting – the producer is likely to sell fewer units of the work. Since the consumer derives greater value from each unit, argues Varian, the producers’ profit might even increase, if pricing is right:

“If willingness to pay for the right to copy exceeds the reduction in sales, the seller will increase profit by allowing that right”.

In a free market, demand and supply will settle on a price for goods where consumer and producer surplus are maximised. When consumers value something highly it increases their welfare, but it does not mean the producer has a right to expropriate that value beyond what the market will bear.

There is no additional direct harm to producers, partly because private format shifting already occurs, but also because format shifting leads to increased consumer welfare, not producer losses. Private format shifting pivots or raises the demand curve and no convincing evidence of demonstrable harm to producers has been submitted to the review.

A risk factor of private format shifting is that it may increase pressure for the introduction of a copyright levy as compensation for detrimental effects on producers. Given that the Review recommends a limited private copying exception corresponding to the expectations of buyers and sellers of copyright content, compensation is already priced into the purchase. By definition, there will be no loss for rightholders. The evidence on copyright levies suggests that they distort the market and affect product introduction as they are both inefficient and negative in terms of distributional and welfare effects”.

These observations are important because they reflect Professor Hargreaves’ acknowledgement and acceptance of some of the key economic assumptions later relied upon by the Secretary of State but which are now criticised by the Claimants.

A further piece of evidence that was before the Hargreaves review team was a literature review prepared for Consumer Focus prepared by Rogers, Tomalin, and Corrigan entitled “The Economic Impact of Consumer Copyright Exceptions: A Literature Review” (November 2009) (“the Rogers Report”). Mr Saini QC drew this to my attention for a number of reasons. First, because (cf. Rogers Report page 15, paragraph 3.1) the authors cite economic literature which distinguished between economic harm (to rightholders) and consumer value. The point made was that the fact that consumers valued the ability to copy content did not, at least necessarily, equate to economic harm to content producers. Consumers might extract incremental value by copying content but absent the ability to copy the consumer would not necessarily go out and purchase a duplicate of the copyrighted product. Hence the subjective value attached to copying by the consumer did not without more translate into lost sales on the part of the producer. This is relevant to the conceptualisation of “harm” for the purpose of Article 5(2)(b) (See Issue II at Section G below).

The second point arising from the Rogers Report is found in section 3.2 (page 17) and concerns “pricing-in” where the following is stated:

“In general, when a consumer buys a CD, or any product, we assume they include in their valuation an estimate of the entire benefits derived from the CD. This includes listening to the CD now, next week and next year, as well as listening to it with friends and family, format-shifting (if legal) and benefits from new technologies in the future.

At any point in time, any specific product will receive a wide range of “valuations”, as consumers are heterogeneous. Economists summarise these evaluations in a demand curve…for a specific product. The market price is determined by a variety of factors, including the competition in the market for the specific product (e.g. the price of similar music or software). The revenue accruing to the agent is shaded and a proportion of this will be forwarded to the creator.

A key result of the standard analysis of demand is that the value of copyright exceptions will be imbedded into a demand curve. Hence, there is an argument that agents and creators automatically extract value from copyright exceptions. Similarly, if the extent of copyright exceptions is changed then the demand curve will automatically shift to reflect these. It will shift outwards if copyright exceptions are increased (as consumers are assumed to gain, hence their valuations increase), and shift inwards if exceptions are reduced.

…

The basic idea that the initial purchase of copyrighted work will embody all the benefits is linked to, but is distinct from, the economic concept of “indirect appropriability” with respect to intellectual property. In economic papers, such as Liebowitz (1985) and Boldrin and Levine (2002, 2004), there is an argument that the “first sale” of a work (e.g. book, song, etc) can embody the entire subsequent flow of revenues from copying and re-selling that work. This is a much stronger argument, which can be used to argue against strong intellectual property rights. In short, the “first sale” argument assumes that the initial buyer has a good ability to on-sell the work (hence the first sale price is high and generates a good return to agents and creators).

In the case of copyright exceptions all that is being assumed is that the consumer understands the benefits they will gain from the purchase. It may be that some of these benefits are uncertain – will their family like the film, or how future technology will expand benefits – but we assume that the consumer is best placed to make these valuations. It may be posited that the prevalence of private copying in the UK, despite its illegality, suggests that consumers assume that paying the first sale price allows them to format shift”.

The authors, thereafter, set out a series of hypothetical examples to demonstrate how extending copyright exceptions could alter the shape of the demand curve and how this might impact upon the initial price and revenues remitted to content producers.

Of importance also is the end conclusion that whilst economic theory posited that pricing-in may exist and be significant – and which is described as a “thought experiment” – its precise extent should be the subject of “empirical study”:

“The above thought experiment points to a number of key issues in assessing the impact of changes in copyright exceptions:

• A change in copyright exceptions will generate consumer benefits in the “secondary market” (e.g. the market for format-shifting, or back-ups). The total benefits generated may be very large relative to any potential economic damage to agents and creators.

• The overall economic damage to agents and creators (e.g. their loss in revenue) depends on how the demand for their work changes. Since increasing copyright exceptions may increase demand (and decreasing copyright exceptions may decrease demand), conceptually it is not clear whether economic damage will be caused at all. However, how demand changes in specific cases should be the subject of empirical study”.

In the conclusions section of the Rogers Report (chapter 6) the authors pointed out the literature review conducted endeavoured to assess the state of current evidence that, inter alia, format shifting caused economic damage to rightholders via lost sales from substitution effects. They commented that the two studies that they had reviewed in depth, however, only indirectly addressed the specific issue and did not “…constitute evidence of economic damage to rights-holders from the proposed consumer copyright exceptions”.

Further, the authors stated that from their review of literature and evidence from those other Member States which had introduced a levy on blank media and digital storage hardware in order to provide “fair compensation”, they were unable to identify any evidence of the economic analysis of damage to rightholders being used in order to set the level of fair compensation.

Their ultimate conclusion was, accordingly, in the following terms:

“As our discussion of economic damage and consumer value makes clear, these are separate concepts. It is not appropriate to claim that increased consumer value derived from the interaction of copyrighted works and new technology is damaging either to artists now, or decreases their incentive to supply creative works. Basic economic concepts suggest that copyright exceptions lead to an increase in consumer value which is significantly greater than the economic damage to rights-holders in the case of format shifting.

Finally, it is clear that there is a gaping void in the current literature on the economic damage caused by consumer copyright exceptions”.

The conclusions which may be drawn from this are: first, that there was no clear empirical evidence to prove the existence of harm to rightholders; secondly, general economic intuition suggested that the exceptions would benefit consumers and cause little or no damage to rightholders because pricing-in occurred; thirdly, that lost (duplicate) sales was an appropriate metric for harm; but fourthly, there was a real need for empirical evidence to substantiate those generalised assumptions.

The Government published a response to Professor Hargreaves in August 2011. In section 6 of that response the Government expressed its concurrence with Professor Hargreaves’ conclusion that copying should be lawful where performed for private purposes and where it did not damage the underlying aims of copyright or restrict copyright holders’ fair remuneration. The Government expressed the view that the introduction of such an exception would cause minimal harm:

“…the Government agrees with the Review’s central thesis that the widest possible exceptions to copyright within the existing EU framework are likely to be beneficial to the UK, subject to three important factors:

• that the amount of harm to rights-holders that would result in “fair compensation” under EU law is minimal, and hence the amount of fair compensation provided would be zero. This avoids market distortion and the need for a copyright levy system, which the Government opposes on the basis that it is likely to have adverse impacts on growth and inconsistent with its wider policy on tax.

• adherence with EU law and international treaties.

• that unnecessary restrictions removed by copyright exceptions are not re-imposed by other means, such as contractual terms, in such a way as to undermine the benefits of the exception”.

Accordingly, the Government indicated an intention to advance proposals for a substantial opening up of the UK copyright exceptions regime upon this basis. However, and consistent with Professor Hargreaves’ observations about the lack of an empirical basis to substantiate the relevant economic assumptions, the Government also stated that it was committed to consulting widely upon the basis of sound evidence.

In December 2011 the Government published a “Consultation on Copyright” (“the Consultation”). Chapter 7 of the Consultation addressed “private copying”. The economic realpolitik which underpinned the Government’s view was that consumers strongly believed that either they were legally entitled to copy content that they had paid for, or, that they ought to be (cf. paragraph [7.24]). Further, that producers (i.e. of content) had already built into their pricing regimes consumers’ expectations about future copying. In paragraph [7.25] the Government stated:

“In recognition of consumer behaviour and belief, UK record labels have not sued the people who copy music they own for their private use. Purchases from many digital download services are accompanied by licences that expressly permit buyers to make private copies. For example, Apple’s iTunes allows copying and backup of purchased content, and Amazon Music Store allows unlimited copying for personal use. Producers are able to protect digital content from copying by applying technological protection measures such as digital rights management (DRM). However, digital music files are increasingly sold DRM-free, reflecting consumer demand for easy private copying”.

In relation to vinyl records, the Consultation recorded (at paragraph [7.26]) that many records were sold with a digital download code permitting purchasers to download digital versions of the music they had acquired enabling them, thereby, to listen on more than one device without the “hassle of copying between vinyl and digital formats”. The Consultation recorded that this practice had been credited as one factor behind the recent increase in vinyl sales. In paragraph [7.27] the following is stated:

“These trends suggest that both consumers and producers of music expect it to be copied between devices, that producers derive value from enabling consumers to format shift content, and that the ability to do so is priced into the purchase price of music”.

The Government did though recognise that the practice of building consumers’ expectation as to future copying into the initial price, evident in the field of music downloads and vinyl, might not necessarily be true of other media: See paragraph [7.28].

In the Consultation the Government squarely posed both the correctness of its conclusion that for many media subsequent copying was an assumption built into the initial price; but also that this might not apply to all media universally. Further, the Government specifically raised as an issue for consultation the provision of private cloud services which permitted people to store digital files remotely accessing them over the internet instead of storing them locally upon a PC or other device. The Consultation stated that, unlike in other jurisdictions, UK copyright law meant that providers of similar services in the UK confronted legal risks and copyright owners could seek licence fees from them. The provisional conclusion arrived at by the Government in relation to cloud services was set out in paragraph [7.37]:

“This means the providers of cloud services and consumers who use them face costs in the UK that they do not in other countries. These costs, including the costs of negotiating and paying for licences, are likely to prevent new businesses, particularly innovative start-ups, entering the UK market for cloud-based services. Moreover, if such services are licensed, consumers may have to pay copyright owners twice for the same right to copy content – once at the point of sale, and again via a subscription to private cloud storage. Certain cloud and music services are already licensed in the UK, although the extent to which these licences relate to the narrow act of private copying under consideration, as opposed to value-added services, is currently unclear. It seems reasonable that people should be able to copy content they have bought, to a private cloud to the same extent that they are able to copy it to a personal device. But we would not want to undermine the ability of copyright owners to charge for value-added services based on the provision of content. We would welcome evidence on the impact of existing copyright law on the development and provision of cloud services, and the potential impact of a private copying exception on future developments in this area”.

In summary the Government concluded that there would be benefits to both consumers and innovative businesses if certain, limited, acts of private copying were permitted and were “…fully factored in to the price of content at the point of sale”.

It was in this context that the Government set out its preferred course, i.e. its predisposition:

“We will ensure that such an exception is sufficiently narrow so that any harm caused to copyright owners through private copying is minimised, and therefore do not intend to introduce a levy on electronic devices or blank media, as exists in some other EU States…”.

The Government indicated that it did not propose to introduce a wide private copying exception which would permit copying of any work from any source including works lent or shared by others for private use. Such an exception would permit people legally to copy any work that they had access to and it would be easy to obtain copies of works without paying for them. As such, the Government concluded that revenues to copyright holders would fall in consequence and the incentive effects of copyright would be undermined. The Government’s guiding principle was that each of the options under consideration included specific measures “…intended to minimise potential harm to copyright owners” (cf. paragraph [7.41]).

As to the position in other Member States the Government did not view these as true comparables because where private copying exceptions had been introduced these permitted copying not only for personal use but also for use within an extended private or domestic sphere, such as a family, household or social circle. The rationale behind the decision to reject this was based upon the uncertainty of the ability of the rightholders to price-in the extent of subsequent copying. In paragraphs [7.44] and [7.45] the Government explained:

“7.44

Some economists liken copying within a private circle to shared uses of traditional media, such as private sharing of books or family television viewing. Someone buying a television for their family home will be prepared to pay more for it as it will benefit their whole family. In a similar way, it may be possible to factor the value of family copying into the purchase price of content, compensating for potential harm to copyright owners. To the extent that this behaviour already occurs, it may already be factored in to prices to some degree.

7.45

However, comparisons with other shared uses of media are imperfect. Sharing a work by copying it results in multiple independent copies that can be used in parallel, and may replace sales. It is unclear whether harm caused by sharing, however limited, can be fully compensated through the price mechanism. The extent to which consumers consider copying for family to be reasonable behaviour is also unclear. In view of these uncertainties, an exception, that permits copying of content within the private sphere, is not currently the Government’s preferred option. However, the Intellectual Property Office intends to commission research into pricing of creative content to further explore this issue”.

The Government was also influenced by the Kretschmer Report which expressed strongly negative sentiment about the EU system of levies. For example, in section 3 of his Summary Report, Kretschmer stated as follows:

“As a mechanism for “fair compensation”, 22 out of 27 European Union members have chosen to meet the requirement through a levy system. The exceptions are the UK and Ireland (only time-shifting of broadcasts is permitted), Malta, Cyprus and Luxembourg (private copying treated as de minimis). Within the 22 countries that provide for a compensated private copying exception, levy schemes vary widely in the following respects:

- levies apply to different media or equipment that can be used to make copies (e.g. recordable carriers, hard disks, MP3 players, printers, PCs);

- levies differ in tariffs for the same media or equipment, and apply different methods of calculation (e.g. memory capacity, percentage of price);

- levies differ in whether they are imposed on the manufacturers, importers or distributors of media or equipment, or consumers;

- levies differ in beneficiaries (music, audio-visual, reprographic rightholders; wider cultural or social purposes);

- regulatory structures differ (processes for setting tariffs and distribution, contestability of tariffs, governance and supervision of agencies).

The system as a whole is deeply irrational, with levies for the same devices sold in different EU countries varying arbitrarily. The following three figures illustrate the variable scope and density of levy schemes, and track the evolution of total revenues raised from copyright levies in the EU. The underlying data can be found in Study I”.

In the light of all these considerations the Government posed a series of questions:

To accompany the Consultation the IPO conducted an initial impact assessment which considered the various options for implementing a private copying exception. Option 0 (i.e. zero) was “do nothing” such that private copying remained illegal; Option 1 (the strongly preferred option) concerned the introduction of a narrow private copying exception permitting copying content legally owned by an individual to another medium or device owned by that individual for their own personal use; Option 2 was to introduce the private copying exception but permitting copying by an individual’s family or domestic circle; and Option 3 would cover content which could be copied and which did not need to be legitimately owned or purchased by the individual performing the copying. This assessment was published at the time of the Consultation so consultees were able to see the more detailed reasoning underpinning the “predisposition”.

In respect of Option 1, which is the option which most closely reflects section 28B CDPA 1988, the assessment noted that initial evidence suggested that the costs to copyright owners would be minimal or zero given that they would be able to charge for private copying through purchase prices. This “initial assumption” was based upon an economic analysis set out in an “Evidence Base” which was summarised in pages 6-10 of the assessment. On page 10 the following was stated:

“Private copying is already widespread, and infringement is not pursued, suggesting any costs to copyright owners and benefits to consumers are already priced in to the purchase price of content. We expect any additional costs not already priced in to be minimal or zero, and capable of being compensated through the purchase price. As such there will be no overall economic cost to copyright owners. In addition to our upcoming consultation on the issue, the IPO is commissioning research to investigate the relationship between the prices charged for content and the permissions to copy which are included”.

This conclusion (that all additional copying was, and could be, compensated through the purchase price) was said to be supported by economic literature in particular: the Rogers Report; the Kretschmer Report; and, and an article by Hal Varian “Copying and Copyright” (2005) in the Journal of Economic Perspectives.

But the assessment acknowledged that the conclusions expressed about pricing-in were merely “assumptions” and “theoretical arguments” and needed a proper evidential underpinning. To this end independent research would be commissioned:

“The issue of harm from a private copying exception needs to be addressed and the consultation document is looking for evidence on this issue. Currently, theoretical arguments by Hal Varian (developing Liebowitz’s concept of “indirect appropriability”)…shows that if sharing takes place, a producer is likely to sell fewer units of the work. In the above examples the potential harm to copyright owners is generally assumed to be through lost sales caused by sharing. The less sharing is possible, the less there is potential for harm.

However, since a consumer derives greater value from content they are able to do more with, a producer could charge more for it. Varian says “…if willingness-to-pay for the right to copy exceeds the reduction in sales, the seller will increase profit by allowing that right”. In a free market, demand and supply will settle on a price for goods where consumer and producer surplus are maximised. It is possible that both consumers and producers will benefit from permitting private copying – consumers from being able to do more with content and producers from being able to charge more.

There is also evidence that the value of copying is already built into purchase prices. For online goods such copying is usually licensed and so included in the price where the retailer (e.g. Amazon) and copyright owner licence the buyer to use the digital file in certain ways including how they can copy and store the file. For example iTunes UK limits the number of devices an electronic music file can be copied to while Amazon UK allows an unlimited amount of copies to be made for personal use. The lack of difference between the prices of content on different MP3 download sites with different terms of use suggests that the ability to copy content does not impact on purchase price to any significant degree. However we will be commissioning economic research in order to test these assumptions during the consultation period”.

The Consultation commenced on 14^th December 2011 and ran until 21^st March 2012. During this period the IPO hosted a series of events in London and throughout the UK to encourage interested parties to express their views. The Consultation stimulated 471 responses from a range of organisations large and small. Mr Robin Stout, a Deputy Director of Copyright Policy at the IPO, produced Witness Statement evidence in these proceedings in which he explained that during the course of policy development in excess of 250 meetings on copyright exceptions occurred which, inevitably, included those representing the interests of copyright holders. He observed that the position adopted by the Third Claimant in these proceedings (UK Music) was not wholly representative of the views expressed by its membership. During meetings with HM Treasury some members of UK Music felt that they could accept a narrowly defined private copying exception without levy upon the basis that this largely reflected current reality. He observed that the responses received demonstrated distinct divergences of views but he accepted that a particular topic of concern was the actual extent to which the proposed exception would have an impact upon the incomes of rightholders.

In relation to remote “cloud” storage Mr Stout stated as follows:

“Some, such as UK Music, also opposed the inclusion of remote “cloud” storage within any exception. I should note that it emerged from some of the meetings held with stakeholders that the music industry’s concern regarding “cloud” services was not the loss of opportunity to licence pure storage of private copies in the cloud (to so-called “dumb lockers”). As noted…below, we were given no evidence that such storage was being licensed and were told that within broader licences such storage was usually “zero-charged”. Rather, the concern of the industry seemed to be to extract licence revenue from non-storage “value added” cloud services such as iTunes Match, and to use the ambiguity surrounding the liability of dumb lockers to persuade cloud providers to adopt such value added licensing agreements”.

Between September and November 2011 the Business, Innovation & Skills Select Committee conducted an Inquiry into the Hargreaves Report and the response of the Government thereto. 117 pieces of written evidence were received and the Committee convened a number of evidence collecting sessions. The Select Committee Report (published 27^th June 2012) endorsed the approach posited in the Consultation in relation to format shifting and concluded that a copyright exception premised upon personal use or use within the private sphere was likely to be the most practicable and justifiable solution. The Committee endorsed, albeit in tentative terms, “pricing in” as “probably a fair point”:

“It is probably a fair point that pricing of content now includes an expectation of personal use extending beyond the original medium, for very practical reasons. Robert Ashcroft of PRS for Music pointed out one such reason, namely that without format shifting it is becoming impossible to play certain content at least through certain (portable) media:

“You can barely go into an electronics shop and buy a portable CD player these days. You cannot make it illegal for people to go and buy a CD and then get it on to the only device that they have that plays music. Where you draw the line is the question””.

In December 2012 the Government published its response to the Consultation in a document entitled “Modernising Copyright: A modern robust and flexible framework”. Annex A addressed the question of private copying. It summarised the points made in the Consultation. Submissions from parties other than copyright owners were almost universally in favour of the exception; the great majority of rightholders however felt that a private copying exception of any type or scope would generate unreasonable harm and that for it to be justified some form of levy, licence or tax upon copying devices would be necessary to compensate them, as was the case in some other Member States of the EU.

The Government concluded, however, in the light of the evidence obtained, that the value to consumers of additional functionality could be factored in at point of sale, i.e. that this was essentially a matter for the pricing discretion of the producer or vendor; and further that this additional functionality “often is” factored in at the point of sale “to some extent”. Under the heading “Minimising harm to rights-holders” (cf. page 23) the Government stated:

“In its Consultation paper, the Government proposed an exception that causes minimal or no harm to copyright owners. The Government considers that its preferred option (an exception which permits copying only by a lawful owner of an original copy) will achieve this aim. The exception will be limited to use by people who already own a lawful copy of a work. It means that the ability to make personal copies can be factored in at the point of sale.

Market evidence suggests that the value to consumers of additional functionality associated with content, including the ability to make private copies, can be, and often is, factored in at the point of sale. When music downloads that could be copied freely for personal non-commercial use were first introduced they were sold at higher prices than the DRM-protected equivalents they were sold alongside, reflecting consumer willingness to pay more for copies that they could use more easily on different types of device. Similar price differences can be observed in relation to other types of content where DRM is still used. Some consumers say they would not buy CDs at all if they were physically unable to format shift them to their personal media devices.

The Government’s own analysis agrees that, to some extent, the value that consumers gain from being able to make personal copies will be factored into the market via positive effects on prices and demand.

The Copyright Directive says that rights-holders must be compensated for unreasonable harm caused to them by users of a private copying exception. In view of the minimal impact on sales expected to arise from introduction of this permitted act, and the opportunity that it provides for the value of private copying to be priced in at the point of sale, the Government believes it will cause minimal, if any, harm to rights-holders.

The Government considers that levies or other compensation are neither required nor desirable in the context of a narrow provision that causes minimal harm. Levies are an unnecessary and inefficient tax on consumers. They are unfair to consumers in that they are payable regardless of the use to which a levied device (for example a hard disk) is put and regardless of whether a user has already paid for the copies that they store on a device. Furthermore, particularly in the current economic climate it is not right to extract more money from the pockets of hard pressed consumers”.

(emphasis added)

The Government did not know the exact extent of pricing-in. The conclusion that it occurred only to “some extent” is one which the Claimants highlighted as being insufficient to provide a definitive answer to the very specific de minimis question which had to be resolved if the Government was, lawfully, to achieve its objective.

The Consultation response was accompanied by a second Impact Assessment dated 13^th December 2012 (“the 2012 Assessment”). This contains a more detailed economic assessment of the costs and benefits of the proposed exception. It pulled together the then available empirical data and research on pricing-in and concluded that pricing-in was both possible and relatively widespread though not universal. An important point made was that no useful data had been provided by third parties during the course of the Consultation to enable Government “…to monetise any costs and benefits to rights-holders arising from this exception” (page 16). The IPO nonetheless expected: “…that costs to rights-holders due to lost sales will be minimal or zero”.

Annex A to the 2012 Assessment provided a description of the results of independent research commissioned by the IPO in the field of music, films, books and (as a control) software. That independent research is considered in section E below. The IPO published the full market studies, data sets and statistical tests on its website which led it to the conclusions set out in this assessment. The affected sectors were hence able to access the independent research and comment upon it.

As I explain below an Updated Impact Assessment was published on 23^rd March 2014 which took account of the evidence and submissions submitted subsequently. The Updated Impact Assessment is the pivotal document for the purposes of the present case since it is there that the definitive reasons justifying the decision to introduce the new Section 28B are to be found. Save therefore to record that from this point onwards the independent research was in the public domain it is not necessary to consider the 2012 Assessment further.

Draft legislation was published by the Government in June 2013. This enabled interested third parties to provide their views and comments within an 8 week period. Provision was made for public round-table meetings held in London to discuss the draft. In excess of 140 organisations made submissions.

On 23^rd March 2014 the Government published the Updated Impact Assessment. The reason that this third assessment was carried out was due to a Ministerial commitment made during a debate in Parliament on the Intellectual Property Bill on 18^th June 2013 that the 2012 Assessment would be updated, to take into account the latest evidence. This assessment, as already observed, sets out the full and final reasoning of the Defendant in adopting Option 1 and it necessitates detailed scrutiny. I conduct this scrutiny as follows: (i) the decision to adopt Option 1; (ii) the submissions of the Defendant about the approach to evidence collection; (iii) the findings made in the assessment; (iv) the findings of the independent experts.

The assessment identified the Government’s chosen option as Option 1. This was described in the following terms:

“…a narrow private copying exception, allowing reproduction of a copy of a work (e.g. a CD or eBook) that is lawfully owned (bought or gifted) by an individual onto any medium or device owned or controlled by that individual (e.g. a tablet device or MP3 player) for their private non-commercial use”.

The option would permit someone to “format shift (i.e. shift content from one format to another); or “space-shift” (i.e. move content to different personal devices or media); and “back-up” copies that they have acquired. The exception would apply to storage by the owner on any device or media over which they had control, and this would include a cloud locker. The exception would not extend to anything other than own-use, for example use by friends or family. It would not permit the copying of rented, borrowed, streamed or broadcast content. It would not permit the copying of illegally-acquired content and it would not permit any distribution of copies.

Mr Stout, the Deputy Director from within the IPO, who produced Witness Statement evidence on behalf of the Defendant, stated that the empirical basis upon which the Defendant ultimately formed his view included, inter alia, consumer focus surveys, a survey produced by the University of Hertfordshire, a survey produced by the University of Leeds, a variety of academic papers, submissions to the Defendant by consultees, statements from the music industry in relation to MP3 pricing on iTunes, the response of BPI to the Consultation, independent research commissioned from the team at Sussex University, evidence from PRS and Warner Music as to the contents of cloud storage licensing arrangements, research commissioned by UK Music from Oliver and Ohlbaum on the value of personal media devices attributable to music copying facilities, and anecdotal evidence provided as part of the Consultation process. In paragraph 71 of his Statement Mr Stout stated:

“…no evidence was received during the Consultation process, or otherwise, which caused the Government to change its mind on the conclusions it reached in respect of the harm that will be caused to rights-holders as a result of the proposed exception”.

Mr Stout acknowledged the limitations of some of the empirical data submitted. In particular he addressed what he describes as a “serious error” in research submitted by Sussex University (i.e. the report of the independent researchers). He stated that at the time of the final Report he articulated his concerns about the quality of the data internally and suggested that it might be more helpful to publish a summary of the research but in the event the decision was taken to publish the research in full. The Updated Impact Assessment of March 2014 explained why, contrary to the researcher’s conclusions, the Defendant considered the results of the research to be consistent with the pricing-in theory. In essence, the independent researchers’ conclusion that CDs allowed “no copies” was inconsistent with practice in the market. Mr Stout discussed the erroneous assumption that the researchers had operated upon with the economists working on the issue and they agreed that as CDs were free from technological copy protection measures they should be treated as if copies could be made of them and, accordingly, having similar copying utility to a MP3 album. He observed:

“this mis-classification of CDs as uncopyable explained why the researchers had found it “puzzling” that an MP3 album, with apparently higher copying utility than a CD, was not sold at a higher price”.

Mr Saini QC took me through the consultative history, from Hargreaves onwards, to show that the conclusions in Hargreaves and the evidence and data which underlay those conclusions were part of the background to the final reasoning in the 2014 Updated Impact Assessment which therefore had to be read in that wider context. I accept this submission. It is, in my judgment, clear that the final decision to introduce the exception, as reflected in the reasoning in the Updated Impact Assessment, was the culmination of the exercise which started with the Hargreaves review.

The core economic principle underlying the decision is based on pricing-in:

“A static model assumes that rights-holders are unable to appropriate the value of private copying, so a lost sale resulting from private copying results in lost income. However, the theory of indirect appropriability, first proposed by Stan Liebowitz and since developed by him and others, shows that rights-holders ought to be able to benefit from the additional value that consumers derive from private copying, as it will be priced in to the market through increased prices or sales…

…

The pricing-in hypothesis is described in Box 1. Essentially, a copy of content that can be format-shifted has greater utility than one that cannot. In theory, this will shift the demand curve and a rights holder can charge a higher price for this copy (or sell more copies at the same price), and capture revenue they otherwise would have lost. Liebowitz illustrates this point using the example of someone who buys an audiocassette copy of a musical work to play in the car as well as a CD copy of the same work to play at home. A private copying exception would permit them to buy the CD copy and make their own audiocassette copy instead of buying two copies. The buyer of the CD will value it more, as it is more useful to them, which means the seller of the CD will be able to charge more for it, indirectly capturing the value of the CD and the audiocassette in the price of the CD alone. The utility of copying thus raises the demand curve for content. The outcome is that right holders do not lose income from such copying. Liebowitz describes certain situations in which the rights holder is able to capture more value by permitting such private copying than they could by restricting it”.

The assessment recognised the limitations upon the theory of pricing-in. The wider the copyright exception the less it could be said that possible “harm” to content producers had been priced-in already at the point of sale:

“Theory only predicts overall benefits for rights holders as a result of indirect appropriability in certain, limited cases. When a consumer copies an original copy they have bought, as would be the case under Option 1, we would expect the value of this copying to have been factored in at the point of purchase. But we would expect indirect appropriability to break down if the exception were to permit excessive sharing of copies – the most extreme example being if it were to cover illegal online file sharing. In such a case, so many copies would become available at minimal cost, flooding the market, that prices would cease to reflect the value consumers place on them and would be driven down to the marginal cost for copies”.

The central finding is that the adoption of Option 1 would not, based upon an assumption said to be backed by research, result in the need to introduce a compensation or levy scheme under Article 5(2)(b). Under the heading “Key assumptions / sensitivities / risks” the following is stated:

“This option is assessed to cause minimal harm to rights holders, based on an assumption (supported by research - see below) that the value of private copying to the extent permitted by this extension will be factored in to the market for original copies. An exception that causes minimal harm to rights holders also minimises risks of non-compliance with EU law”.

On page 17 the conclusion (which still only applies “to some extent”) is said to flow from two categories of evidence (i) that which is set out “above” and (ii) “the independent research”:

“In view of the above, and the independent research commissioned by IPO, we expect that the value of the private copying which is permitted by this exception will be factored in to the market, at least to some extent. This will mean that, to the degree that rights holders experience lost income due to lost sales of duplicate copies, they will be compensated for this to – at least to some extent – via a positive impact on prices or sales of content”.

(emphasis added)

On page 18 the following is set out:

“No useful data was provided to the consultation to enable us to monetise any costs and benefits to rights holders arising from this exception. However, based on the above analysis, we expect that costs to rights holders due to lost sales will be minimal or zero”.

The critical section of the reasoning is found on pages 16-18. It is here that the evidence the Secretary of State relies upon is identified. It is therefore necessary to set out, in full, the conclusions about the evidence which is said to justify what are very strong end-conclusions about the extent of pricing-in:

“Evidence for copying being priced in

The question of whether additional copies are priced in to the final price, or indeed the wholesale price charged by producers, is ultimately an empirical one. Respondents to the Government’s consultation made a number of suggestions as to how this hypothesis could be tested.

One approach is to compare downloads sold under different usage restrictions, in particular those whose use is physically restricted by DRM technology. An example of pricing reflecting additional consumer value was observed when iTunes introduced DRM-free music downloads (which can be copied freely for personal, non-commercial use) in 2007. Previously, iTunes music downloads could be copied reasonably freely, but only on to 10 approved devices. When DRM-free copies were introduced they were priced at a higher price ($1.29) than copies which could be copied to a more limited extent due to DRM ($0.99). Thus, tracks that could be copied onto 10 devices (already quite permissive) were priced 30% higher when restrictions were removed to allow any copying for personal non-commercial use, in reflection of their added value to consumers. This example appears to show strong evidence of pricing-in, as theory predicts.

Subsequent to the first publication of this impact assessment in December 2013, some music stakeholders have commented on this example, noting that iTunes dropped its DRM-free download prices from $1.29 to $0.99 in October 2007 (to the same price as DRM-protected copies), and arguing that this demonstrates that consumers are not prepared to pay more for the ability to copy music. We do not consider this to be the case. “Pricing-in” does not simply mean a producer will be able to charge more for more useful product. It also means that a producer which chooses to maintain the same price, for an improved product would expect to sell more units of it. In either scenario, overall revenues to the producer would increase.

So if freely-copiable downloads are made available at the same price as DRM-protected download, we would expect overall demand for copies to increase. This is what we would expect to be the result of the iTunes price change, and is consistent with iTunes’ own explanation for it.

iTunes no longer sells music in DRM-restricted format and most music download licences now permit copying for personal non-commercial use so it is difficult to base any comparative analysis on the current music download market. However, other download formats (e.g. video) continue to use technological measures to restrict the number of copies that can be made, and responses to the Government’s consultation suggested that rights holders in this sector expect as a consequence to charge more for formats which physically permit more copying.

The IPO commissioned independent research based on a comparison of prices of copies of different media sold under different usage restrictions. This research, which is summarised in Annex A, appears to confirm that pricing-in is possible, and is taking place.

As noted above, “pricing-in” is expected to impact on demand as well as price. Therefore, another approach to research in this area could be to look at impacts on consumer demand and sales of copies. As noted by contributors to the Government consultation, without format shifting it is becoming increasingly difficult to play CDs on portable devices, as MP3 players (including mobile phones) become widespread, and demand for portable CD players declines. This suggests that many people buying CDs do so with the expectation that they will format shift in order to play their CDs on a portable device, and that sales of CDs (and prices) would be lower if they could not do this (for example, if CDs came with copy-protection). A consumer survey on purchasing considerations would be a way to investigate this effect.

Two further ways to examine the effect of pricing in were suggested by stakeholders. One was to compare the price of an average digital download album and the price of an average CD album. As the licence attached to a digital download album (licensed for unlimited copying for personal, non-commercial use) permits greater copying than the CD album (not licensed for personal copying), some stakeholders argued that the digital album ought to cost more if there was pricing in. We do not think this conclusion is consistent with the pricing-in theory, however. CD copying for personal, non-commercial use is widely considered to reasonable and takes place regardless of copyright. Therefore, the value of being able to make private copies from CDs is likely to already be factored in to the market, and should be similar to the value of making copies from MP3s. Moreover, other factors raise the price of CDs compared to downloads, including the fact that a CD costs more than a download to manufacture and distribute and the value that buyers of CDs are likely to place on physical aspects including packaging. Therefore, average CD prices are higher than MP3 download prices, and this appears to be consistent with the pricing-in hypothesis.

Stakeholders also proposed to compare pricing between countries with different private copying exceptions, including those with device-based levies and those without. We do not consider such an analysis to be very useful, however, due to a number of factors affecting pricing in different markets, including exchange rate fluctuations.

In view of the above, and the independent research commissioned by IPO, we expect that the value of the private copying which is permitted by this exception will be factored in to the market, at least to some extent. This will mean that, to the degree that rights holders experience lost income due to lost sales of duplicate copies, they will be compensated for this to – at least to some extent – via a positive impact on prices or sales of content.

In addition, some of the savings that consumers make from no longer having to buy duplicate copies of the same content is likely to be spend on alternative copies (see consumer benefits, above).

No useful data was provided to the consultation to enable us to monetise any costs and benefits to rights holders arising from this exception. However, based on the above analysis, we expect that costs to rights holders due to lost sales will be minimal or zero”.

With regard to the independently commissioned research the terms of reference and the instructions given to the experts were included in documents disclosed by the Defendant in these proceedings. The initial terms of reference reflect the clear acknowledgement of the need for additional empirical evidence:

“…the potential effect of private copying to right holders is still unknown. More empirical evidence is needed about the UK market of copyright protected content. This study is aimed at filling this gap, providing empirical evidence on the price of copyright content and terms associated with copying”.

The instructions made it perfectly clear that the purpose of the research was to “test whether allowing a certain degree of copying is reflected in the price for content”. Mr Saini QC, for the Secretary of State, explained to me that this was a reference back to the conditional conclusion of Professor Kretschmer in his 2011 Report at page 19:

“7.

Distinction between “priced into purchase” and “statutory licence”

Reconsider the consumer activities listed in section two above. For (i) [Making back-up copies / archiving / time shifting / format shifting]; and (ii) [Passing copies to family / friends], a certain amount of copying appears to be already priced into the purchase (Varian’s argument). For example, right holders have either explicitly permitted format shifting, or decided not to enforce their exclusive rights. There is no lost sale, and the European criterion of harm may be treated de minimis, i.e. no compensation is due. Commercial practice will not change as a result of introducing such a narrowly conceived private copying exception.

A more widely conceived exception that would cover private activities that take place in digital networks [activities (iii) to (vii)] might be better understood as a statutory licence. Possible rationales for issuing such a licence include: making the copyright system more permissive for consumer led innovation, as well as non-economic arguments (such as influencing the bargaining position of creators versus producers, or preserving fundamental rights of privacy). The EU concept of “compensatable harm” contributes little towards assessing an appropriate scope and tariff for such a licence. There is no case for copyright levies unless the payment of levies is linked to clear consumer permissions, and an argument is made why scope and tariff of these permissions cannot be left to the market”.

As this quotation makes clear Professor Kretschmer was not saying that pricing-in was proven to be a universal practice but only that a “certain amount” of pricing-in occurred.

The experts appointed to perform the empirical research task were Roberto Camerani, Nicola Grassano, Diego Chavarro and Puay Tang, from Sussex University. I refer to their Report henceforward as the “IPO Research Report”. In the assessment a summary is provided in Annex A. I turn now to consider Annex A before considering the IPO Research Report itself.

In Annex A the IPO made reference to the size of the data set collected by the researchers. This embraced 18,958 “observations” which incorporated 17,272 albums. Under a heading “What the data suggests” the IPO drew the following conclusion which included the doubts of the civil servants as to the correctness of the researchers’ conclusions:

“Music: the most significant finding of the research on music came from the work to try and understand the copying conditions contained in the terms and conditions of sale of different types of music. More than in any other type of content there was real difficulty in interpreting the small print in the terms of different suppliers and formats. Often the terms were confusing (for our researchers let alone ordinary consumers) and in some cases internally contradictory. In part we suspect that this reflects two basic facts of the music market:

-

The ability to control copying as part of the business model, through DRM, is absent from most of the digital music market.

-

Most consumers do not accept (from the research cited earlier) that there should be limits on their ability to copy for personal use.

Variation in prices which might indicate “pricing in” of the ability to copy is difficult to find in music content. This could be because private copying for personal use is so widespread in the sector that it is already largely or fully priced in. Consumers and producers of music appear to expect that music will be copied for personal non-commercial use, and all downloads are explicitly sold on these terms; so we would expect to see little variation in prices compared to other types of content, particularly those which physically restrict and permit different degrees of copying such as film downloads and eBooks.

We note that, when attempting to draw conclusions from their pricing analysis, the researchers appear to make the assumption that a CD which does not, by way of a licence, expressly allow its owner to make personal copies will not be copied, so the value of copying will not be priced-in. Given that consumers commonly format-shift CDs, regardless of the law, we do not consider this to be a valid assumption. As explained above, the current law on private copying, and the terms of use attached to copies, are unlikely to factor strongly in purchasing decisions as most consumers are either unaware of them, or ignore them because they consider them unreasonable.

If we instead assume that consumers do expect to be able to copy CDs, regardless of licence terms (which is what consumer surveys suggest) the data gives the result we would expect. Both the CD and the download have similar copying utility, so this does not noticeably affect the price differential. Instead, the higher cost of the CD is likely to derive from its physical appeal to consumers and higher manufacturing costs”.

The other areas examined by the researchers were films, books and (the control) software.

In relation to films, the researchers examined 3,515 products which included blue-ray, digital copies, DVD, VHS and bundles where a film was offered in multiple formats within a single package. The IPO recorded the researchers’ basic conclusion which was that:

“There is a positive correlation between the retail price and number of physical copies included in a bundle, but the relationship appears to be different for the number of copies allowed. This is explained by differences in the price premiums charged in bundles for different formats. So when an additional standard DVD or Blu-Ray disc is added, the average price of film bundle is negative or statistically insignificant; the price rises with an additional digital copy (£3.33), 3D Blu-Ray discs (£4.30) and Ultraviolet licences (£5.42). Ultraviolet is particularly interesting as it offers a very broad license with options to watch films on multiple devices, with up to five copies in some cases, and a possibility to share the license with up to five other people. These regressions explain 28% of the variation in price, and running the same tests for the natural logarithm of prices (testing to see if there is a percentage change in the average price) produces much the same results with the same explanatory power”.

In relation to books the data set covered 2,071 books including physical and eBooks. The conclusion was that:

“There is a statistically significant average price difference between physical books (£12.87) and e-books which allow 6 copies (£6.69) and those which allow five copies (£5.43). The premium for a physical book is of course likely to be explained by manufacturing costs and the value consumers place on physical features, rather than the number of copies that can be made from it. The number of allowable devices does not appear to influence the price, and regression analysis on the price does explain 34% of the variation in price. That said, while the number of copies for e-books appears significant, it could be driven by Amazon’s strategy to offer lower cost books (similar to Amazon results for software), and the number of copies allowed may have no bearing on this. Similarly differences between retailer and publisher can change the copy permissions on individual books, but these conditions are usually only available on purchasing each book, and so not useful for analysis without further publisher figures”.

Finally, the IPO records the researchers’ conclusion on the control category of software (a control because software is not within the scope of the section 28B exception). The data set included 1,008 pieces of software. The conclusion was:

“This market case is partly a control case as it would not notionally be affected by a private copying exception, but it is a sector which actively tries to price additional copies. Regression analysis which explains 55% of the variation in price suggests that an additional copy is associated with a 1% increase in price holding all other criteria constant. Moreover, when adding the fact that the price increase seems to diminish with more copies a quadratic term is introduced into the regression which confirms that hypothesis, and does not change the model except to increase the pricing-in association to 3%”.

On 12^th February 2014, shortly prior to the adoption of the Updated Impact Assessment, the IPO received from the Motion Picture Association of America (“MPAA”) a short Report addressing the economic impact of the private copying exception on the film and television industry. It referred to “new” evidence but, of greater significance, it also referred to evidence that the Defendant had specifically relied upon in the Updated Impact Assessment which it was now said had been misconstrued. In particular it argued that the inferences which the Defendant purported to draw from a particular survey could not in actual fact properly be drawn therefrom. This Report is considered later in this judgment at paragraphs [257] - [261].

The Updated Impact Assessment noted that a consumer survey on purchasing considerations would be one way to investigate the pricing-in effect (see the quotation at paragraph [100] above (see the seventh paragraph therein)). No such survey evidence was, however, proffered to the IPO. In April 2014 UK Music informed the IPO however that it had commissioned just such a survey. An executive summary of a Report prepared by Compass Lexecon/FTI was provided on 30^th April 2014. Further background material was supplied on 6^th May 2014.

Mr Stout explains that the Government’s initial view was that this Report appeared to indicate that consumers were willing to pay more for materials they could copy but that there appeared to be a number of gaps and flaws in the assumptions relied upon in the Report and its survey methodology and analysis.

The independent economic analysis commissioned by the Government: The “IPO Research Report”

I turn now to the IPO Research Report. The relevance of this is that, as I set out below in Section I, the Claimants submit that the inferences drawn from this research set out in the Updated Impact Assessment are simply unwarranted when the Report itself is considered. It is accordingly necessary to examine with care this evidence and then to compare and contrast its findings with those in the Updated Impact Assessment (I set out conclusions in section I below at paragraphs [264] – [267]).

The specific terms of reference for this research have been set out at paragraph [101] above. In the Report itself the researchers summarise their remit as: “Whether and how four copyright industries – music, film, publishing and software – have (or not) adopted private copying measures, and whether the price of the products in the UK reflect a right to private copying”. They do not express their remit in terms of an evaluation of the actual extent of pricing-in and whether it satisfies a de minimis threshold.

The inclusion of software as a control case was considered to be relevant as a benchmark because it was a sector which actively sought to price-in additional copying. It is, therefore, convenient to start with this “control case” as the benchmark for assessing the conclusions in the other sectors which were within the scope of the proposed exception.

The researchers drew their conclusions from in excess of 1,000 observations contained upon their database. The type of software spanned the full gamut from operating systems, productivity, utilities, design and publishing, photography, finance, music and audio and others. For each the database identified whether the delivery mechanism was download with or without a back-up disc, and CD/DVD. The average price of the product was logged as the average number of copies and the average price per copy. The analysis showed that products sold through direct download permitted a larger number of copies than those sold on a physical medium. The researchers correlated the relationship between copying and price. Their conclusion (cf. page 14 figure 2a) was that: “it is clearly observable that the price increases with the number of copies allowed”. To confirm their conclusion the researchers conducted a regression analysis using as the dependent variable the natural logarithm of price as opposed to the price itself. This was because an analysis of the frequency distribution of price showed that it was “very skewed” (page 15) indicating that distribution was not symmetrical. The researchers concluded that calculating the natural logarithm of price was one potential solution for this type of asymmetrical price distribution. The regression analysis using the logarithm of price as the dependent variable showed a more normal distribution of product “…with a better defined bell shaped distribution thereby solving the skewness”. The researchers then applied three different regressive models using a variety of variables (cf. page 24). One model (Model 3) added a quadratic term to the regression (i.e. number of copies squared) to capture a decreasing or increasing marginal effect. This had the result of demonstrating that the price per copy appeared to decrease with an increase in the number of copies. The inclusion of the quadratic term enabled the researchers to test if this effect was influencing the price of software. The conclusion which emanated from this analysis, which might (from a common sense perspective) not be viewed as particularly surprising, was that software producers charged users to make a single additional copy of the software but the additional, incremental price gradually reduced with the increase in the number of copies permitted. In other words, this reflected the fact that the value to consumers of the right to copy decreased as the number of copies increased:

“…analysis suggests that software producers charge a price that reflected the possibility of making additional copies. In other words, a higher number of copies is associated with an increase in the price. However this effect becomes smaller as the number of copies increases”.

(ibid, page 24)

It can properly be concluded that the research showed that in relation to the control, software, vendors did in fact seek to price-in subsequent copying. The control thus supports the basic economic intuition which is that wherever they can vendors will seek to capture the value of such copying as is taking place on the part of consumers, irrespective of whether that copying is legal. The researchers did not, however, go further and provide an answer to the question whether pricing-in was so endemic that it obviated the risk of “harm” remaining uncompensated for.

I turn now to the findings in relation to music (cf. pages 6-11). The researchers focused upon a large sample (17,272) of digital albums. The results here were said to be “ambiguous” (cf. page 11). A number of possible explanations for these ambiguous results were posited. In particular, the analysis could have been affected by a series of omitted variables. Further, music was essentially an “experience” good whose value depended upon a variety of subjective (“latent and intangible”) factors such as taste, fads, etc which were difficult to measure. Moreover, other potentially relevant factors were theoretically measurable but difficult to collect in practice (such as advertising expenditure, promotional campaigns, etc). Finally, and perhaps most pertinently, all the stores sampled permitted the possibility to copy digital music on to a CD or other portable device for personal use and the exact number of copies could not be quantified since the limit was defined as “personal use”. The general approach of the researchers was to start by examining the effect of copying on price. This, however, did not provide “…any evidence in support of a widely-held view that online stores charge higher for additional copies” (cf. page 9). In order to advance the price/copies correlation analysis the researchers applied 2 series of regressions using two dependant variables viz, album price and average track price, and then they estimated four models for each defendant variable (cf. page 9 Table 5).

The conclusion of the researchers (cf. page 4) in relation to music was cast in the following terms:

“We did not find any evidence in support of a widely-held view that stores are including in their price the permission to copy. All digital albums are allowed to be copied for personal use, as opposed to CD albums that cannot be copied. In spite of this we found that digital music is associated with a lower price compared to a CD. Whether the lower price depends on the store, on the intangible nature of the product, or on the possibility to copy, cannot be determined due to the lack of variability in the explanatory variables. However as private copying for personal use is widespread and allowed in the UK, it is plausible that private copying is already largely or fully priced in the UK market. Consumers and producers of music appear to expect that music will be copied for personal non-commercial use and all downloads are sold on these terms”.

The researchers thus concluded (cf. page 11) that the results in relation to music were “puzzling” and “ambiguous”:

“Overall, the analysis on the impact of the number of copies on price for a sample of digital and CD albums has produced puzzling results. These could be explained by a number of factors. First, the analysis is probably affected by an omitted variable issue. The R2 shows the percentage of the variability in the dependent variable (price) which is captured by the model. The models estimated using price as dependent variable capture approximately 40% of price variability (and much less in the case of price per track). Second, a main problem with music recordings is that music is essentially an “experience” good whose value depends on several factors. Some of these factors are latent and intangible (such as tastes, fads, consumers’ subjective preferences, etc.) and are very difficult to measure. Third, other factors are technically measurable but very difficult to collect (such as advertising expenditures, radio plays, promotional campaigns, etc.). The fourth factor pertains to the variability of the explanatory variables. All the stores sampled allow the possibility to copy the digital music into a CD or into a portable device for personal use (with the only exception of 7Digital, which does not explicitly mention the possibility of burning a CD with the purchased music, already noted above). This creates a problem as it is not possible to control for both the copying possibilities and the store at the same time because these variables would overlap. Moreover, the exact number of copies cannot be quantified since the limit is ‘personal use’.

In summary, none of the results seems to support the claim that online stores are charging for digital copies. Therefore, according to our sample, an assumption or view that online stores embed an additional cost into their product price for copying remains ambiguous”.

The conclusion on music may, fairly, be said to provide no empirical evidence that pricing-in occurred but more importantly, none that showed that even if it did occur it was so extensive that it rendered “harm” minimal or zero.

In relation to films, the data set consisted of 3,515 products divided into 5 categories: Blu-Ray discs, DVDs, VHS tapes, digital films, and, bundles of products including one or more physical media but which might include the right to download a digital copy. It is evident (cf. pages 12-16) that the analysis was easier to conduct than in the case of music because here the researchers were able to determine average prices. This analysis led to the following conclusion:

“The product price and average price per disc follow a very similar pattern. The relationship between number of copies and price appears to be a positive. But this result holds only up to four copies. An increase in the number of copies seems to be associated with an increase in the product price. However, the average price dramatically drops when the copies are equal to five. One plausible explanation for this could be related to the fact that the only kind of product allowing five copies is iTunes digital films. In the case of digital films it is possible to create an exact copy of the movie with virtually no additional costs, while in all the other cases, one additional copy means one additional physical medium. Therefore, according to figure 1 above, film sellers seem to establish a price that incorporates the possibility of having one or more additional physical copy of the films, and not for the option of making more digital copies”.

The authors tested whether the number of copies exerted an impact on product price by performing regression analysis of the film data set. They applied two models with different variables in each. This produced interesting results. In particular, it revealed that film publishers operated sophisticated pricing regimes which reflected the formats included within the bundle of rights. For example, digital copy, Blu-Ray 3D, and UltraViolet digital copy exerted a positive effect on price; but this was not so evident in relation to DVDs or ordinary Blu-Ray. The authors concluded:

“This seems to indicate that film publishers charge users not simply to have more copies, but to have a higher variety of available choices. This is also confirmed by the fact that the more sophisticated kinds of formats, such as Blu-Ray 3D and UltraViolet Digital Copy, are also those with the highest coefficients”.

The researchers confirmed these conclusions when the same analysis was performed with the dependent variable in natural logarithm which meant that the effect of an increment in the independent variable could be interpreted as percentage changes. Table 10 (cf. page 16) showed that, expressed in this way, the results were more or less the same as those generated using normal regression techniques (albeit with some minor differences). The researchers thus found evidence of pricing-in but were not able to say whether it was so widespread as to lead to only minimal or zero residual harm. The end conclusion to this was expressed in the following terms:

“In parallel with traditional films, which come with only one licence (e.g. Blu-Ray and DVD), the film industry has found different ways to allow users to have more than one copy of the films they purchase. Examples can be found in the sales of digital films (which could be copied on to a number [of] devices), by bundling multiple formats in the same package (which may also include a complementary downloadable digital copy), or through new kinds of licences, such as UltraViolet. We found a positive relationship between price and film copies. In particular, since different types of formats have different impacts on price, we can conclude that film stores charge users not simply to have more copies, but to have the opportunity to choose from a higher variety of available choices”.

In relation to the publishing industry (books) the researchers conducted similar forms of analysis to that in relation to films, music and software. The results highlighted a number of obvious differences between physical books (which have significant production costs) and eBooks which do not. They observed that there was a price premium for physical books which was explained by manufacturing costs and the value consumers placed on physical features rather than upon the number of copies that could be made from it. In relation to digital books they observed that all of the book sellers that they examined in the sample permitted users to make copies of the books. There was not however a great deal of information enabling them to compare and contrast prices for digital books with and without copying rights. From a statistical perspective they concluded that the analysis of the relationship between book prices and the number of copies allowed produced “indeterminate results” (cf. page 21). In relation to digital books these were, almost universally, allowed to be copied for a specified number of times. To this extent, the copying right was, inevitably, already factored in to the initial price. There was however little information which enabled the researchers to examine the relationship between price contingent upon variables in the number of times that a book could be copied.

Issue I: The relevant legal principles

There are 3 main areas of legal principle which govern the issues in this case:

The Claimants submitted that in relation to all of the issues arising the Court must apply a full “merits” review, i.e. decide the issues on a “correct or incorrect” basis and not therefore according any margin of appreciation to the decision maker. As an alternative it was submitted that insofar as the Defendant had a discretion it was very limited and consequentially the review to be performed by the Court was very intensive. The Defendant, to the contrary, submitted that this was an area where the Secretary of State had a very broad margin of appreciation because, in particular, it was evident from Article 5(2)(b) and the recitals to the Directive that in determining where harm lay the Member State was permitted and required to take into account the complex socio-economic balance which existed between consumers and copyright holders. It was submitted that this meant that Member States enjoyed a very wide margin of appreciation which, in turn, implied a high degree of deference on the part of the Court. Mr Saini QC went so far as to say that the Court was not “competent” to determine whether the Defendant had erred or not. There is, accordingly, a sharply defined difference of view as to the approach to be applied between the parties.

Mr Mill QC and Mr de la Mare QC, for the Claimants, both orally and in writing, set out comprehensively the reasons why they submitted: (i) the issues arising were, in effect, black and white issues for the Court to decide with no deference being owed to the Defendant as decision maker; or (ii), insofar as some deference was due and owed that it is was very modest.

In support of these submissions the principal points relied upon by the Claimants were as follows.

First, the Directive demanded a standard of protection for rightholders which could only be guaranteed by close scrutiny by the courts of decisions affecting those rights. The general policy of the Directive is reflected in Recital (9) which provides that harmonisation “must take as a basis a high level of protection, since such rights are crucial to intellectual creation”. Further, Article 5(2)(b) was a derogation from a basic norm and therefore had to be interpreted narrowly and strictly, implying rigorous judicial control. Article 2 required Member States to secure to rightholders “the exclusive right to authorise or prohibit direct or indirect, temporary or permanent reproduction by any means and in any form” of their works. That exclusive right is subject only to the exceptions in Article 5. Recital 32 makes clear that those exceptions are exhaustive: “This Directive provides for an exhaustive enumeration of exceptions and limitations to the reproduction right […]”. Recital 42 emphasises the need for Member States to arrive at a “coherent” application of the exceptions and limitations. Article 5(5) provides that exceptions must only be applied in certain cases, not conflicting with normal exploitation and not unreasonably prejudicing the legitimate rightholders’ interests. The Claimants submitted that: “Those constraints forbid a relaxed approach to the assessment of a Member State’s compliance with the requirements of the Copyright Directive. The starting point is that the Copyright Directive requires a high degree of copyright protection and that the ability of a Member State to depart from that high standard is tightly constrained”. In Case C-5/08 Infopaq International A/S v Danske Dagblades Forening (16^th July 2009) at paragraphs [56] and [58] the Court of Justice stated:

“56…the provisions of a directive which derogate from a general principle established by that directive must be interpreted strictly”; and,

“58.

This is all the more so given that the exemption must be interpreted in the light of Article 5(5) of Directive 2001/29, under which that exemption is to be applied only in certain special cases which do not conflict with a normal exploitation of the work or other subject-matter and do not unreasonably prejudice the legitimate interests of the rightholder”.

Secondly, the Directive incorporated a presumption that the introduction of any exception under Article 5(2) to the exclusive Article 2 rights caused harm to the rightholder and accordingly Member States could only avoid introducing compensation schemes if they had cogent evidence demonstrating an absence of harm. This followed from Articles 5(2)(b) and Recital 38 which provided that a private copying exception could only be established, “on condition that” rightholders receive “fair compensation”. Article 5(5) limited the scope of any exception to copyright protection to “certain special cases which do not conflict with a normal exploitation of the work or other subject-matter and do not unreasonably prejudice the legitimate interests of the rights-holder”. Recital 35 made clear that, unless some other payment to rightholders has already been made, making no payment to rightholders after the establishment of a private copying exception is permissible only in the exceptional case of a particular activity where the prejudice to rightholders is “minimal”. The Claimants cited the Court of Justice in Case C-467/08 Padawan [2010] ECR I-10055 (“Padawan”) at paragraph [44] which held that: “…copying by natural persons acting in a private capacity must be regarded as an act likely to cause harm to the author of the work concerned”. It followed that the introduction of an exception must itself be presumed to cause harm to rightholders. The Court also noted at paragraph [57] that Recital 35 mentioned, as a valuable criterion for the determination of the level of fair compensation, not only ‘harm’ but also ‘possible harm’ – from which it followed that harm may be relevant even if it is not proven to have occurred.

Thirdly, this is a case of “jurisdictional fact” given that proof of de minimis harm is a pre-condition of the introduction of a copyright exception which is unaccompanied by a compensation scheme. English law recognises that in cases of precedent or jurisdictional fact the Court will determine on a definitive basis whether the condition precedent exists, or not. The assessment of harm is not therefore a matter for the discretionary judgment of the authorities of a Member State absent full verification of the correctness of that decision by the Court. The Claimants relied upon Recitals 6, 31, 32 and 38 which they said must be construed as supporting the conclusion that the appraisal of harm was a hard edged question and a decision by a Member State on this was justiciable as to its merits. Support is also said to arise from paragraph [36] of Padawan where the Court, in answer to the first question referred, held that ‘fair compensation’ is an autonomous concept of EU law which had to be interpreted uniformly in all the Member States that had introduced a private copying exception:

“… although it is open to the Member States, pursuant to Article 5(2)(b) of Directive 2001/29, to introduce a private copying exception to the author's exclusive reproduction right laid down in European Union law, those Member States which make use of that option must provide for the payment of fair compensation to authors affected by the application of that exception. An interpretation according to which Member States which have introduced an identical exception of that kind, provided for by European Union law and including, as set out in recitals 35 and 38 in the preamble thereto the concept of 'fair compensation' as an essential element, are free to determine the limits in an inconsistent and un-harmonised manner which may vary from one Member State to another, would be incompatible with the objective of that directive….”.

(Emphasis added)

The Claimants rely also on Case C-462/09 Stichting de Thuiskopie v Opus Supplies (“Stichting”) where the Court at paragraph [34] stated:

“…unless the [safeguards provided by Article 5(2)(b)] are to be deprived of all practical effect, those provisions impose on a Member State which has introduced the private copying exception into its national law an obligation to achieve a certain result, meaning that it must guarantee…the effective recovery of the fair compensation intended to compensate the authors harmed for the prejudice sustained”.

That ‘guarantee’ can be effective only if the decisions that give effect to it are properly reviewable by a court.

Fourthly, the conclusion that the issue is a hard-edged question for the Court is supported by the fact that copyright is internationally acknowledged to be a property right. Articles 16.2 and 47 of the Charter of Fundamental Rights and Freedoms of the European Union (“the Charter”) require intellectual property to be protected and require ‘effective protection’ of rights conferred by EU law. There is no effective protection of intellectual property rights if those rights can be abrogated upon the basis of a decision on a question of fact which is incapable of being properly reviewed, save in the case of irrationality. In addition under TRIPS, and under the Berne Convention ‘Three-Step Test’ (which TRIPS enshrines and which is also given effect to in Article 5(5) of the Directive), the decision to introduce an exception without compensating rightholders engages international obligations going beyond those recognised in the Directive itself. Finally, Article 1 Protocol 1 of the European Convention on Human Rights (“ECHR”) is engaged since a property right is being expropriated without compensation and this curtails the discretion of the State to take expropriating decisions without very careful scrutiny by a reviewing court. The Claimants refer to case law under the ECHR where the Strasbourg Court decided expropriation cases on the merits, e.g. Sporrong and Lonnroth v Sweden (1982) 5 EHRR 35.

Finally, it was submitted that even if the Defendant enjoyed a margin of appreciation that discretion was highly circumscribed. The Claimants relied upon all of the above factors in support of this lesser proposition.

My conclusions on the issue of the standard of review are as follows:

In the text below I set out the factors I consider to be relevant to the intensity of the review to be applied to issues of fact and assessment and reasoning (Issues III and IV). The sorts of relevant governing factors to consider include the nature of the decision; the identity of the decision maker; the reasons for the decision; and the effect of the decision: See e.g. R(on the application of Rotherham MBC) v Secretary of State for Business, Innovation & Skills [2015] UKSC 6 (“R(Rotherham MBC)”) endorsing the Court of Appeal at [2014] EWCA Civ 1080 at paragraphs [54ff]. In the present case I approach this issue by considering the following which seem to me to be most relevant to the question of intensity:

The nature of the issue being decided: The limits of the Secretary of State’s discretion will, in large measure, turn upon the legal issue in question. It is trite that the margin of discretion allowed by EU law may be broad or narrow depending upon the circumstances (e.g. R(Rotherham MBC) in the Court of Appeal at paragraph [54] per Lord Dyson MR). Most issues of policy in this area were predetermined by the Member States (participating in the Council of Ministers) and by parliamentarians in the European Parliament when they adopted the Directive. At that stage decisions were taken as to the scope of the Directive including as to the economic sectors to be covered. The decisions then taken also addressed whether there should be exceptions to the core right (in Article 2) to prohibit unauthorised reproduction and, if so, in what field(s) and for what activities and, critically, whether if exceptions were to be introduced they were mandatory or optional. Policy decisions were also made at that time as to whether if exceptions were implemented they should be subject to a right of compensation or not. For illustration should copying by libraries or archives be subject to compensation? Each of these issues involved large issues of policy. These “big picture” issues are now no longer on the table.

However not every issue was predetermined since one “policy” decision taken when the Directive entered into force was that certain issues were best left to the individual Member States themselves to decide upon. This included whether to introduce the exceptions in Article 5(2).

Thus: the UK had a discretion whether or not to implement a private usage exception under Article 5(2)(b); and if it did so it then it had a further right to decide only to implement one part or facet of that wider private usage right. In this case the UK both exercised the basic right under Article 5(2)(b) and then chose to introduce only a limited personal private usage right exception. There is no challenge to either of those two particular exercises of discretion. Thus it is not said that the Defendant erred in deciding to introduce a private use exception; nor is it said that the decision to introduce a specific and limited version of private use exception was per se unlawful.

Equally there is no dispute between the parties that if the introduction of a limited personal private use exception is introduced and causes some relevant harm which is more than de minimis then compensation must be paid to the rightholder.

As such, this case boils down to the narrow question: Whether on the facts there is evidence of harm beyond the de minimis level for which no compensation mechanism has been provided?

Whether the decision incorporates any margin of appreciation: The policy content of this issue can fairly be described as micro, not macro. This case is clearly distinguishable from cases such as R(Sinclair Collis) v Secretary of State for Health [2011] EWCA Civ 437 (“Sinclair Collis”) (whether bans on certain types of cigarette marketing should be introduced in the name of public health) or R(Gibraltar Betting and Gaming Association Ltd v Secretary of State for Culture, Media & Sport [2014] EWHC 3236 (“Gibraltar Betting and Gambling”) (whether Parliament could introduce legislation to regulate off-shore internet gambling in the name of fraud prevention and consumer protection) or R(Rotherham MBC) (the allocation of EU structural funds between the countries of the United Kingdom and the English regions), all of which were concerned with large “macro” questions of policy.

In this case the Defendant acknowledged that EU law “limited” the scope of his discretion and that he therefore had to address his mind to a narrow de minimis issue: See Updated Impact Assessment page 21 (bottom paragraph). However, this does not lead to the conclusion that a merits review follows. This is not like the de minimis principle which is applied under other equivalent tests in EU law such as Articles 101 and 102 TFEU (the competition provisions) where decades of experience have enabled regulators to reduce the de minimis principle to near-mathematical market share and value thresholds. In Article 5(2)(b) the concept is much more elusive. The de minimis concept is not defined in the Directive. The Court of Justice has ruled that where the Directive does not define a matter Member States have a discretion (see case law cited at paragraphs [194] – [198] below) to choose their own parameters. The Secretary of State thus had a certain margin of appreciation to select a sensible de minimis threshold. In defining de minimis a Member State must strike a “fair balance” between the competing interests of consumers, rightholders and manufacturers of copying devices: cf. Recitals 31 and 35. All of this means that even within the confines of a narrow legal issue there are still choices to be made by the Secretary of State and that the Court should exercise caution in second guessing those choices. Yet these are still choices to be exercised within relatively tight bounds, in particular because the property rights of natural and legal persons are at stake and because this is an exception from a basic norm and because if harm exceeds a de minimis level compensation must be paid. In large measure I accept the submissions of the Claimants on these points but to the effect that the factors they rely upon limit the Member State’s discretion but not to the extent of making the judicial review a merits exercise (See paragraphs [128] – [134] above). On the sliding scale of margins of appreciation this is not a case of “high level policy and economic, social and political judgments” (R(Rotherham MBC) [2014] EWCA Civ 1080 at paragraph [57] per Lord Dyson MR). This is towards the other end of the scale.

The “economic” nature of the decision: The Defendant contended that since the issue was “economic” it was by definition complex and this entitled the decision maker to a wide margin of discretion. In my view the mere fact that the decision was taken on the back of “economic” evidence is not, in and of itself, decisive. It is an error to suggest that simply because the subject matter of a decision, or the evidence used to justify it, is “economic” or “technical” that courts should recoil in terror and move gratefully into judicial reticence mode by reference to “margin of appreciation”. If this were the judicial default position Courts would find it hard indeed to hold in favour of Claimants in clinical negligence cases where, almost invariably, the case turns upon complex scientific evidence. In R(Rotherham MBC) the Supreme Court recognised the dangers of “judicial timidity” ([2015] UKSC 6 paragraph [65] per Lord Neuberger). Decisions of the utmost importance to individuals, to companies and to society are routinely “economic” and “technical” and errors in those decisions should be as much susceptible to judicial review as other equivalent but less technical decisions. There should be no lacuna in judicial review simply because the nature of the decision under challenge is a difficult one.

But this does not imply that the Courts will substitute their own view of the correct decision for that of the decision maker. There is a wealth of difference between the court exercising proper supervisory jurisdiction over an “economic” decision and a court acting as the decision maker itself. Courts are well able to: ensure that the decision maker has collected sufficient relevant evidence to meet the needs of the decision to be taken; ensure that the inferences that are drawn from the evidence are rational and reasonable and within the boundaries of that which may properly be inferred; and assess whether errors that are made are material to the final decision. Equally a court will ensure that, where appropriate, a proper procedure has been followed to collect and review the economic evidence and that affected persons have been treated fairly.

In Gibraltar Betting and Gaming at paragraph [100] in the quite different context of a challenge to legislation where an exacting standard based upon “manifest error” was required before a Court would strike down or declare invalid legislation, I stated that even within this context:

“an error which is far from being obvious or palpable may nonetheless prove to be fundamental. […] The fact that the calculation is complex and that only an accountant, econometrician or actuary might have exclaimed that it was an ‘obvious’ error or a ‘howler’, and even then only once they had performed complex calculations, does not mean that the error is not manifest. […] An error will be manifest when (assuming it is proven) it goes to the heart of the impugned measure and would make a real difference to the outcome”.

See also at paragraphs [96] – [98] in relation to the application of a “thorough” review process even to an area where the decision maker had a very broad margin of appreciation. A Court may be thorough in its review of evidence even where the decision maker has a margin of appreciation.

The effect on rightholders: I accept the thrust of the Claimants’ analysis summarised at paragraphs [130] and [133] above but in the context of the need for a relatively intensive and careful review process. In particular, it is appropriate to conduct a thorough and close analysis of the evidence because: the consequence of the Defendant making an error is to strip proprietors of valuable and important property rights; because it is in the best interests of the economy for rightholders to be properly rewarded and protected since this is integral to the system of rewards for creativity which intellectual property laws respect; because copyright is a right protected under Article 1 of the First Protocol to the ECHR and under the Charter and it is EU policy to ensure observance of that principle; and because Article 5(2)(b) is an exception to a norm which should be construed strictly. In more concise form Lord Wilson made a similar point in relation to the Court’s supervision of consultation exercises: See citation at paragraph [167] below.

The relatively intensive review: In the present case in view of the above considerations I propose to review, carefully and thoroughly, the evidence relied upon by the Defendant to justify the decision to introduce section 28B. I am not conducting this exercise to see, for instance, whether I agree with the conclusion arrived at by the experts who advised the Defendant. I am concentrating upon: the process of evidence collection; its fairness and thoroughness; any acknowledgments as to its inherent limitations; and the inferences drawn from the evidence by the Defendant which ultimately led to the decision in issue.

The next issue concerns “direct effect”. The Intervener submitted that the right to compensation in Article 5(2)(b) was a sufficiently precise and unconditional obligation imposed for the benefit of individuals (i.e. rightholders) that it had “direct effect” which meant, in EU law, that it was capable of being invoked against the State in national law proceedings. A corollary of this was that the national court had to determine the issues arising on a merits basis since it was only upon that basis that the Courts could guarantee effective protection to the directly effective rights contained in Article 5(2)(b). Mr Kennelly (for the Intervener) also relied upon the direct effect of Article 5(2)(b) as part of the State aid argument (See e.g. paragraph [294] below). Mr Saini QC, for the Defendant, argued against the direct effect of Article 5(2)(b).

For the reasons set out below I do not accept that the provision has direct effect. I accept that it is intended to confer upon rightholders specified economic benefits (i.e. compensation) but the obligation is not sufficiently precise or unconditional in the way that is required by case law with regard to (i) the content of the obligation; and/or (ii) the identity of the person required in law to pay.

Member States have a duty to achieve the objectives set out in directives through their implementing measures. In the present case where a copyright exception is introduced under Article 5(2) the Member State is under an obligation to “…guarantee the effective recovery of…fair compensation”: Stichting (ibid) paragraph [34]. But where there has been a failure (whole or partial) to transpose a directive then in principle private persons who are affected may rely upon the direct effect of the directive in place of the transpositional measure. For a measure of EU law to have direct effect it must meet the test of precision and unconditionality. Guidance on this test is given by case law.

In the seminal Joined Cases C-6/90 and C-9/90 Francovich v Italian Republic [1991] ECR I-5403 (“Francovich”) a directive was intended to guarantee employees a minimum level of protection in the event of the insolvency of their employer. The directive left it to Member States to lay down detailed rules for the organisation, financing and operation of the guarantee institutions and conferred on them a discretion including as to who was to pay or fund the guarantees. At paragraph [10] the Court explained that “direct effect” was concerned with answering the question whether: “…the persons concerned can enforce those rights against the State in the national courts in the absence of implementing measures adopted within the prescribed period”. This made clear that the concept of direct effect is essentially concerned with locus. The Court then (ibid, paragraph [11]) articulated the test for direct effect:

“11.

As the Court has consistently held, a Member State which has not adopted the implementing measures required by a Directive within the prescribed period may not, against individuals, plead its own failure to perform the obligations which the directive entails. Thus wherever the provisions of a directive appear, as far as their subject-matter is concerned, to be unconditional and sufficiently precise, those provisions may, in the absence of implementing measures adopted within the prescribed period, be relied upon as against any national provision which is incompatible with the Directive or insofar as the provisions of the Directive define rights which individuals are able to assert against the State (judgment in Case 8/81 Becker v Finanzamt Munster-Innenstadt [1982] ECR 53)”.

Thereafter, the Court applied the test to three aspects of the legislation:

“…the identity of the persons entitled to the guarantee provided, the content of that guarantee and the identity of the person liable to provide the guarantee”.

With regard to the question of the “identity” of the intended beneficiaries of the measure, the Court held that the law was sufficiently precise and unconditional. It sufficed (ibid, paragraphs [13] and [14]) for the category to be as broad as “an employed person under national law”; there being no need for any greater degree of precision in the categorisation of the intended class of beneficiaries. That categorisation enabled the national court to determine whether a particular person should be regarded as a person intended to benefit under the Directive (ibid, paragraph [14]).

With regard to the “content of the guarantee” the Court explained that the fact that a Member State had discretions was not, per se, a reason to deny the provision direct effect. On the facts the directive conferred upon Member States a discretion to choose a date prior to which the payment of outstanding claims which resulted from contracts of employment had to be honoured. The directive identified three alternatives and a further option (which itself depended upon the choice of date selected) to restrict liability to periods of three months or eight weeks respectively. None of these discretions and options however led the Court to the conclusion that the directive lacked unconditionality and sufficient precision:

“…the right of the State to choose among several possible means of achieving the result required by a Directive does not preclude the possibility for individuals of enforcing before the national courts rights whose content can be determined sufficiently precisely on the basis of the provisions of the Directive alone”.

(ibid, paragraph [17])

The Court held that the “content” of the obligation could be determined by taking the date whose choice entailed the least liability for the guarantee institution (ibid, paragraph [19]), and since this could be determined with finality it was hence unconditional and precise.

However, the directive lacked direct effect because there was uncertainty with regard to the identity of the person liable to provide the guarantee. Article 5 of the directive required Member States to lay down detailed rules for the organisation, financing and operation of the guarantee institutions. It also set out only broad principles which Member States had to observe when formulating the rules. This was sufficient, in the view of the Court, to render the provision as a whole lacking in direct effect:

“25.

Under Article 5, the Member State has a broad discretion with regard to the organisation, operation and financing of the guarantee institutions. The fact, referred to by the commission, that the Directive envisages as one possibility among others that such a system may be financed entirely by the public authorities cannot mean that the Member State can be identified as the person liable for unpaid claims. The payment obligation lies with the guarantee institutions, and it is only in exercising its power to organise the guarantee system that the State may provide that the guarantee institutions are to be financed entirely by the public authorities. In those circumstances the State takes on an obligation which in principle is not its own.

26.

Accordingly, even though the provisions of the Directive in question are sufficiently precise and unconditional as regards the determination of the persons entitled to the guarantee and as regards the content of that guarantee, those elements are not sufficient to enable individuals to rely on those provisions before the national courts. Those provisions do not identify the person liable to provide the guarantee, and the State cannot be considered liable on the sole ground that it has failed to take transposition measures within the prescribed period”.

In subsequent cases the Court of Justice has tended to hold that where broad options are conferred on Member States this is not normally conducive to direct effect. For instance in Case C-365/98 Brinkmann Tabakfabriken GmbH v Hauptzollamt Bielefeld [2000] ECR I-4640 a provision of a directive which imposed minimum excise duties for manufactured tobacco other than cigarettes was not directly effective because it gave Member States a choice between three different tax formulas: See paragraphs [38] and [39]. In Case C-157/02 Reiser Internationale Transporte GmbH v Autobahnen-und Schnellstraßen-Finanzierungs-AG [2004] ECR I-1477 a provision of a Directive required toll rates to be related to the costs of constructing, operating and developing the infrastructure network concerned, but did not explain the nature of the relationship, define the three heads of costs in question or define the concept of infrastructure network concerned. It was held that the provision did not provide any specific mode of calculation and left Member States a level of discretion which was incompatible with direct effect: See paragraphs [40] – [42].

For the reasons I set out below Article 5(2)(b) is not capable of exerting direct effect.

First, with regard to the identity of beneficiaries I accept that, as a class, persons who hold copyright are sufficiently identifiable. Upon the basis of the approach adopted by the Court in Francovich I conclude that this aspect of the test for direct effect is met.

Secondly, with regard to the “content” of the rights and obligations contained within Article 5(2)(b) I accept, in principle, that the mere fact that the content of any compensation scheme is subject to a series of discretions is not conclusive against direct effect. However, in Francovich it was not fatal to direct effect because the discretion and options conferred by the Directive did not prevent a fixed minimum rate of payment being determined by reference to the terms of the Directive itself. In the present case, however, there are no fixed minimums provided in the Directive and a discretion of a far broader type is conferred upon Member States to define the compensation scheme and set its parameters. The breadth of the discretion as to content is, in my view, inconsistent with the requirements of unconditionality and precision required for direct effect to exist.

Thirdly, with regard to the persons who are subject to the obligation to pay the Directive is, once again, unspecific. It is open, for instance, to Member States to pay compensation themselves. However, it is also open to them to impose the payment obligation upon third parties such as device manufacturers or even, through a sales levy, upon consumers. Recital 35 leaves it to Member States to determine the “form, detailed arrangements and possible level” for fair compensation. The recital also makes clear that in deciding upon compensation Member States must take account of “the particular circumstances of each case” and are responsible for “evaluating these circumstances”. Recital 35 thus confers on Member States a role in factual appreciation which is insufficiently precise to have direct effect. In Copydan at paragraph [20] the Court stated, in terms which are inconsistent with direct effect:

“Since [the Copyright Directive] does not provide any further details concerning the various elements of the fair compensation system, the Member States enjoy broad discretion in that regard. It is for the Member State to determine, inter alia, who must pay that compensation and to establish the form, detailed arrangements for collection and level of compensation”.

Similarly, in Case C-521/11 Amazon.com (11^th July 2013) at paragraph [20], in relation to the person who has to pay, the Court confirmed that “…since the provisions of Directive 2001/29 do not expressly address the issue of who is to pay that compensation, the Member States enjoy broad discretion when determining who must discharge that obligation”. In short, there is no unconditionality or precision about the identity of the person subject to the obligation. As in Francovich I conclude that this is antithetical to a conclusion that direct effect arises.

Even if I had concluded that Article 5(2)(b) had direct effect that does not mean that adjudication at the national level is always or inevitably a “hard edged” merits issue. In principle, direct effect is primarily concerned with creating a locus for an affected person to seek private law relief from the Courts (see paragraph [152] above). Typically, this would be an individual rightholder seeking a financial or other remedy from an infringer in civil proceedings. But, in the present case, the Claimants and the Intervener are all representative bodies who have been accorded locus before the High Court in public law judicial review proceedings and they are entitled to seek relief from amongst the wide range of remedies available to the Court if they prevail. As such, recourse to direct effect is otiose as a route to locus. So far as it is contended that if Article 5(2)(b) is directly effective this entitles the Claimants to a hard-edged merits review, this is a separate issue. There is agreement between the parties that if the Claimants prevail they are entitled to remedies which are effective, proportionate and dissuasive (see paragraph [39] above; Recital 58 and Article 8 of the Directive). The actual process by which this Court analyses the issue and selects proper remedies is one however which is prima facie determined by domestic law (the principle of the “procedural autonomy” of national law) subject to the well known principles of “effet utile” and non-discrimination. In many instances it will be appropriate for a national court to apply a merits review to a directly effective right. But in my view there may be circumstances where something less intense may apply yet still amount to an effective remedy. This might, for instance, arise where a party seeks to enforce direct effect via public law proceedings in a case where some degree of discretion remains vested in the Member State (see, for example, paragraph [284] below).

In my view, whether the provision has direct effect or not does not in this particular case alter my conclusion about the intensity of the review.

The next legal issue arising concerns the law governing consultations and the appraisal by decision makers of evidence. This latter aspect (appraisal) includes: (a) the duty not to predetermine the evidence and (b) the duty to consider the evidence and arrive at rational conclusions from it. This area of law is relevant to the Claimant’s submissions as to the adequacy of the consultation process, to its complaint about predetermination of the result and to its complaint that the Secretary of State simply did not draw proper or rational conclusions from the evidence. There are also indications set out in the case law which go to the intensity of the review process.

In some cases, the duty to consult is generated by statute. However, and not infrequently, it is generated by the duty imposed by the common law upon a public authority to act fairly: See for example per Lord Wilson JSC in R(Stirling) v Haringey LBC [2014] UKSC 56 at paragraph [23] (hereafter “R(Stirling)”). In R(Osborn) v Parole Board [2014] AC 1115, Lord Reed JSC (at paragraphs [67] and [68]) explained that the purpose behind the requirement that a consultation should be fair was that it was liable to result in better decisions because the decision maker received all relevant information which was properly tested; and because it avoided the sense of injustice which a person who was subject to the decision might otherwise feel. In the present case the Secretary of State was not obliged under the Directive to consult. However, his chosen end result of a copyright exception absent compensation was by its very nature bound to be highly controversial and risked, if he got the decision wrong, denying many persons and companies rightful compensation. It is hence easy to see why fairness compelled that he should conduct a detailed consultation.

As to the principles to be applied in R(Stirling) (ibid) in 2014 the Supreme Court decided that the time had come to endorse the so-called “Sedley” requirements for a fair consultation process. These were:

“First, that consultation must be at a time when proposals are still at a formative stage. Second, that the proposer must give sufficient reasons for any proposal to permit of intelligent consideration and response. Third…that adequate time must be given for consideration and response and, finally, fourth, that the product of consultation must be conscientiously taken into account in finalising any statutory proposals”.

(paragraph [25] citing with approval per Hodgson J in R v Brent London Borough Council ex p Gunning (1985) 84 LGR 168 at page 189 citing in turn and with approval the four principles advanced by way of submission in that case by Mr Stephen Sedley QC, as he then was).

In R(Stirling) Lord Wilson also endorsed the observation (which has resonance in the present case) about the intensity of review in cases involving the deprivation of rights of Simon Brown LJ in ex p Baker [1995] 1 ALL ER 73, at page 91:

“The demands of fairness are likely to be somewhat higher when an authority contemplates depriving someone of an existing benefit or advantage than when the claimant is a bare applicant for a future benefit”.

The fourth “Sedley” principle is that the product of consultation must conscientiously be taken into account by the decision maker. This reflects two broader principles (which apply also outside the context of consultations). First, a decision must be based upon a reasonable or rational view of the evidence it is said to be based upon. This is a principle of long standing: see e.g. R v Deputy Industrial Injuries Commissioner ex p. Moore [1965] 1 QB 456 at page 488 (CA) per Diplock LJ; Mahon v Air New Zealand Ltd [1984] 1 AC 808 (PC) pages 820, 821; see also the commentary in “Judicial Review – Principles and Procedure” (Auburn, Moffett, and Sharland) (2013) paragraphs 20.09ff pages 454, 455 (“Auburn”). If the final decision adopted bears no sensible relationship to the evidence relied upon then the reasoning is unreasonable or irrational but, in a consultation context, it may be said also that the “product” of the consultation (the evidence) has not “conscientiously” been “taken into account”. The second broader principle is that the outcome of the consultation must not be predetermined. If it is then the decision maker will not have acted “fairly” (the leitmotif of the principle governing consultations) and will not have “conscientiously” taken into account the evidence. The law focuses upon actual predetermination but also the appearance of predetermination though commentators point out that the issue of the appearance of predetermination is very much harder to establish in the case of administrative or “political” decisions than in the case of judicial decisions since the former involve decision makers who are entitled to have strong predispositions whereas Judges should not (or at least should keep them firmly to themselves): See generally Auburn (ibid) paragraphs [8.84ff].

Issue II: The meaning of “harm”

Issue II raises an important and threshold point of law. It focuses upon the concept of “harm” which underpins the duty to compensate in Article 5(2)(b) and the various objectives and principles set out in the recitals including the concept of a “fair balance of rights and interests” in Recital 31 and the concepts of “fair compensation”, “possible harm to the rightholders” and “prejudice” in Recital 35. In the text below, I address this issue in the following order. First, I set out the rival contentions of the parties. Secondly, I address a significant concession made by the Defendant. Thirdly, I deal with a related dispute between the parties about an error of law alleged to have been made by the Defendant. Fourthly, I deal with a particular point of construction of the Directive upon which the parties were agreed. Fifthly, I address what is meant in law by “harm”.

The Secretary of State submits that the “harm” which he took into account for the purpose of his conclusion that the introduction of section 28B would not trigger a requirement for compensation was based upon the expectation that costs to rightholders due to lost sales would be minimal or zero. In arriving at this conclusion the Defendant took as a starting point that the endemic copying performed by users, in the United Kingdom, did not, to any material degree, thwart duplicate sales which might otherwise have been made by the purchaser had copyright law been rigorously enforced. The metric which the Defendant has used in order to determine “harm” was thus lost sales.

The Claimants adopted a very different stance. They argued that the measure of their loss or “harm” was the total licence fee that might have been charged upon the hypothesis that every violation of copyright deprived the rightholder of some value. The Claimants accordingly computed “harm” upon the basis of a hypothetical licence fee which would be charged to the user in a counterfactual market where enforcement was all pervasive.

The Claimants adduced in support of this analysis a report by Compass Lexecon/FTI Consulting which concluded that, in a world of viable enforcement, consumers would be willing to pay in the order of £9 more for a CD which would permit them to make unlimited licensed copies than for a CD for which they were permitted to make no copies. The experts translated this into a £2.4 billion loss in revenue for the period 2006-2012. The experts also concluded that consumers would be willing to pay £0.72 more to make one licensed copy of a CD. When the total number of CDs sold between 2006-2012, together with the likely number of digital copies that consumers would be expected to make of those CDs, was taken into account the experts assessed the potential value that consumers would be prepared to pay for such copying as between £100 million and £624 million. The Defendant criticised this analysis as being exaggerated, unworldly and wholly unreflective of the real market. I shall refer to the Defendant’s test as the “lost sales” test; and to the Claimant’s test as the “licensing test”.

Differences in the conceptualisation of “harm” under the Directive might thus lead to very different outcomes in this case. It is unnecessary for me to decide whether the Claimants’ calculations are exaggerated or not in order to conclude that the difference between the “lost sales” and “licensing” tests for harm are material in economic and legal terms.

Mr Saini QC accepted, on behalf of the Secretary of State, that were he to be wrong in his submission that “harm” under the Directive encompassed lost sales only then the decision which led to the introduction of section 28B was predicated upon an error of law. He accepted that, in those circumstances, the conclusion in the Updated Impact Assessment that the loss to rightholders was “minimal or zero” would be based upon an erroneous understanding of the law.

The correct meaning of “harm” is thus a threshold issue which is fundamental to this litigation. In the analysis below I have reached the conclusion that the Secretary of State acted lawfully in adopting the “lost sales” test. However, if I am wrong in this then the entire consultation was, as is now conceded, based upon a flawed legal premise and would be unlawful. I have also concluded (section I below) that the actual reasoning which led to the new copyright exception was flawed for reasons unrelated to the meaning of “harm”. However since that particular flaw could, at least in principle, be cured by the further collection of adequate evidence, it is conceivable that the present copyright exception could be maintained. However, if I am wrong in my conclusion about relevant “harm” no amount of repair work to the consultation and decision making process could save it from being quashed and set aside in its entirety upon the basis that it was permeated through by an error of law. This is relevant to the question of relief that I have posed in the Conclusion section of this Judgment at section K below.

Before addressing the law on “harm” I briefly set out my conclusion upon a submission of the Claimants that there was a second legal error inherent in the consultation process, which they called the false “incrementalist approach”. The Claimants submitted that it was evident from a reading of the documents leading up to the decisional Updated Impact Assessment and also from the pleadings, submissions and witness statement evidence submitted in this litigation by the Secretary of State that he had taken the view that the harm to be made subject to compensation was only that which flowed out of such new and incremental copying as would arise out of the new copyright exception. In other words the Defendant had, in effect, written-off as legally irrelevant, all that vast body of illegal copying which had historically occurred. It was for this reason that the Claimants thus labelled the Defendant’s reasoning the “incrementalist approach”. They used this phrase in contra-distinction to what they submitted should have been the correct legal principle which they called the “legitimisation approach” which was harm which flowed from all that copying which was now legitimised but which hitherto had been illegal. I was taken through numerous sentences and phrases, dotted around and amongst the documents, in which it was said the flawed “incrementalist approach” was evident.

I do not accept that this is a justifiable criticism of the Defendant for two reasons.

First, there is indeed a dispute between the Defendant and the Claimants about duplicate or incremental lost sales, and the licensing approach. This is a real and important issue and it is encapsulated in the disagreement about the meaning of “harm”. In my view when the Defendant was, throughout the documents, referring to incremental sales he was referring to this debate. I do not accept that, fairly read, the Secretary of State was purporting to say that in the computation of harm he simply wrote off as irrelevant all historical, unlawful copying. His point was quite different - it was that as a matter of fact not every act of past unlawful copying led to actual loss or harm. He based this upon economic literature which drew the distinction between a consumer’s willingness to engage in unlawful copying and that same consumer’s preparedness to go out and buy a duplicate (incremental) copy. I have referred to this evidence in Section D above; see e.g. at paragraph [57].

Secondly, as to the distinction drawn by the Claimants between the incrementalist approach” and the “legitimisation approach” the Secretary of State accepted without qualification that it was the latter which was the correct approach. He thus agreed with the Claimants. He simply contended that in applying the legitimisation approach it was necessary to go on and assess the extent to which the now legitimised acts of copying would lead to lost sales.

For these reasons, I reject the Claimants’ criticism on this ground.

Before turning finally to the merits of the point of law it is necessary to record one further matter concerning the interpretation of the Directive which is common ground between the parties. On one possible construction of the Directive the obligation to provide for a compensation mechanism is mandatory whenever a Member State exercises Article 5(2)(b).

In the present case, the Defendant has however proceeded throughout the entire consultation exercise upon the premise that he would only introduce a limited private use exception if it would not be accompanied by a compensation scheme, which he took fundamental objection to. Since the Defendant came to the view that a private use exception could be introduced which generated minimal or zero harm to rightholders the Secretary of State therefore saw no need to introduce any form of compensation or levy scheme.

But if the interpretation of the Directive that I have just set out is correct, the Defendant faced a dilemma. He was required to institute a compensation or levy scheme even though when it was applied it always resulted in zero compensation payments being made. This would be an exercise in the utmost futility.

Recognising the logic of the Defendant’s position the Claimants have conceded that if the Defendant is correct and that the introduction of the limited exception would cause no or no material harm to rightholders, then there is no requirement in law to go through the bureaucratic rigmarole of instituting a compensation mechanism which would, by definition, lead to no payments out ever being made. The Claimants thus accept that assuming the Defendant’s assumption as to “no harm” is correct he may lawfully introduce the exception without the need for the concomitant introduction of a compensation or levy scheme.

I agree entirely with the concession made by the Claimants. It would not, in my view, be a proportionate interpretation of the Directive to lead to the absurd conclusion that a compensation or levy scheme was required even though it would lead nowhere.

However, this has, in my view, an additional implication. It means that the facts and matters identified as criteria in Recital 35 (on the operation of compensation) can be applicable in two different scenarios. First, in relation to the modus operandi of an actual levy or compensation scheme introduced by a Member State. But also, secondly, in the context of a copyright exception regime which does not introduce any system of compensation. The criteria in Recital 35 will work differently in each scenario. In the case of an actual scheme it is likely that the facts and matters in Recital 35 will be built into the detailed scheme rules, e.g. where the thresholds for de minimis apply, who pays, etc. But in the second scenario these facts and matters will be taken into account in a quite different way and in particular in the process of reasoning which leads up to the decision that no compensation scheme is needed.

I turn now to the heart of the matter, the meaning of “harm”. In my view, the analysis of the Defendant is to be preferred. In the text below I set out seven reasons why I prefer the conclusion of the Secretary of State upon this point. I have added a caveat to this at paragraph [206] below.

First, a growing body of consistent judgments of the Court of Justice establishes that the concept of “harm” is an “autonomous” concept of EU law which is unrelated to national law. A central point raised by the Claimants is that in domestic (common) law enforcement proceedings one basis for the calculation of damages is the fees that would be obtained in a hypothetical licence. Thus “harm” should be understood to take account of every act of copying undertaken by an infringer (the “licensing test”), and whether or not the consumer would have made a duplicate sale is therefore not the correct optic through which to view the right to compensation. The difficulty with this argument is that in classifying the concept of “harm” as an “autonomous” concept of EU law the Court has rejected the notion that harm should be defined by reference to the sorts of remedies that would be awarded in national courts. Indeed, conspicuous by its absence from Recitals 31 and 35 is any reference to the concept of being harm being defined by reference to national remedies. It would have been straightforward had “harm” been intended to have been defined by reference to remedies in domestic civil law for that to have been spelled out in the recitals, but it was not.

The above conclusion is consistent with the judgment of the Court of Justice in Padawan paragraphs [31]-[33]:

“31.

It should be noted at the outset that neither Article 5(2)(b) nor any other provision of Directive 2001/29 refers to the national law of the Member States as regards the concept of ‘fair compensation’.

32.

In such circumstances, according to settled case-law, the need for a uniform application of European Union law and the principle of equality require that the terms of a provision of European Union law which makes no express reference to the law of the Member States for the purpose of determining its meaning and scope must normally be given an independent and uniform interpretation throughout the European Union; that interpretation must take into account the context of the provision and the objective of the relevant legislation (see, in particular, Case 327/82 Ekro [1984] ECR 107, paragraph 11; Case C-287/98 Linster [2000] ECR I-6917, paragraph 43; and Case C-523/07 A [2009] ECR I-2805, paragraph 34).

33.

It is clear from that case-law that the concept of ‘fair compensation’ which appears in a provision of a directive which does not contain any reference to national laws must be regarded as an autonomous concept of European Union law and interpreted uniformly throughout the European Union (see, by analogy, as regards the concept of ‘equitable remuneration’ in Article 8(2) of Council Directive 92/100/EEC of 19 November 1992 on rental right and lending right and on certain rights related to copyright in the field of intellectual property (OJ 1992 L 346, p. 61) and Case C-245/00 SENA [2003] ECR I-1251, paragraph 24)”.

The same points were made by Advocate General P Cruz Villalon in Case C-435/12 ACI Adam (10^th April 2014) at paragraphs [57] – [58]. Indeed, the Advocate General pointed out that far from being defined by national law the concept was more influenced by international law, such as the Berne Convention: ibid, paragraphs [59] – [60].

Secondly, the Claimants’ argument does not sit well or easily with the very unusual and particular circumstances which have led to the decision to introduce the private use exception in the first place. These are that the advent of digitalisation has led to a market where device sellers and consumers assume they may copy and where rightholders have not sought private law remedies against infringers. It is a particular feature of this case that there is a widespread consensus that the law has signally failed to keep up with market reality and with reasonable consumer expectations and indeed has been brought into disrepute by its condemnation as illegal of activities which are now accepted by consumers as lawful and which in actual fact form the basic commercial premise upon which copying and storage devices are actively sold throughout Europe.

This view is not just a view expressed by those with an economic interest in seeing the exception introduced (device manufacturers and consumers); it is also a view widely acknowledged by content suppliers. The reality (at least in the UK, but it would seem elsewhere in the EU as well) is that copyright holders do not use the courts to obtain relief against the widespread infringing activities of consumers. Civil redress is not seen as a rational or logical response by the industry to market developments. A compensation scheme based on judicial remedies would thus fail to reflect how the world has changed with digitalisation. This supports the argument, advanced by the Defendant, that it is not relevant to look to national law remedies as a determinant of the meaning of “harm”. As a means of rewarding and protecting content creators civil redress has in this particular context failed. As such there is no reason why “harm” should be defined and determined by the entirely theoretical prospect of obtaining judicial compensation based upon lost licence fees. In my view “harm” should be understood in a wider context which takes account of actual consumer expectations and the very specific evolved features of the markets which are covered by the exception. The Directive (and indeed the construction placed upon the Directive by the Court) makes clear that compensation is to be assessed for loss or prejudice or harm “caused” by the copying activity in question. The Defendant in his assessment of causality has applied a real world counterfactual, i.e. one in which civil claims are not pursued against bona fide purchasers of content who engage in subsequent copying. I have explained elsewhere in this judgment why I consider this to be a rational decision upon the evidence: See section H below. Against this counterfactual the loss sustained is not that which could hypothetically be obtained from a court but is measured in terms of lost duplicate sales. This conclusion is, in my view, consistent with the clear recognition in Recitals 31, 35 and 38 that the exceptions to copyright and the related compensation provisions are closely connected to actual, real-world, developments in the digitalisation of markets.

Thirdly, as was pointed out by Advocate General Trstenjak in Padawan the Directive represents and reflects a compromise which sought to balance the declared objective of harmonisation with “…the differing legal traditions and views in the Member States of the European Union – including in particular the common law concept of copyright and the continental European concept of copyright protection”. The Advocate General was of the view that the significant number of discretions conferred upon the Member States to create exceptions to copyright meant that Member States enjoyed “considerable flexibility in the transposition of the Directive” (ibid paragraph [43]). The Advocate General also referred (in footnote [12]) to literature which analysed the Directive as an attempt to reconcile continental and civil law concepts of copyright protection with common law concepts. The Advocate General described this as “the greatest challenge” of the Directive. This highlights the conclusion that the concept of “harm” is not one with its roots in national law; it is a blended or hybrid concept drawing inspiration from civil and common law traditions and, as such, should not be defined by reference to remedies available in specific Member States, such as an account of profits or a right to licence fees, etc.

Fourthly, all of the above points support the conclusion that since “harm” is not a defined notion it is one about which Member States have a discretion, albeit one supervised by the Court of Justice. This is a key point because it indicates that there may be more than one test for “harm”, and that the real question is whether adoption of a “lost sales” test is within the discretion of the Member State. In Case C-463/12 Copydan (5^th March 2015) (“Copydan”) the Court of Justice did not itself define harm but instead ruled that as a matter of EU law a discretion was conferred upon Member States to define the issue: See ibid, paragraph [62]. The Claimants submitted that there was an inconsistency in the Court on the one hand ruling that the concept of “harm” was an autonomous concept of EU law but then remitting it to the discretion of the Member States. I do not agree. There is no such inconsistently. For the Court to remit the issue to the discretion of the Member States is not the same as remitting the issue to national law; and it does not in any event mean that Member States have carte blanche to introduce (or not) any scheme that they please. Member States thus do not have to define harm by reference to existing national law or judicial remedies and they may choose some other metric. But the discretion remains a fettered one. For example Member States could not refuse to implement a compensation scheme where the harm was more than “minimal”; and equally they may not choose a solution that offends against general principles of EU law such as the principle of non-discrimination or proportionality (see in this regard Copydan paragraph [62] cited below); and further (as case law demonstrates) because the issue of compensation is an autonomous EU law issue the Court of Justice retains jurisdiction over a range of technical questions such as what sorts of acts attract the compensation duty, and who pays, etc.

All of this is in my view clear from (inter alia) Copydan (ibid) where the Court was asked, upon a reference from the Ostre Landsret (Denmark), to interpret Article 5(2)(b) and Recital 35 which stipulates that in certain circumstances where the prejudice to the rightholder would be minimal no obligation for payment arose. The Court concluded as follows. First, it noted (ibid paragraph [57]) that the object of the Directive was to harmonise “only certain aspects of the law on copyright and related rights” and that the general scheme of the Directive was circumscribed by a number of provisions which disclosed an intention on the part of the EU legislature to grant “a degree of discretion to the Member States in the implementation” of the Directive. Secondly, the Court stated that Article 5(2)(b) incorporated two different discretions:

“In the same way that Member States may elect whether or not to adopt any of the exceptions set out in Article 5(2)…they also have the option, as confirmed by recital 35 in the preamble to the Directive, to provide – in certain cases covered by the exceptions which they have freely established – for an exemption from payment of fair compensation where the prejudice caused to rights-holders is minimal”.

The Court then (ibid, paragraph [61] - [62]) said this in relation to the discretion of Member States in this regard:

“61.

For the reasons given in the preceding paragraphs, the setting of a threshold below which the prejudice may be classified as “minimal”, for the purpose of recital 35 in the preamble to [the Directive], must also be within the discretion of Member States, provided, inter alia, that the application of the threshold is consistent with the principle of equal treatment, as referred to in paragraph 31 above.

62.

In the light of the foregoing, the answer to Question 3 is that Article 5(2)(b) of [the Directive], read in the light of recital 35 in the preamble to that Directive, must be interpreted as permitting Member States to provide, in certain cases covered by the private copying exception, for an exemption from payment of fair compensation, provided the prejudice caused to rights-holders in such cases is minimal. It is within the discretion of the Member States to set the threshold for such prejudice, it being understood that that threshold must, inter alia, be applied in a manner consistent with the principle of equal treatment”.

The quotation set out above makes clear that Member States have a discretion with regard to “the setting of a threshold below which the prejudice may be classified as “minimal”; and, “…the threshold for such prejudice”.

As to the breadth of the discretion accorded to Member States, in Padawan Advocate General V Trstenjak, at paragraph [82], referred to Member States having “…a wide discretion…as to how their respective national systems implement such fair compensation” citing academic literature to support this view (cf. footnote 61). In Stichting (ibid) the Court pointed out that where the Directive was silent as to a matter (such as who paid compensation) then Member States had a “broad discretion” (paragraph [23]). It appears that the Court took the same view of “…the notion and level of fair compensation” (ibid, paragraph [24]). In Copydan Advocate General Cruz Villalon stated that Member States creating copyright exceptions “…have the greatest latitude to adopt provisions derogating from the charging of fair compensation where the harm is minimal” (ibid paragraph [99]); and the Court (ibid paragraph [62]) referred to Member States having “…the discretion…to set the threshold for such prejudice”. There are thus indications that the discretion is broad. In other cases however the Court itself has been more circumspect. It has recognised that a broad discretion might undermine consistency and hinder the internal market. In ACI Adam (ibid) at paragraphs [21] – [24] the Court, when referring to the “scope of those exceptions and limitations” (i.e. in Article 5(2)) emphasised that they were derogations from “a general principle established” by the Directive (i.e. in Article 2) and that accordingly they “…must be interpreted strictly” and not in a manner conflicting “with a normal exploitation” of the work or in a manner which would “unreasonably prejudice the legitimate interests of the rightholder”. Viewed overall, Member States do have a discretion with regard to the nature, extent and modalities of the exception and the concomitant compensation right; but it is one which needs to be exercised carefully and with proper regard to the rightholders’ interests. In my judgment, unless I can say that the choice of a lost sales test for computing harm is outwith the scope of the Member States’ discretion I should not interfere with it. And for the reasons that I have already given I consider it to be a perfectly rational option for the Defendant to adopt.

In my view, it follows from the above that the concept of “harm” is a matter governed by EU not national law; that it is a concept which seeks to meld disparate common law and civil traditions; and it leaves the Member States a discretion as to the threshold below which “harm” is “minimal” and, a fortiori, above which “harm” must be compensated for.

The fifth point flows from the considerable importance attached to Recital 35 by the Court of Justice as a guide to the construction of Article 5(2)(b). By way of illustration only, the substantive text of Article 5(2)(b) contains no express de minimis exception; that flows exclusively from the recital. Yet the Court of Justice has made clear that a de minimis exception does arise in Article 5(2)(b) despite its express omission from the terms thereof. It is therefore important to see what additional guidance is available from the recitals and in particular Recital 35. This identifies a series of considerations which are matters which Member States may take into account when determining the nature of the scope of the de minimis exception. Mr Saini QC, for the Secretary of State, dissected Recital 35 into its constituent parts. He identified the following as relevant considerations which, he submitted, the Defendant was entitled to take account of in determining whether or not the posited copyright exception would cause “harm” (or not) to rightholders:

The sixth point relates to the manner in which the Court of Justice has, recently, defined the specific subject matter of copyright. In Joined Cases C-403/08 and C-429/08 Football Association Premier League Limited & Others v QC Leisure & Others; Murphy v Media Protection Services Limited (4^th October 2011) the Court of Justice was required, inter alia, to consider the circumstances in which derogations from the principle of free movement could be permitted upon the basis that they were justified for the purpose of safeguarding rights which constituted the specific subject matter of copyright. The Court acknowledged that the specific subject matter of copyright included ensuring for the rightholder protection of the right to exploit commercially the marketing or the making available of the protected work by the grant of licences in return for payment of remuneration (ibid, paragraph [107]). However, the Court then sought to impose a limit on the ability of the rightholder to maximise revenue by reference to a benchmark of “reasonable remuneration”:

“108.

However, the specific subject-matter of the intellectual property does not guarantee the rightholders concerned the opportunity to demand the highest possible remuneration. Consistently within the specific subject-matter, they are ensured – as Recital 10 in the preamble to the Copyright Directive and Recital 5 in the preamble to the Related Rights Directive envisage – only appropriate remuneration for each use of the protected subject-matter.

109.

In order to be appropriate, such remuneration must be reasonable in relation to the economic value of the service provided. In particular it must be reasonable in relation to the actual or potential number of persons who enjoy or wish to enjoy the service…”.

The Defendant submitted that the Court did not view an ability to maximise revenue as a necessary constituent of the specific subject-matter of copyright. Support for this conclusion is found in Article 5(5) of the Directive which links the exceptions to the “normal exploitation” of the work and where the legitimate interests of the rightholder are not “unreasonably” prejudiced. These considerations support an understanding of Article 5(2)(b) through the optic of the law on the specific subject matter of copyright. The judgment acknowledges that a right to maximise revenue cannot, without more, be assumed to be part of the specific subject-matter of copyright and introduces a species of reasonableness test which, in the context of the present case, lends some support to the conclusion that Member States have a margin of discretion as to the choice of relevant thresholds for “harm” and need not choose a metric which maximises rightholder value.

Finally, there has been placed before the Court a variety of academic texts and reports prepared by consultants in which differing views on this issue are set out. For instance the Secretary of State drew the courts attention to the report by CRA “Assessing the economic impacts of adapting certain limitations and exceptions to copyright and related rights in the EU - analysis of specific policy options” (May 2014) in which the authors endorsed the pricing-in methodology and, of particular relevance to the present issue, the computation of harm based upon lost sales. This was a report prepared on behalf of the European Commission (DG Markt). For instance at page 5 the authors state:

“Private copying relates to the reproduction of creative content for use in the private sphere. An exception is already in place for private copying in most member states. Rightholders are often financially compensated for this exception when it is assumed that private copying implies a loss of revenues for them. We however argue that harm does not necessarily arise as a result of private copying. In particular, there is no market failure if the price of original content already reflects the value attached by consumers to copying or if there is no substitution between original content and copies because of the valuation of consumers and the difference in cost of production”.

(emphasis added)

Elsewhere in this detailed report the consultants explain why levy and compensation systems themselves can be market distorting and are often, from an economic perspective, unsatisfactory. However, the basic point here is that they take the view that harm only arises in the case of lost duplicated sales, i.e. they support the Defendant’s case.

The Claimants drew my attention to the important report by Mr Antonio Vitorini, a former Commissioner for Justice and Home Affairs, who was tasked by the then Commissioner to consider solutions to the differences which existed across Europe in the manner in which rightholders and users paid levies: cf. “Recommendations resulting for the Mediation on Private Copying and Reprography Levies” (31 January 2013). In his report he described this issue as a “source of friction” and “important and sensitive” (ibid. page 1). From pages 19ff Mr Vitorini examines “the notion of harm” and he sets out the reasons why a licensing test for harm is appropriate. His conclusions – which are endorsed by the Claimants – are as follows. It is worth setting them out fully:

“In my opinion, to determine the 'harm' sustained by rightholders because of the copies made within the scope of the private copying and the reprography exceptions one needs to look at the situation which would have occurred had the exception not been in place. In particular, one needs to assess the value that consumers attach to the additional copies of lawfully acquired content that they make for their personal use. It would allow the estimate of losses incurred by rightholders due to lost licensing opportunities ('economic harm'), i.e. the additional payment they would have received for these additional copies if there were no exception. I think it is fair to say that, in the majority of cases, this amount would neither reach the level comparable to the value of the initial copy nor would it be so negligible that it could be completely ignored.

Although such a definition of 'harm' is, at its most basic level, not contested by the majority of stakeholders, it seems that views on its practical implications diverge. Some stakeholders are of the opinion that once 'harm' is defined uniformly in the above manner, it would still simply refer to the number of copies made by consumers under the private copying exception, and all of them would need to be fully be taken into account when setting the levies. For others, the definition of 'harm' outlined above implies that the customers' willingness to pay for additional copies decreases proportionally to their number and that, therefore, the hypothetical licence-fee(s) which the rightholder would have obtained - had there been no exception - decreases with each additional copy made. Such an approach also seems to imply that, in many instances, the value of each additional copy could be so small that the 'harm' sustained by the rightholders should be considered as minimal, with no compensation obligation arising. Accordingly, it would be necessary to assess not the actual number of copies made but rather the hypothetical (lower) number of copies that could have been licensed in the absence of the exception.

It seems to me that, since the main rationale underlying the private copying exception is linked to the practical difficulty of the licensing of copies made by consumers for their private use, it is fair and reasonable to compensate rightholders precisely for lost income opportunities, e.g. via the licence agreements they would have concluded if there were no exception. Moreover, in my view it would be justified for the level of compensation to reflect the actual value attached by consumers to such additional 'private' copies. The latter also depends on the form in which the copyright protected content is copied (analogue/digital). For instance, the value a consumer attaches to the fact that he can copy a book (instead of acquiring a subsequent copy of it) is different from the one he attributes to the possibility of copying a CD on multiple devices being in his possession. In any event, I am far from saying that these additional copies would cause minimal 'harm' and would not need to be compensated. But I equally think that measuring all the copies actually made by virtue of the exception without taking into account the consumers' willingness to pay for these copies if there were no exception could lead to compensating to a greater degree than EU law actually requires.

Finally I am aware that the value attributed to each private copy by a single consumer is rather small, and hence the resulting 'harm' may be minimal, but I also believe that the CJEU made clear in the judgment in Case C – 467/08 (Padawan v SGAE) that the 'harm' which is subject to fair compensation arises not from one single copy but from a number of a natural persons' activities which, taken together, amount to relevant 'harm' caused to rightholders.

RECOMMENDATION:

 Ensure more coherence with regard to the process of setting levies by defining 'harm' uniformly as the value consumers attach to additional copies in question (lost profit)”.

With great respect I am not entirely convinced by this. The report misses the salient point that the Court (in Padawan, inter alia) has concluded that the exact parameters of “harm” are to be remitted to the discretion of the Member States who can choose an appropriate method for calculating harm. That might, as Mr Vitorini postulates, be based upon lost licence fee income but it might equally, depending upon the particular circumstances prevailing in a Member State, be defined by lost sales. I am also not convinced that the “main rationale” behind Article 5(2)(b) is linked to “practical difficulties” in licensing. The recitals to the Directive paint a far broader picture than this. Nonetheless, the views expressed in this report cannot be ignored; they are serious and Mr Vitorini raises important policy questions and considerations.

All of this leads me to the conclusion that the Defendant acted lawfully in adopting the more limited “lost sales” concept of “harm” as opposed to the broader “licensing test” advanced by the Claimants. In my view there was a proper basis in economic and legal logic for choosing to adopt this basis in the United Kingdom. The decision taken by the Secretary of State was a rational one which is consistent with EU law. My conclusion is therefore not that the licensing test is wrong in principle; but only that the issue properly falls within the discretion of the Member States and the choice of the sales test was a lawful choice.

Nonetheless, this is an issue of great significance. I accept that if I am wrong and Member States have no discretion and must adopt the licensing test then a conclusion based upon the Claimants’ analysis would go to the heart of the legality of the decision to introduce section 28B without a compensation scheme. And I also accept that the Claimants’ contrary arguments are neither trivial nor can be dismissed so easily that I can reject them as acte claire wrong.

I will consider the implications of this when I consider remedies.

Issue III: The alleged irrationality and/or inapplicability of the pricing-in principle

The Claimants submitted that the Secretary of State’s reliance upon the pricing-in theory was flawed because pricing-in could not obviate harm where the seller had no ability to price discriminate, and such an ability did not exist on the facts of the present case. In the 2012 Assessment and in the Updated Impact Assessment the Defendant relied upon early academic literature by Professor Stan Liebowitz for the proposition that where pricing-in occurred there might be no loss from subsequent unauthorised copying. However, in the course of the litigation the Claimants relied upon a newly commissioned expert report from Professor Liebowitz (“Report regarding Indirect Appropriability”, November 2014) in which he qualified and explained his earlier analysis. Professor Liebowitz explained that whilst pricing-in might have some application in this market sector it would never enable all the relevant “harm” to be captured:

“4.1

Whilst the theory of “pricing-in / indirect appropriability may be applicable in the context of the PCE, it does not provide a basis for concluding that rightsholders will be fully or even significantly compensated for the harm that they will suffer arising from private copying as legitimised by the PCE”.

Pricing-in was effective to capture all harm “only” if either (a) all consumers received a similar value for the content or (b) if sellers could price discriminate:

“7.7

Indirect appropriability is unlikely to compensate rightsholders for all of the lost sales arising from private copying and possibly may not compensate them for any of them. When I have previously written about indirect appropriability, I have always pointed out that special conditions are needed for indirect appropriability to work, and even when it works it might only partially compensate the rightholders. I have never claimed that indirect appropriability will always fully compensate rights owners for copying.

7.8

In particular, indirect appropriability can work effectively to compensate rights holder only if all consumers derive similar values from their copying or if sellers can adjust the price to each copier based on the value of their copying”.

(emphasis added)

In the present case consumers are not similar and are heterogeneous and therefore the “only” way for a seller to capture full value is to be able to engage in price discrimination between consumers:

“7.14

The only way for indirect appropriability to work well when there is volatility in consumers, is for the seller to charge higher prices for authorised units which will have multiple copies made than for those units sold to individuals making zero copies. Charging different prices to consumers, based upon their willingness to pay, is known as “price discrimination”. Without price discrimination, indirect appropriability cannot be counted on to work when consumers are heterogeneous, which in fact they are”.

Professor Liebowitz’s ultimate conclusion was that by using a (flawed) model based upon pricing-in the Defendant had ensured that “…private copying is virtually assured of harming rightholders relative to a case where no private copying occurred”.

In addition the Claimants relied upon the Compass Lexecon / FTI Report which set out the academic literature which underpinned the pricing-in theory, including that of Professor Liebowitz. The report pointed out that consumers attached very different values to products and consequently had a very different willingness to pay (or “WTP”). The Report then stated:

“2.24

It is therefore clear that the extent to which indirect appropriability is effective depends on the circumstances. Different consumers tend to place different values upon the incremental benefit of the product and indirect appropriability is not fully effective when it is not possible to price discriminate between consumers according to their WTP. The different values placed by consumers upon the incremental benefit could be a result of the desire to make different numbers of copies, for example…”.

The thrust of the complaint was that across the music, film and books market content suppliers had no real ability to price discriminate and hence the Defendant had erred in relying upon a theory of pricing-in that was economically inapplicable. It was an error to rely upon the early theories of Professor Liebowitz without verifying (with him) that he considered that this theory could apply in the relevant sectors; and moreover, in any event it should have been plain from that early literature that the context in which Professor Liebowitz was considering the pricing-in theory (or “indirect appropriability”, as Professor Liebowitz described it) was inapplicable to the music, film and book markets covered by section 28B.

A critical cog in this analysis is therefore that without an ability to price discriminate there is no way for a seller to capture full value through pricing-in. Professor Liebowitz goes so far as to say (cf. paragraph 7.14 cited in paragraph [209] above) that it is “only” where there is such an ability that pricing-in becomes effective. The thrust of the Compass Lexecon / FTI report is to the same end.

The Claimants submit that I should decide this argument upon its intrinsic merits; but in the alternative it is an argument that should be subjected to an intensive scrutiny test consistent with the narrow margin of appreciation that the Defendant possesses in this area. The Secretary of State submits that his choice of the pricing-in theory was a perfectly rational decision, consistent with evidence, and squarely within his (very broad) margin of appreciation which arises in cases of this sort. Indeed, Mr Saini QC submitted that the High Court was not “competent” (jurisdictionally) to review this particular economic choice.

I have in Section F above addressed the extent to which a merits review arises in a case such as this. For the reasons set out there I reject the submission that I should decide this issue as if it were a right/wrong adjudication. Further, for the reasons set out below I do not accept that, even applying a relatively narrow margin of appreciation, the Defendant erred.

Before setting out my reasons I set out the approach that I have adopted towards the assessment of this specific issue.

The first point is that in relation to an issue such as this the actual distinction between a merits review and a judicial review might be more apparent than real. The extent to which courts will adjudicate upon technical or expert issues varies according to context. For instance courts are routinely required to determine, on a merits basis, clinical negligence cases where the case frequently turns upon the assessment by the court of complex, scientific and medical expert opinion. But even in such merits cases the test actually applied by courts is a hybrid judicial review type test. The Judge does not “decide” the science; on the contrary the judge hears expert evidence and then applies an approach which is akin to a public law rationality challenge. The approach adopted in law to determine these cases has a strong public feel about it. The test is known as the “Bolam” test. It essentially accepts that there may well be a number of different competing medical or scientific theories in a case and there may also be a range of quite different yet reasonable views. In such cases the court only rejects a Defendant’s view if it is outwith that reasonable range. If the Defendant’s case rests upon a theory or body of expert opinion that is within the range the Defendant will succeed even though the Claimant’s case also rests upon expert evidence which is reasonable and within the range. A Court must assess the evidence itself and cannot delegate, in effect, the judicial task to experts and case law lays down various criteria that the court must consider when reviewing the expert evidence. But the judicial task is essentially to review the evidence to determine whether the Defendant’s approach was within the range of reasonable options: See by way of an illustration of the articulation of this principle and its interrelationship with expert evidence - C v North Cumbria University Hospitals NHS Trust [2014] EWHC 61 at paragraphs [20ff].

In other areas where statutory appeals are mandated and where the court or tribunal is required to adjudicate upon the merits of decisions containing economic assessments (i.e. the case is not a de novo hearing), there are still occasions where a margin of appreciation has been recognised. For instance, in T-Mobile (UK) Limited et ors v Office of Communications (OFCOM) [2008] CAT 12, a case where the issue was the legality of a detailed administrative decision that relied upon complex economic arguments, the Competition Appeal Tribunal stated:

The underlying point of relevance is that the distinction between a merits review and a judicial review can be as much one of degree as it is of sharply defined, binary, black and white lines.

In the present case in order to perform an intensive review I have considered the following issues:

Conclusion

Applying these criteria I conclude that the Defendant acted lawfully. This is for a number of different reasons.

First, it is clear from evidence before the Court that there was a reasonable literature base upon which the Defendant was entitled to rely for support for the pricing-in theory. Indeed, the Defendant relied upon Professor Liebowitz as only one amongst a number of experts and economists who advanced the theory of pricing-in. Thus in the Updated Impact Assessment (ibid., page 215) the IPO refers to the theory “first proposed” by Professor Liebowitz “… and since developed by others” (such as Hal Varian “Copying and Copyright”, Journal of Economic Perspectives 19:2 121-138). In the course of argument Mr Saini QC referred to the CRA Report (see paragraph [6] above) and to other literature and expert analyses which endorsed the pricing-in theory as applicable in the relevant economic sectors and where the authors came to the conclusion that pricing-in was in principle capable to some degree of eliminating “harm”. In this body of literature there was no qualification that this could “only” arise where there was an ability to price discriminate. I have in the course of this case reviewed a number of different pieces of literature and in my view it is clear that many experts agree with the broad thrust of the reasoning behind the Defendant’s reliance upon pricing-in. In these circumstances it was within the Defendant’s margin of appreciation to accept the theory as prima facie applicable to the facts of the present case. It is not a rogue theory nor one that is generally rejected by experts. Indeed it seems to me that a Court would be entitled to come to this view even if a merits approach was adopted. This might be an area where there is quite properly room for more than one legitimate view.

Secondly, I turn next to a concern that I have about the price discrimination qualification. I have difficulties with this argument at two different levels. First, the concept seems divorced from the way in which prices are set in most normal markets for non-fungible goods. As such it is divorced from reality and not the sort of argument that is so powerful and compelling that could result in the Defendant being required to reject pricing-in as a theory of relevance. Secondly, it implies a legal concept of “harm” for the purpose of Article 5(2)(b) which assumes that sellers must be able to extract the very last gram of value from the copyright. However, in my view, the concept in law is not so purist or absolutist; “harm” as understood in Article 5(2)(b) is a broader concept which balances the interests of rightholders and users and incorporates notions of “fair balance” (cf. Recital 31). It is also a concept which should reflect market realities: see generally paragraphs [191-192] and [200] above. In most normal, everyday, markets sellers simply do not have the ability to engage in price discrimination and do not do so yet it is never said that they do not have the ability to gain a reasonable or fair value for their products. I thus have difficulty in accepting that an ability to price discriminate should be a sine qua non to the ability for seller to use pricing–in to capture relevant “harm”.

Professor Liebowitz’s criticism of the decision is that a content seller cannot extract full value unless that seller knows how frequently the consumer will use (i.e. copy) the content and hence the extent to which it is valued by the consumer. If the seller knows this he can differentiate in price as between consumers and (it is assumed) obtain more revenue. He argues, in effect, that because there is a substantial informational asymmetry between seller and buyer about these matters the seller does not know how, in effect, to pitch the level of the price and as such is not able to recover the full value of the copyright protected work. At one level this is probably correct because if a seller were possessed of perfect information that seller might be able to modulate prices so as to extract the very last penny of profit. However, the argument does not reflect the real world where in virtually all normal markets there is informational asymmetry and sellers of non-fungible goods (presumably not fungible goods, which get consumed and hence not copied or replicated) typically do not know how many times customers will use their product or how they (subjectively) will be valued and they do not thereby have the chance to price discriminate as between different consumers or groups thereof in order to capture the maximum value attributable to those factors. Almost every item sold by a department store is sold without the vendor knowing how often the chair or lawn-mower or cup and saucer will be used or whether, on purchase, the new owner will like the item or come to hate it. In other words normal markets do not operate upon the basis that when the vendor comes to set price and hence extract value that vendor has perfect foresight about rates of future usage. Yet sellers to a degree price-in predictions of subsequent use; a product that is stronger and more durable will generally attract a higher value to reflect those attributes, than one that is more fragile and less likely to bear continued usage. In the present case, as reflected widely in evidence submitted to the Hargreaves review team and to the IPO during the consultation, vendors have a clear understanding of when consumers are likely to copy the work and when they can use DRM to limit that copying. To an extent the informational asymmetry can be plugged by market research and other normal data and information collecting techniques. I do not perceive any systemic inability to price discriminate as suggesting any form of market failure.

Mr de la Mare QC, for the Claimants, argued that the analogy with ordinary goods was not an apposite one since there was no issue of copying which arose in those cases, for instance with a lawnmower. But in my view this does not address the concern that I have. Professor Liebowitz’s argument rests upon an inability to price discriminate to reflect a variable rate of usage. This is the same conundrum that other vendors of ordinary household goods face. A music download might be played once or a hundred times (if copied); and the same goes for the usage of a lawnmower (where usage is not predicated upon copying): the problem vendors face of not being able to link price to usage is hence as much an issue for the lawnmower seller as it is for the content seller.

In my view the criticism made that in the absence of an ability to price discriminate “private copying is virtually assured of harming rightsholders relative to a case where no private copying occurred” (see paragraph [210] above) is one which takes a highly purist and abstract view of “harm” which seems divorced from normal markets. It is also not a concept of “harm” which fits easily with the case law (see Section G above).

Ultimately, and save to set out my doubts about the price discrimination argument, I do not in this judgment express any final or concluded view on the merits of the debate. I confine myself to concluding that Professor Liebowitz’s criticism is not, in my view, so self-evidently true or compelling that it could ever amount to a reason to disturb the Secretary of State’s exercise of discretion to adopt the pricing-in theory as the test for determining when “harm” arose and therefore when the duty to compensate arose.

The third matter that I consider relevant focuses upon the procedure adopted by the Secretary of State which led to the adoption of the pricing-in theory. Where a decision maker adopts an economic principle which has been the subject of wide ranging consultation then a challenge based upon analysis which could have been, but which was not, advanced during the consultation must carry far less weight in the subsequent review exercise, including in one conducted with relative intensity. In this case Professor Liebowitz’s pricing-in theory was canvassed during the Hargreaves review (see e.g. paragraphs [50], [56] and [58] above) and was squarely set out at the beginning of the consultation (see paragraphs [66ff] above) and it was hence a legitimate target for any of the many, invariably sophisticated, consultees to shoot at. The attack, based upon the new reports by Professor Liebowitz and Compass Lexecon / FTI, on pricing–in was advanced after the consultation had finished and after section 28B had been adopted and necessitated an application to be made to me to admit the evidence. Mr Stout, who gave written evidence on behalf of the Minister, explained that evidence was actively sought on, inter alia, pricing-in but that “…the response was limited”. In assessing the lawfulness of the decision to adopt pricing-in I cannot ignore the fact that all of this analysis could have been advanced during the consultation itself and the fact that it was not is a factor weighing against a conclusion that the Secretary of State erred. In my view this answers also the criticism made by the Claimants that the Defendant should have verified the theory with Professor Liebowitz himself; this point lacks credibility. Professor Liebowitz was only one of the economists whose writings were relied upon by the Defendant. If the consultees were able to consult him after the event they could equally have done so before it. If they wished to persuade the Defendant that he was in error in relying upon Professor Liebowitz’s analysis then they could have sought his opinion at the outset of the consultation and presented it during the consultation. Even where the margin of appreciation is relatively narrow it is, in my view, a relevant consideration that the decision maker put his basic economic theory on the table at the outset and asked for views upon it.

Fourthly, the ability of any court to conduct an intensive review will depend also upon the evidence put before it by the parties. In this case the manner in which the attack on the economic logic of the decision has been advanced makes it difficult to accept. If a Court is to overturn an economic assumption made by a decision maker then it has to have before it all of the evidence that the decision maker considered so that it can be assessed in the round. It cannot be open to a Court to reject the Defendant’s assessment if only a small portion of the relevant evidence is relied upon for that challenge. By its nature - and especially in relation to an issue which splits an industry and stimulates partial submissions reflecting defined economic interests - if only that portion of the evidence which reflects but one side of the argument is put before the court then it will, inevitably, appear to be a powerful and coherent body of evidence which is inconsistent with the decision maker’s reasons. However, that might be for the very reason that it is only a portion of the evidence that was before the decision maker. To then say that there is a coherent body of evidence that contradicts the decision is true but an inapposite conclusion since it does not necessarily indicate that the impugned decision was outwith the decision maker’s proper discretion. In this case submissions inviting rejection of the Defendant’s economic pricing-in theory have rested upon only a small portion of the actual evidence before the decision maker. I have now read the contrary evidence. Submissions made have not focused upon why these other experts are wrong nor have they sought to weigh the pros and cons of competing economic theories.

In conclusion, even applying a relatively intensive standard of review, the Defendant was within his margin of appreciation to adopt the pricing-in theory in the relevant sectors of music, film and books.

Finally, an issue arose as to the admissibility of the expert reports of Professor Liebowitz and Compass Lexecon / FTI. In the event the Secretary of State did not object to their being admitted and submissions were therefore made about them. For the record I grant permission for those reports to be admitted into the evidence in this case.

The next issue concerns the Claimants’ submission that the decision was flawed because the evidence relied upon to justify the conclusion about harm was inadequate or even “manifestly” inadequate.

The conclusion that I arrive at below is that the decision adopted by the Secretary of State was nowhere near to being justified by the evidence that the Minister specifically accepted and endorsed. This is a conclusion I would arrive at applying almost any test of judicial review, howsoever intense or relaxed and deferential.

An important feature of this case is that the error in the Defendant’s reasoning flows from the evidence that he specifically endorses.

A challenge based upon a piece of evidence advanced during the course of the Consultation but not accepted by the decision maker is intrinsically less likely to sound in judicial review because the Secretary of State must consider all of the evidence in the round and the simple fact that he rejects, wholly or partially, one piece of evidence is not, in and of itself, an indication of unlawful conduct. In many consultations many hundreds of submissions and documents may be tendered. A court, assessing one piece of evidence, cannot readily, or even at all, replicate the assessment process conducted by the decision maker without also examining the totality of the evidence that was taken into account. And so it is likely to be only in rare circumstances that a court would accept as sufficient to grant relief submissions advanced on the back of only a portion of the evidence before the decision maker. For this to occur the evidence in question would have to be singularly compelling in its probative quality (and hence material to the outcome) and there would have to be evidence that the decision maker ignored it or gave it insufficient weight.

A very different conclusion, however, might arise in relation to the evidence which the Secretary of State himself accepts as admissible and relevant to the final decision. As to this, the Court is entitled to ensure that the inferences that the decision maker draws from this, ex hypothesi, important evidence are rational, i.e. they are inferences that may properly and reasonably be drawn. In such a case the Court is not questioning the decision maker’s selection of relevant evidence but simply ensuring that the evidence the decision maker has chosen to rely upon is sufficient to justify the inferences and conclusions drawn from it (see paragraph [168] above).

There are, in my view, 7 points to make about the evidence in this case which bear upon the outcome.

First, the court must examine the inferences drawn in the light of the relevant legal issue. In the present case the legal issue is whether pricing-in will lead to “minimal or zero” harm since if it does not then, in law, a compensation scheme must in law be introduced. The legal issue guides the exercise: The evidence accepted by the Secretary of State must be sufficient to enable a reasonable inference to be drawn from it that this de minimis threshold has been met.

Second, a court will examine the evidence to see if there are, logically, gaps in it. Such might well be the conclusion if experts selected by the Government to collect and analyse evidence themselves identify questions and queries and matters requiring further investigation which are not then followed up or which are ignored; or if consultees identify issues which the decision maker accepts are relevant but which for no identified reason are not followed up.

Third, a court will, even in relation to a narrow question such as whether the evidence established a de minimis threshold, accord to the decision maker an appropriate margin of discretion. So, if the evidence had suggested that (say) 93% of harm was obviated by pricing-in then the court might well conclude that it had no remit to quibble as to whether 95% or 97% should be treated as the relevant de minimis threshold. But the conclusion might be very different if the evidence either fails to address what is to be understood by de minimis at all or suggests an answer that (say) between 40-50% of rightholders would be deprived of compensation if the exception were introduced. In short the court will look to see if the inferences drawn are within the range of reasonable and rational inferences which are capable of being drawn from that evidence.

Fourth, the court must recognise that relevant evidence may be qualitative and/or quantitative. Experience shows that, even in areas of technical or economic complexity, decisions may, of necessity, be properly taken upon the basis of a mix of quantitative and qualitative evidence. Indeed, a decision maker might well adopt a particular proposition because it is simply intuitive economic common sense: See Sinclair Collis (ibid) at paragraphs [237], [238] and [242], and, Gibraltar Betting & Gaming (ibid) at paragraphs [110], [119], [120], [163] – [168].

Fifthly, when appraising the reasons in the decision based upon the evidence a court should not construe the decision as if it were a statute. The Court is concerned not to pick holes in infelicitous or badly chosen language but, rather, with seeking to understand and review the true substance of the reasoning. The courts have adopted this approach in relation to reports of the Competition Commission (Tesco Plc v Competition Commission [2009] CAT6 at paragraph [79] (“Tesco”)); and officers’ reports in the context of planning (e.g. South Somerset District Council v Secretary of State for Environment [1993] 1 PLR 80 per Lord Justice Hoffmann at page 83). In the present case, the critical and final reasoning is found in the Updated Impact Assessment of March 2014. Evidence before the Court demonstrates how that reasoning evolved over time. The task is to focus upon pith and substance and to avoid the temptation to pick holes in loose language.

Sixthly, a court needs to exercise care in placing too much store by witness statement evidence from decision makers, prepared for the purposes of litigation. In many cases claimants submit that the challenged decision should stand or fall upon its own merits and after the event (“boot strapping”) attempts in witness statements to add new or supplemental reasons, caveats or “spins” must be ignored: See, for example, Ermakov v Westminster City Council [1995] EWCA Civ 42; Lanner Parish Council v The Cornwall Council [2013] EWCA Civ 1290 at paragraphs [61] and [64]; and per Ouseley J in Ioannou v Secretary of State for Communities & Local Government [2013] EWHC 3945. In the present case, a different approach has been adopted by the Claimants who submitted that the Defendant’s witness statement evidence should be taken to reflect the reasoning of the Defendant and that it disclosed errors of law and evidence of predetermination. In my view, the reasoning the Defendant accepted has been set out comprehensively in the Updated Impact Assessment of March 2014 and it is to that document that I must primarily turn to identify the Defendant’s reasoning.

Finally and seventhly, even if the court concludes that there is an error a remedy will not be ordered if that error is not material to the outcome: see in this regard R v MMC ex parte National House Building Council [1993] ECC 388 at [398] (upheld on appeal: [1995] ECC 89); Tesco (ibid) at paragraph [79]; Gibraltar Betting and Gaming (ibid) at paragraphs [100] – [102].

I turn now to the reasoning in the Updated Impact Assessment.

I accept the Secretary of State’s starting point which is that he was entitled to draw inferences from the facts about the operation of the market. Mr Saini QC took me through the background to the decision which, as I have set out above, started with the Hargreaves Review and passed through a consultation and evidence collecting exercise. In my judgment, the Defendant was entitled to infer and accept, as a starting point for analysis, that: (i) not every consumer who engaged in unlawful copying would have purchased a duplicate copy of the copyrighted work absent that copying; and (ii), that pricing-in is a rational and normal pricing strategy for vendors, producers and suppliers to adopt in circumstances, such as the present, where there is uncontrollable copying of the product being sold and where the market has evolved upon that basis. The Defendant did not arrive at these inferences upon the basis of an evidential vacuum or by speculation (see below). He drew these economically intuitive inferences upon the basis of other facts that he was entitled to accept. These were set out in the Hargreaves Review Report and recorded in documents promulgated as part of the subsequent Consultation and which are also found in research documents and other academic literature and are set out in the Updated Impact Assessment. These facts include the following:

The two inferences referred to above in paragraph [246] fall, therefore, within the Defendant’s margin of appreciation. He could arrive at these conclusions as proper inferences to draw from a combination of the evidence collected and economic common sense.

At various points the Claimants accused the Defendant of treating speculation as fact. It is therefore appropriate to distinguish between speculation and the drawing of inferences. Speculation and the drawing of inferences are quite different things. A decision maker can properly draw inferences from facts through ordinary processes of deductive reasoning and by applying common sense. Speculation, however, is the forming of conclusions for which there is no evidential foundation and which are neither self-evident nor common sense. This is not a mere semantic quibble. In criminal trials, for instance, where juries are required to find facts, it is universal practice to direct juries to distinguish between speculation (which they cannot and should not do) and the drawing of inferences upon the basis of facts they accept (which they are entitled to do).

However, the fact that certain inferences may properly be drawn is not a complete answer in a case such as the present. In this case, the Defendant’s discretion is fettered by the fact that he is only permitted to avoid the introduction of a compensation scheme if the harm caused by the introduction of the copyright exception is de minimis or zero. This means that the evidence which he must collect and the inferences which he may draw must be sufficient to provide an answer to the specific question: Is harm minimal or zero? It is therefore no answer to this question to say that economic common sense suggests that “to some extent” pricing-in will obviate harm. That perfectly sensible conclusion might go some way to answering the requisite legal question but it does not go all the way. For example, if “to some extent” means that 60% of harm is obviated by pricing-in then unless the residual 40% can be treated in law as de minimis then economic intuition is not enough.

The Updated Impact Assessment posed the correct legal question which was whether there was evidence to support the conclusion that introduction of the copyright exception (Option 1) would cause minimal or zero harm. Having posed the question the Secretary of State answered it: Introduction of section 28B would cause minimal or zero harm. The evidence relied upon for this conclusion is clearly identified from the Updated Impact Assessment. It is set out on pages 16-18 and falls into two categories: First, the IPO Research Report; secondly, the other matters referred to and recited in full at paragraph [100] above. In the text below I consider, piece by piece, the evidence referred to and relied upon. I start my analysis by considering this “other” evidence.

iTunes: The first piece of evidence concerns iTunes. This is set out at paragraph [100] (2^nd to 4^th paragraphs therein) above. In the iTunes example, upon introduction downloads were priced at a higher price than content which could be copied to a more limited extent due to DRM. The reasoning states that this appeared to show strong evidence of pricing-in, as theory predicted. However, subsequently, the differential pricing exercise was abandoned by iTunes and this led, in the assessment, the IPO to say that accordingly:

“…it is difficult to base any comparative analysis on the current music download market”.

At its highest therefore this single incident, where there was prima facie evidence of consumers being prepared to pay more for unrestricted downloads, proved to have little probative value. As such it cannot provide more than the faintest of support for the strong proposition that pricing-in is endemic.

Use of DRM: The next piece of “evidence” relied upon by the Defendant is to be found at pages 16 to 17 of the assessment. It is contained within the quotation at paragraph [100] above (5^th paragraph therein). Here, the IPO, having downplayed the relevance of the iTunes incident, states that in relation to other download formats, such as video, these continued to use DRM to restrict the number of copies that can be made and responses to the Government’s Consultation suggest that rightholders expect in consequence to charge more for formats which physically permit more copying. They cite (in footnote 45) by way of example evidence tendered by the British Screen Advisory Council. This illustration arises in the film market but not the music market. It is in any event hard to describe the point as “evidence” relevant to the specific de minimis issue.

Consumer survey: The third piece of “evidence” relied upon is that which might be generated by a survey of consumer purchasing considerations: paragraph [100] above (7^th paragraph therein). The Defendant points out here that one approach to research could be to examine the impact on consumer demand and sales of copies. The tenor of the observation is that such information could have significant probative value. However, no such survey was ever commissioned. In the course of the hearing Mr Saini QC, in response to my question, confirmed that this was so but he could not, on behalf of the Defendant, explain why. Mr Mill QC, for the Claimants, pointed out that prior to the adoption of the decision (but after the Consultation closed) the Claimants submitted a summary report prepared by Compass Lexecon / FTI which did address this particular evidential lacuna but this is nowhere referred to in the reasons in the assessment. For present purposes it suffices to record that evidence the Defendant accepted would have been relevant and potentially valuable was not collected. This particular reference in the assessment is not, therefore, “evidence” in any rational or meaningful sense that could be relied upon to justify the challenged conclusion.

Price comparisons: The next piece of “evidence”, also found on page 17 of the Updated Impact Assessment: paragraph [100] above (8^th paragraph therein), lies in the explanation that a number of consultees proposed to the Secretary of State that a further way to examine pricing was to perform a price comparison exercise between an average digital download album with that of an average CD album. No such comparative exercise was, however, conducted. This was because the Defendant concluded, upon the basis of generalised objections, that it would not be useful. Once again, I am not in this analysis concerned to appraise the Defendant’s reasons for not pursuing this evidence. I am concerned only to identify what the “evidence” base for the conclusion that harm would be minimal or zero was. Accordingly, as with the survey of consumer perceptions, no price analysis was undertaken and, again, this cannot amount to “evidence” in any ordinary sense of that word which can be relied upon to justify the decision.

Comparables in other Member States: The next piece of “evidence” referred to in the reasoning was an analysis of comparative prices between Member States of the EU with different private copying exceptions including those with device-based levies and those without: paragraph [100] above (paragraph 9 therein). Once again the Defendant did not pursue such an investigation. The reasoning is set out in the assessment:

“We do not consider such an analysis to be very useful, however, due to a number of factors affecting pricing in different markets, including exchange rate fluctuations”.

It follows that a comparative analysis of the position in the UK and other Member States did not arise. Yet again I am not concerned with the correctness of the conclusion. I am concerned simply to record that no analysis was performed.

Ultimately, it is very difficult (impossible) to see how these matters referred to in the reasons can, upon any rational basis, be said to amount to “evidence” which justified the decision. On the contrary, the Updated Impact Assessment does no more than provide explanations as to why investigations were not pursued.

Films/LECA: I should also mention films. As to this, evidence was submitted by the MPAA (see paragraph [110] above) which included a summary report by Law and Economics Consulting Associates (“LECA”). This was submitted prior to the promulgation of the Updated Impact Assessment and it is evident from the disclosed documents that it was accepted as being relevant and admissible to the Defendant’s on-going investigations. LECA purported to present new evidence that undermined the Government’s conclusions that private copying would not inflict serious harm on the film and TV sector and which suggested that the proposed copying exception could reduce revenues in the UK film and TV industry by as much as 15%. However, of far greater relevance in my view was the alleged identification of key errors in the interpretation of the data relied upon in the Updated Impact Assessment. In other words MPAA focused upon evidence that the Defendant accepted as relevant and probative and submitted that it had been misconstrued. It is a criticism that can fairly be said to go to the core of the Defendant’s reasoning on film. LECA identified the claim in the 2012 Impact Assessment that “private copying for personal, non-commercial use is already widespread”. LECA submitted that their own assessment of the underlying sources of this statement demonstrated, to the contrary, that private copying for personal, non-commercial use was not in fact widespread. LECA found that only “…around 1% of film and TV buyers engage in private copying of the type covered by the proposed new law, and that the most common reasons people give for not copying are that they were satisfied with the product purchased and that there is a current law against it”. LECA pointed out that for the assumption that illegal private copying was widespread the Defendant had relied upon a report published in 2010 by Consumer Focus which was itself based upon a Kantar Media Survey commissioned in 2009. The Consumer Focus survey had been specifically identified and relied upon by the IPO in the 2012 Impact Assessment. For example in that assessment the following was stated:

“In a recent Consumer Focus survey, 85% of consumers thought it was already lawful to copy a CD or DVD to an iPod, mobile phone or other mobile device and 91% thought it should be”.

LECA however argued:

“The Government appears to have relied upon this result to assume that between 85% and 91% of people currently engage in private copying – and all that a new private copying exception would do is to legalise this behaviour without other effects.

To further assess the Government’s assumptions, we used a Freedom of Information request to obtain the actual raw data direct from Consumer Futures. That data underlies the Consumer Focus Report from 2010. Our review of this survey data shows clearly that the extent of illegal private copying is much more limited than assumed by the Government. The Government’s own survey showed that the proportion of the population copying DVDs or CDs for personal use as a whole – for example (i.e. not limited to specific devices) – is not extensive. Only 28% of the total population surveyed engaged in making copies of CDs or DVDs for personal use (561 out of 2026 surveyed). This is far from widespread. Instead, as outlined below, our recent survey properly measured private copying of the type covered by the proposed new law. It showed that less than 3% of the population were engaged in such behaviour, much less than the 28% measured in the Kantar Media study. Thus the Government’s claim that private copying for personal, non-commercial use was widespread exaggerated the result from the Government’s own data (28%) – and relied upon a measure which was in any event inflated and which failed to accurately measure the behaviour in the first place (less than 3%)”.

LECA then identified other areas within the 2012 Assessment where the Defendant relied upon the same Kantar Media survey.

Internal emails disclosed by the Defendant showed that the points made in the LECA analysis were taken seriously by civil servants who recognised that there was a “need to give a well argued, measured response”. However, it is apparent from the Updated Impact Assessment that the Defendant persisted in relying upon both the Consumer Focus report and the Kantar Media survey, both of which are cited upon a number of occasions for precisely the propositions which were challenged by LECA.

In the course of oral argument, Mr Saini QC, having taken instructions, was unable to explain what internal steps were taken to verify or address the complaint made by LECA that the Defendant had drawn incorrect inferences from data which he relied upon. It therefore appears that LECA’s criticisms were not followed up.

As with other pieces of the evidence relied upon I am not concerned with the ultimate correctness of LECA’s analysis, though prima facie it raised arguable questions as to the reliability of the inferences drawn and relied upon in the 2012 Assessment and later in the Updated Impact Assessment. My concern here lies in the fact that a properly arguable challenge to the reliability of inferences drawn from data the Defendant accepted was launched, and was then seemingly ignored.

Historical, pre-digitalisation sales: Next, it is submitted that the Defendant omitted, in the Updated Impact Assessment, to consider a possibly material category of loss. This is copying based upon purchases made historically, pre-digitalisation. The point was made by consultees during the consultation process that this was a category of harm for which compensation was required. This covered for instance purchases of vinyl records from (at least) the 1960s onwards; and CDs from the 1980s onwards. The copyrighted work deployed in the course of argument to illustrate the Claimants’ point (because it was no doubt speculated that it would appeal to the Court) was Pink Floyd’s “Dark Side of the Moon” released in March 1973. When this was first released there was, of course, no material scope for purchasers to copy the album (save possibly by playing it and simultaneously recording it through the air onto a tape cassette, generating a passable but tinny sounding rendition) and hence the basis upon which it could be said that there was pricing-in of subsequent copying was absent. However, “old vinyl” records were now very popular not least because record players were now sold with in-built copying facilities (such as USB ports etc). Dark Side of the Moon could now be copied extensively from old records with excellent results. The Claimants also adduced evidence to the effect that between 1983 and 2000 about 1,506,741,000 CDs where shipped into the UK with a trade value exceeding £7 billion and that it was during this period that copying became more prevalent. They accordingly submitted that in any evaluation of the extent to which pricing-in could capture all relevant “harm” an assessment had to be made of the modern copying of old vinyl records and CDs where, on no proper basis, could it be said that pricing-in occurred. However, there was no evidence that such as assessment had ever occurred.

Mr Saini QC, for the Secretary of State, accepted that such an assessment had not in fact occurred and indeed it is clear from the documents that no such exercise did take place. He submitted instead that any vendor of content even in pre-digitalisation days would have priced-in “such copying as would occur over time”. With respect, this is no answer to the criticism which is made. The Defendant’s case on pricing-in takes as one of its core factual premises that it is digitalisation that has made copying endemic and that pricing-in would not exist but for digitalisation. But no such premise can be taken to hold true pre-digitalisation. The truth is that the Defendant simply did not think of this category of relevant harm during the consultation and hence it is not included in the Updated Impact Assessment as part of the reasoning for the conclusion that pricing-in obviates harm. I am unable to say that, had it been investigated, it would have turned out to be material; but the point cannot be discarded without proper investigation. In my view this is another area where the inquiry was incomplete.

The IPO Research Report: This brings me to the remaining, and principal, category of evidence relied upon, namely the independently commissioned IPO Research Report which I have summarised in Section E above. In this part of the judgment, I confine myself to setting out my conclusions only on this research.

First, it provides useful evidence in relation to the control (software) which supports the Defendant’s intuitive conclusion that to some extent pricing-in occurs in copyright related industries. However, even in the software sector there is no analysis of whether pricing-in is complete or leaves some harm which is unaccounted for by pricing-in; and if so whether this residual harm is de minimis (see paragraphs [115] – [117] above).

Second, a similar conclusion occurs in relation to books and in relation to films (see paragraphs [122] – [125] above). These provide some evidence supporting the proposition that pricing-in arises but it does not provide evidence that it is total and would leave only “minimal or zero” harm.

Third, in relation to music (see paragraphs [118] – [121] above), there are three critical findings and points to be made: (i) the empirical research did not show evidence of pricing-in as between CDs and downloads but, on the contrary, it showed precisely the opposite effect; (ii) the researchers found this conclusion to be “puzzling” and the evidence, at best, “ambiguous”; (iii) the researchers speculated about the reasons for these unexpected results and whether they were, for example, due to “a lack of variability in the explanatory variables” and/or the private copying was “already largely fully priced-in in the UK market” but this speculation served to highlight that the research was incomplete as an exercise designed to answer the de minimis question. Put another way it raised more questions than it answered.

Absence of any explanation of de minimis concept: Finally, there is nowhere set out what the Secretary of State understands by the de minimis concept. In other areas of EU law (such as competition law) the de minimis concept is defined by market share and turnover figures. That level of exactitude might not however be possible in the present case (of paragraph [143] above) but the Defendant has not examined whether this is so. Nonetheless, since the test in law is framed in terms of a de minimis threshold it assumes that Member States will be able to determine when it is met, or not. In my view, the Secretary of State will need to consider both the global sums which might be uncompensated for; but also the scale of the sums likely to be payable to individuals. If, for instance, the Secretary of State concluded that there were a very large number of rightholders who would be due some minor payment I do not consider that the fact that the sum total of all outstanding payments might appear to be a significant figure necessarily means that the de minimis exception would not apply. I consider that the Secretary of State is also entitled to take proportionality into account; if the payments due are overwhelmingly trivial then it might be disproportionate to set up a system that may be complex and bureaucratic, and which was criticised in the Hargreaves Report, simply to administer the payment out of a large number of very minor payments. At all events, what is lacking from the Updated Impact Assessment is any articulation of what is meant or understood by the concept of de minimis in the context of harm flowing from the copyright exception. Given the fundamental importance of this concept this is a material failing.

Where does this lead to? In my judgment the inferences drawn in the Updated Impact Assessment about de minimis are not remotely supported by the evidence which the Defendant asserts and claims justify the conclusion drawn from it.

In my judgment:

I accept that there was a sufficient evidence and literature base for the Defendant to draw certain common sense economic intuitions about pricing-in. The Claimants do not dispute this proposition. However, these common sense intuitions were sufficient as a starting point but were not capable of answering the very much more specific legal question which was whether pricing-in was so extensive as to render residual harm minimal or non-existent. It is one thing to say that “to some extent” harm is avoided by pricing-in; it is altogether another thing to say that it is avoided so completely as to pass a de minimis threshold.

For the avoidance of any doubt, I am not saying that the Defendant had to prove on a balance of probabilities with hard-edged quantitative evidence that the residual harm was minimal or zero. It is possible (I express no view) that it is not feasible to obtain the evidence to prove this categorically and this might be a lesson which could be drawn from the IPO Research Report. This is why it lies within the competence and discretion of the Defendant to use a proper combination of quantitative and qualitative analyses (for instance surveys) in order then to draw inferences. I thus accept that there is an element of art and judgment as well as science inherent in the exercise. But that exercise needed to be conducted, and it was not.

In conclusion, the decision to introduce section 28B in the absence of a compensation mechanism is unlawful.

The Claimants submit that the Secretary of State predetermined the outcome of the consultation exercise in that he was determined from the outset to introduce the copyright exception but without any accompanying compensation scheme and pursued this objective without regard to the evidence. The submission is couched in terms of actual predetermination but I have taken it as a submission as to the appearance of predetermination as well: see paragraph [168] above. It is submitted that this violates the fairness principle which is reflected in (in particular) the fourth “Sedley” principle (see paragraphs [166ff] above) and amounts to unlawful predetermination. They rely upon the Updated Impact Assessment (ibid., page 22) which states that:

“In view of the Government’s intention to implement the exception without introducing a levy or similar mechanism, we have made every effort to minimise harm caused by it”.

The Claimants submit that this intention “…suffused the entire process” and that it can be inferred from the facts that the Secretary of State entered upon the Consultation exercise with an unshakable and fixed determination to introduce copyright exception without any concomitant compensation provisions and, in effect, that no amount of convincing and compelling contrary evidence was going to dissuade the Secretary of State from this fixed position.

I can deal with this submission quite briefly. I reject it for the following four reasons.

First, the Secretary of State was entitled to have a strong predisposition. The distinction between a predisposition and predetermination is well understood in the law. A decision maker may consult upon an issue that he has a firm view about. Indeed, if the decision maker’s cards are laid squarely upon the table consultees are fully informed as to that predisposition and have the clearest possible target at which to aim their submissions. A strong predisposition is not, therefore, inimical to a fair consultation assuming, of course, that the decision maker is prepared to keep an open mind and be willing to change his or her views if the evidence and submissions tendered are properly persuasive.

Secondly, the Secretary of State was not in law obliged to introduce the copyright exception and was entitled, if he concluded that a compensation scheme was the price that would have to be paid for the exception, simply to revert to the status quo and not introduce an exception at all. Accordingly, a strong predisposition to push through a copyright exception without a compensation scheme was a lawful position for the Defendant to adopt. The Defendant consulted upon four options, one of which (Option 0) was “do nothing”, i.e. not introduce a copyright exception and another option (Option 1) was to introduce an exception only if it led to minimal or zero harm and thereby did not necessitate the introduction of a compensation scheme. In my judgment, these options were lawful. Under Article 5(2) of the Directive each Member State has a discretion to create an exception; but they are not obliged so to do. Hence, in leaving open the option to refrain from introducing any exception, the Defendant was doing no more than reserve his right under Article 5 to decide not to introduce an exception. Choosing Option 0 would not therefore have been inconsistent with EU law; and choosing Option 1 was lawful provided the exception could be introduced without more than de minimis harm to rightholders.

Thirdly, no evidence was placed before the Court which showed that the Secretary of State had, in actual fact, predetermined the outcome. Indeed there was no evidence at all of the actual thought processes of the Secretary of State in 2014 when the decision was taken to introduce the new section 28B. The only evidence before the Court of “internal” thought processes were certain emails disclosed by the Defendant in the course of the litigation emanating from civil servants. The Claimants submitted that these emails were consistent only with the conclusion that the possibility of introducing an exception with provision for fair compensation had been finally rejected long before the Consultation concluded. In my judgment, the emails prove no such thing. There were two emails which were relied upon both authored by officials from within the IPO. The first, dated 24^th April 2012, stated:

“[Name] and I were discussing private copying today and we wonder if we should put this one on hold pending the current ECJ case. This should either say harm always occurs in which case no exception without levies or that harm can be minimal in which case a green light to legislate. Without the ECJ case we’d have to flag up the possibly significant risk that any exception would immediately be subject to judicial challenge”.

The reply, dated 25^th April 2012, stated:

“I am rather of the view that we must press ahead with “difficult” exceptions like private copying. […] if the Court rules we are right, then great. If they rule that fair compensation = 0 in only the most utterly trivial of circumstances, then Ministers will have to face the choice between paying that compensation (whether by levy or otherwise) and not having the benefits of private copying”.

These emails certainly reflect the strong predisposition which the Minister transparently maintained from the outset. However they also show that if the evidence (or the law as laid down by the ECJ) demonstrated that there would be harm through the introduction of an exception then the officials would need to put to the Minister the option either to pay the compensation or not introduce the exception at all. For the reasons set out above, this was a lawful choice to place before Ministers. It certainly does not evince any form of predetermination as to the outcome, by the Secretary of State. It is of course also obvious that officials do not necessarily speak for a Minister and it cannot be inferred simply from the recitation of the views of officials (in 2012) that these reflect the views of the ultimate decision maker (in 2014). For it to be possible to attribute the views of officials to those of a Minister there would have to be evidence indicating that the internal views expressed by the officials accurately reflected the Minister’s view and had been acted upon by him when the decision was taken. These emails do not provide that evidence.

Fourthly and finally, as to the invitation made by the Claimants to me to infer from the end result that the Minister had apparently predetermined the outcome, this is to elide two different matters which are (a) the drawing of incorrect inferences from evidence, and (b) the decision maker’s thought processes. I accept that one component of a conclusion that a decision maker has unlawfully closed his mind, fettered his discretion, and predetermined a matter, might in principle be a finding that unwarranted inferences were drawn from evidence. However, in the present case the conclusions as drafted and set out in the Updated Impact Assessment would have indicated to the Minister when he read them that he could legitimately proceed to introduce the copyright exception without a compensation mechanism. The fact that (as I find) the Defendant erred in this regard does not prove predetermination or the appearance thereof. It suggests only that the Defendant erred in accepting the reasons in the Updated Impact Assessment. It will thus be rare for inferences of this type to amount to conclusive evidence of predetermination. The evidence may suffice to mount a rationality challenge but it does not, necessarily, prove predetermination. Mr Mill QC confirmed that the Claimants did not allege bad faith on the part of the Defendant. The emails relied upon show the thought processes of officials, not the decision maker. The Defendant has given disclosure of relevant internal documentary material and had there existed evidence that the Minister had, for instance, expressed the position that he would proceed with the copyright exception regardless of the evidence, then that would have been disclosed. But there is no evidence that the Minister did anything other than accept the conclusions in the Updated Impact Assessment. Ultimately, in the absence of any evidence that the Defendant in actual fact prematurely closed his mind to any evidence which undermined Option 1, I cannot accept the submission that the outcome was predetermined.

It is for these reasons that I do not accept that the application succeeds on grounds of alleged predetermination.

The issue and the approach to be adopted

The final issue concerns State aid. The Incorporated Society of Musicians Limited, the Intervener, was granted leave to intervene upon the basis that it wished to contend that the effect of the introduction of section 28B in the absence of a compensation provision amounted to unlawful state aid contrary to Article 107(1) of the Treaty on the Functioning of the European Union (“TFEU”). This provides that:

“Save as otherwise provided in the Treaties, any aid granted by a Member State or through State resources in any form whatsoever which distorts or threatens to distort competition by favouring certain undertakings or the production of certain goods shall, in so far as it affects trade between Member States, be incompatible with the internal market”.

Article 107(2) and (3) TFEU stipulate that aids of certain types either shall or may be compatible with the internal market and that the determination on whether an aid is so compatible is the exclusive prerogative of the European Commission: see for example Case C-354/90 FNCEPA [1991] ECR I-5505 at paragraph [14]. Article 108(3) TFEU imposes upon Member States an obligation to inform the Commission of plans to grant or alter aid and it prohibits the implementation of proposed measures pending such notification. It has been long established that Article 108(3) TFEU is directly effective. As such, it may be relied upon before domestic courts as a ground for impugning the legality of a legislative measure said to constitute unnotified and hence unlawful state aid. In such a challenge the court must form its own view as to whether the impugned measure or act does, or does not, involve the grant of an aid within the meaning of the Treaty: see, for example, R v Customs & Excise Commissioners ex p Lunn Poly Limited [1999] 1 CMLR 1357 at paragraphs [22] – [24] per Lord Woolfe MR. If the measure does amount to “aid” and it has not been duly notified to the Commission then it is unlawful.

The analysis to be undertaken of “aid” by a Court may involve the consideration of a number of quite different components. In my judgment the question whether there is aid “through State resources” is an objective question for the Court and does not involve the conferral of any margin of appreciation upon the decision maker. The facts which must be considered by the Court do not involve any evaluative judgment on the part of the Defendant; the Court simply has to identify the manner in which the advantage allegedly comes about and then assess the nature of the link between the advantage and the State budget in terms of the closeness and strength of the nexus. The relevant facts are fixed and within a relatively narrow compass. It is possible that the Court, in another case involving other component parts of the definition of “aid”, might need to adopt a more limited review. So, for instance, if the issue was whether the market investor test was satisfied and it could be shown that on one reasonable analysis the test was met a Court might be loathe to substitute its own view for that of the decision maker. I do note in this regard however that the Court of Justice has stated that even where the analysis of whether “aid” exists is “technical or complex” the Court (in casu a judicial review by the General Court of a decision of the Commission) must conduct a “comprehensive review”: see e.g. Case C-487/06P British Aggregates Association v Commission [2008] ECR I-10515 paragraph [114]. I emphasise however that no such complex technical or economic issue arises on the facts of the present case and I do not therefore express any view on how a Court would address other more complex components of “aid”.

In a leading text (Bacon, European Union Law of State Aid (2^nd ed, 2013) (“Bacon”) it is stated that: “The jurisprudence of the European Court has not yet provided a consistent and comprehensive interpretation of the conditions for State aid under Article 107(1). Whilst the Court’s definition of aid is often still based on the actual wording of Article 107(1), in practice the Court has not adhered rigidly to that formulation” (ibid., page 20, paragraph [2.02]).

It is also stated (ibid) that, from case law, the following four conditions must be met before the provision is engaged:

The nub of the dispute between the parties arises in relation to the Intervener’s submission that there has been an aid “through state resources” (cf. the second condition above). A large number of legislative measures will have winners and losers. The administrative practice of performing impact assessments means that Government will routinely endeavour to estimate in monetary terms the economic advantage accruing to the winners. That advantage is invariably an inherent part of the legislative instrument. It would not therefore ordinarily be treated as a revenue opportunity foregone by the State and it is artificial to seek to categorise it as revenue foregone by arguing that the State should have charged the winning category of persons in order to negative the advantage and that the omission to do so is therefore lost revenue. The question in this case is whether the advantage to the “tech-sector” identified in the Updated Impact Assessment can be said to amount to revenue foregone by the State and therefore aid “through State resources”? The particular issue is whether the advantage which has been identified is so closely connected to State resources that, contrary to the normally expected position, it can properly be said to flow through the State’s budgetary resources and hence constitute “aid”. As I explain below the Court of Justice has developed two tests for drawing the line between advantages which are and are not “aid”. The first test focuses upon the strength of the link between the alleged advantage and the State’s budget; the second test starts from the premise that an advantage does arise but then asks whether it is inherent in a legislative or regulatory scheme designed to meet another purpose.

This case focuses upon one critical component of “aid”; there has been no detailed evidence in this case as to the other components and conditions, including whether the alleged advantage is liable to distort competition and/or affect trade between Member States.

The Intervener relies upon the conclusion in the Updated Impact Assessment that the copyright exception whilst causing no harm to rightholders would create substantial benefits for technology firms. The assessment (page 2) estimates the benefits to such firms as being in the order of £258 million over ten years. The Intervener cites the benefits attributable, in particular, to cloud storage providers from the removal of the requirement upon them to pay rightholders for licences:

“Benefits…are expected to arise from: (a) reduced costs due to not having to pay for, or reduce licence fees for, storing copies in a private cloud; (b) reduced admin costs due to not having to negotiate licences; (c) reduced legal risk and uncertainty”.

The Defendant has been unable to quantify, with precision, the transfer of value to technology firms. In the Updated Impact Assessment the transfer was said to be unlikely to be as high as £8 million per annum but unlikely to be nil. The benefit to technology firms was said to be “slightly” in excess of the cost to rightholders. The Intervener submits that, notwithstanding the relatively small scale of the benefit, it was treated as being significant having been specifically identified in the assessment under the heading “Key non-monetised benefits”. The Intervener submits that the economic advantages which technology firms are accorded by operation of the new copyright exception are not benefits they would receive under ordinary market conditions. It is submitted that they thereby meet the test for an “aid”.

In their written submissions, the Intervener stated as follows:

“In effect, they are granted a statutory licence which (on the Defendant’s own reasoning) reduces their costs and facilitates market entry and expansion. It is precisely on the basis that those economic advantages would not occur in ordinary market conditions in the absence of State intervention that the Defendant in the UIA preferred Option 1…”.

The Intervener then submits that this aid to technology firms, being granted by means of secondary legislation creating an exception from the general domestic law on copyright, is both attributable to the State and granted “through” State resources, for two reasons.

First, the Defendant, in creating the exception, had the power under EU law to impose a private copying levy on technology firms pursuant to Article 5(2)(b) but chose not to impose such a levy and has thereby foregone potential revenues which it could otherwise have collected under the levy. Alternatively, a levy could have been imposed under the European Communities Act 1972. This amounts to a grant directly or indirectly through State resources, there being a direct connection between the gratis creation of the exception and the loss of revenue. They cite in support Case C-279/08 P Commission v Netherlands (Emissions trading scheme) [2011] ECR I-07671 at paragraph [111] (the “Netherlands (Emissions trading scheme)” case). I return to this authority later.

Secondly, it is submitted that the aid is granted through State resources because by creating the exception without providing any domestic compensation arrangements the UK has itself assumed a direct liability to compensate rightholders under Article 5(2)(b) which is of direct effect alternatively upon the basis that the unlawful refusal to compensate rightholders exposes the State to liability.

The Defendant rejects the above submissions. He submits that the Intervener’s case fails “at the first hurdle” because there has been no grant of aid to technology firms out of State resources. Instead, the Secretary of State has done no more than adopt a beneficial legislative measure which gives rise, inter alia, to potential economic benefits for online service providers. However, those positive benefits to technology firms have not been financed by any corresponding cost, expense or burden on the State. In particular the Secretary of State advances the following three propositions.

First, with regard to the contention that aid is granted because the UK has foregone potential revenue that it could have derived from the imposition of a private copying levy pursuant to Article 5(2)(b), the mere fact that the UK had the (un)exercised power to impose such a levy does not indicate that it should be classified as granting aid out of State resources. There is no cost or expense which has been absorbed by the State by reason of the failure to impose a levy. Indeed, it is strongly doubted whether any charge could ever have been imposed under Article 5(2)(b) since such charges are levied to fund compensation to rightholders not negative collateral advantages to third parties.

Secondly, in relation to the submission that aid is granted through State resources because by creating an exception without providing for any domestic compensation arrangements the UK has itself assumed a direct liability to compensate rightholders, the Secretary of State rejects the notion that in law he has any obligation to compensate rightholders under the Directive which, he submits, is not in any event of direct effect.

Thirdly, the Secretary of State advances a broader point: if the mere fact that a legislative measure exerted positive effects for particular industries meant that they became State aid then virtually any beneficial legislative intervention would be prohibited, at least without the prior consent of the European Commission. It is commonplace for legislative measures to create benefits for some and detriments for others; but State aid laws are neither designed nor intended to prohibit such measures by reason of these obvious and often inevitable effects. The Defendant relies upon the principle, well established in case law, that the legislative allocation of financial burdens or benefits as between private undertakings does not amount to the character of State aid nor does the fact that this may result in reduced tax receipts by the State. The Defendant refutes the Intervener’s interpretation of the Netherlands (Emissions trading scheme) case and refers to the ruling of the Court of Justice in Case C-379/98 PreussenElektra [2001] ECR I-2099 at paragraphs [58] – [62] (“PreussenElektra”).

In my judgment there is no aid granted through state resources: Article 107 TFEU does not therefore apply. Moreover, had I come to a different conclusion I would not have held that Article 107 TFEU applied but I would have remitted the issue of State aid to a further hearing.

State aid issues arise in public law litigation not infrequently. Article 107 TFEU is an important provision of law but it has acquired a reputation as an argument of last resort because it is often thrown in at the end of a long list of other arguments and is not always given the level of attention that is required if the arguments are to be made good. The present case is an illustration. The Intervener did not seek its own judicial review; it joined the Claimants’ but with its own discrete argument. As such it has had to be content with life as an Intervener which has strictly limited the evidence and arguments it can advance. The Defendant for his part was confronted with these arguments late in the day, shortly before the hearing, and has therefore chosen to contest the issue on a limited battle ground. In the event this strategy has succeeded. But I have doubts as to whether other conditions of the definition are in fact met. In particular, I am not clear that there is “aid” or (whilst recognising that the thresholds in law are relatively low) whether even if there is “aid” that the benefits and advantages said to amount to the “aid” are sufficiently large and focused to be liable to distort competition, or, affect inter-state trade. It seems to me to be at least arguable that the benefits identified are so remote and diffuse that they do not amount to “aid” and/or that they would exert no sufficient impact on the market or trade between Member States. The market is vast and international and at best the benefits amount to circa £26m per annum spread throughout that market in ways that might well not be such as to cause any discernible alteration in competitive or commercial behaviour. Whether this is so or not would require some detailed (and possibly expert) analysis and this has not arisen in this case. The consequences of a Court applying Article 107 TFEU may be very serious and should not be so applied, by a side wind, without that Court having undertaken a proper analysis of all the factual conditions precedent. Had I therefore agreed with the Intervener on the point in issue I would not have granted relief in relation to this piece of important legislation upon that score, without first being fully satisfied that Article 107 TFEU was properly engaged in all its facets.

For the purpose of the analysis below I will therefore assume for the sake of argument – but without making any positive findings to that effect – that the advantage referred to by the Intervener (i) is liable to distort competition and (ii) is liable to affect trade between the Member States.

The following are my reasons for the conclusion that the aid does not arise through state resources.

In Bacon (ibid.) it is explained that until recently there was confusion in the case law as to whether Article 107(1) TFEU should apply to any measure which conferred economic advantages on specific undertakings which was the result of conduct attributable to the state regardless of whether it involved any financial burden for the State (ibid., page 61, paragraph 2.94). It is also stated (ibid., paragraph, 2.95) that the Court of Justice has now clarified the state resources argument in PreussenElektra. This authority is relied upon by the Secretary of State. In that case at issue was the German Stromeinspeinungsgretz, a legislative measure which required private electricity distribution undertakings to acquire a fixed amount of electricity from renewable energy suppliers, at fixed minimum prices. The difference between those prices and market prices was paid in part by conventional electricity suppliers. It was accepted that the net economic effect of the scheme was to confer advantages upon one class of trader. The Court held however that this did not amount to State aid:

“58.

In that connection, the case-law of the Court of Justice shows that only advantages granted directly or indirectly through State resources are to be considered aid within the meaning of Article 92(1). The distinction made in that provision between ‘aid granted by a Member State’ and aid granted ‘through State resources’ does not signify that all advantages granted by a State, whether financed through State resources or not, constitute aid but is intended merely to bring within that definition both advantages which are granted directly by the State and those granted by a public or private body designated or established by the State (see Case 82/77 Van Tiggele [1978] ECR 25, paragraphs 24 and 25; Sloman Neptun, paragraph 19; Case C-189/91 Kirsammer-Hack [1993] ECR 1-6185, paragraph 16; Joined Cases C-52/97, C-53/97 and C-54/97 Viscido [1998] ECR I-2629, paragraph 13; Case C-200/97 Ecotrade [1998] ECR I-7907, paragraph 35; Case C-295/97 Piaggio [1999] ECR I-3735, paragraph 35).

59.

In this case, the obligation imposed on private electricity supply undertakings to purchase electricity produced from renewable energy sources at fixed minimum prices does not involve any direct or indirect transfer of State resources to undertakings which produce that type of electricity.

60.

Therefore, the allocation of the financial burden arising from that obligation for those private electricity supply undertakings as between them and other private undertakings cannot constitute a direct or indirect transfer of State resources either.

61.

In those circumstances, the fact that the purchase obligation is imposed by statute and confers an undeniable advantage on certain undertakings is not capable of conferring upon it the character of State aid within the meaning of Article 92(1) of the Treaty.

62.

That conclusion cannot be undermined by the fact, pointed out by the referring court, that the financial burden arising from the obligation to purchase at minimum prices is likely to have negative repercussions on the economic results of the undertakings subject to that obligation and therefore entail a diminution in tax receipts for the State. That consequence is an inherent feature of such a legislative provision and cannot be regarded as constituting a means of granting to producers of electricity from renewable energy sources a particular advantage at the expense of the State (see, to that effect, Sloman Neptun, paragraph 21, and Ecotrade, paragraph 36)”.

The effect of this (according to Bacon, ibid., page 63, paragraph 2.96) is that “… notwithstanding the wording of Article 107(1), the conditions of a grant of aid and a grant through state resources are to be interpreted as cumulative rather than alternative”. The Intervener must therefore show both that the measure involves state resources and also that it is imputable to the state.

As to the circumstances when an advantage conferred may properly be defined as being through state resources the law has evolved through the judgment of the Court of Justice in Netherlands (Emissions Trading Scheme) in particular at paragraphs [104] - [111]:

“104.

In that regard, it should be noted that, according to settled case-law, it is not necessary to establish in every case that there has been a transfer of State resources for the advantage granted to one or more undertakings to be capable of being regarded as a State aid within the meaning of Article 87(1) EC (France v Commission, paragraph 36).

105.

Also, the distinction made in that provision between ‘aid granted by a Member State’ and aid granted ‘through State resources’ does not signify that all advantages granted by a State, whether financed through State resources or not, constitute aid, but is intended merely to bring within that definition both advantages which are granted directly by the State and those granted by a public or private body designated or established by the State (PreussenElektra, paragraph 58, and Case C-222/07 UTECA [2009] ECR I-1407, paragraph 43).

106.

In the present case, an advantage granted by the national legislature, that is, the tradability of NO_x emission allowances, could entail an additional burden for the public authorities in the form of an exemption from the obligation to pay fines or other pecuniary penalties (see, to that effect, Case C-295/97 Piaggio [1999] ECR I-3735, paragraph 42). By establishing the ‘dynamic cap’ scheme, the Kingdom of the Netherlands gave to the undertakings covered by the measure in question the possibility of buying emission allowances in order to avoid the payment of fines. In addition, the consequence of that system is the creation, without real consideration supplied to the State, of emission allowances which, because of their tradable character, have an economic value. It must be concluded that the Member State could have sold such rights, or where appropriate put them up for auction, if it had structured that scheme differently (see, in that regard, paragraphs 63, 64 and 86 to 96 above).

107.

Thus, the General Court’s conclusion in paragraph 75 of the judgment under appeal that the undertakings concerned are free to sell their emission allowances even though they are linked to a maximum ceiling, is not called into question by the arguments submitted. As the Advocate General considered in point 87 of his Opinion, the Member State, by conferring on those emission allowances the character of tradable intangible assets and by making them available to the undertakings concerned free of charge instead of selling those allowances or putting them up for auction, foregoes public resources.

108.

In addition, the fact that such a measure permits undertakings to offset among each other the surpluses or deficits in relation to the standard laid down and that that measure creates a legal framework for limiting the discharge of NO_x emissions in a profitable manner for undertakings with large facilities, show that the undertakings covered by the measure in question have an alternative to the imposition of a fine by the State.

109.

With regard to the arguments concerning PreussenElektra, the General Court correctly distinguished that judgment from the present case. According to that judgment, legislation of a Member State which, first, requires private electricity supply undertakings to purchase electricity produced in their area of supply from renewable energy sources at minimum prices higher than the real economic value of that type of electricity, and, second, allocates the financial burden arising from that obligation amongst those electricity supply undertakings and upstream private electricity network operators, does not constitute State aid within the meaning of Article 87(1) of the Treaty.

110.

The Court considered, in that judgment, that even if the financial burden arising from the obligation to purchase at minimum prices was likely to have negative repercussions on the economic results of the undertakings subject to that obligation and entail a diminution in tax receipts for the State, that consequence was an inherent feature of such a legislative provision and could not be regarded as constituting a means of granting to producers of electricity from renewable energy sources a particular advantage at the expense of the State.

111.

On the contrary, in the present case, as the Advocate General stated in point 92 of his Opinion, that foregoing of resources cannot be considered as ‘inherent’ in any instrument designed to regulate emissions of atmospheric pollutants by an emission allowance trading scheme. Where it has recourse to those instruments, the State has in principle a choice between allocating those allowances free of charge or selling or auctioning them. Furthermore, in the present case there is a sufficiently direct connection between the measure in question and the loss of revenue, a link which did not exist between the imposition of the obligation to purchase and the possible diminution in tax receipts at issue in the case which led to the judgment in PreussenElektra. The facts of the two cases are therefore not comparable and the solution adopted by the Court in PreussenElektra thus cannot be applied to the present case”.

These two authorities (themselves relying upon earlier authorities) establish a series of propositions:

How do these principles apply in the present case? I accept, in broad terms, the Defendant’s submissions summarised at paragraphs [295] – [298] above.

First, there is no revenue or resource foregone to the State and hence there is no aid “through state resources”. The Intervener’s argument is that because there is an advantage it is to be treated as a gift or a gratis advantage which has been wholly unearned and unwarranted. It flows from the State and could very easily have been negatived by the State charging the benefited class for that advantage. Such a charging scheme could have been introduced under Article 5(2)(b) or the European Communities Act 1972. As such because the advantage has not been charged for it reflects revenue foregone.

However, in my view, this is exactly the sort of hypothetical, remote and indirect diminution in revenue that the Court of Justice has been keen to avoid becoming ensnared within the State aid rules. I reject the submission that there is revenue foregone because the State could have imposed a levy in order to neutralise the advantage, or because the State is said to be “liable” for breach of Article 5(2)(b) and in consequence, the liability represents “…a clear and concrete risk” to the budget. These are not, in my view, the sorts of linkages which the Court of Justice would view as remotely sufficient. The Intervener sought to draw an analogy with the facts of Netherlands (Emissions Trading Scheme) in which the Court held that there was a revenue foregone and hence “aid”. But in truth the facts of the present case could hardly be more different. In that case the Court found as fact that the disputed emissions trading scheme in the Netherlands resulted in revenue foregone because the state had instituted a formal market in emission credits specifically conferring “allowances” which the Court felt able to categorise as “intangible assets” (see paragraphs [102]), not least because it was a specific object of the emission credit scheme that these assets should have two valuable characteristics. First, they were capable of being traded in a formal market (cf. paragraphs 88ff) and were intended thereby to have an “economic value (cf. paragraph [106]). Secondly, these credits were the deliberate act of the State in establishing the dynamic cap scheme which enabled undertakings to “avoid the payment of fines” i.e. they served as fiscal or penal exemptions. There was thus revenue foregone because the State gave away for no consideration valuable, tradeable, assets which were capable of being used to offset financial charges which would otherwise have accrued to the State. The present facts do not even come close to exhibiting the closeness and formality of the connection between the advantage and foregoing of State resource which arose in the Netherlands case. The other side of the coin (relied upon by the Secretary of State) is PreussenElektra where the linkage between the advantage and the debit to the State budget was held to be too indirect and remote to amount to an advantage through state resources. In my judgment comparing and contrasting the facts of Netherlands (Emissions Trading Scheme) with the facts of PreussenElektra enables a clear line to be drawn between the sorts of advantage conferred by the state which do and do not amount to revenue foregone and an advantage conferred through State resources. In the present case the alleged aid is no more than an estimation of a diffuse advantage spread across a wide class with no clear shape or definition or formality to it. The facts of the present case fall some considerable distance away from the Article 107(1) TFEU line and do not have the characteristics of anything that could be said to amount to revenue foregone by the State.

A more recent illustration where the Court, in order to avoid a finding of “aid”, also endeavoured to distinguish, on the facts, a national scheme from those in Netherlands (Emissions Trading Scheme), was Case C-518/13 Eventech Ltd v London Borough of Camden, and, Transport for London (14^th January 2015) (“Eventech”) at paragraphs [28] ff.

It seems generally from all of these cases that the Court has little appetite for classifying the benefits which flow from generalised legislative schemes as “aid” falling under Article 107(1) unless there is a clear and direct nexus of a relatively formal character between the advantage and the foregoing of revenue. In Eventech at paragraph [34] the Court stated of the concept of “aid”: “… for the purposes of determining the existence of State aid, it is necessary to establish a sufficiently direct link between, on the one hand, the advantage given to the beneficiary and, on the other, a reduction of the State budget or a sufficiently concrete economic risk of burdens on that budget”.

In these cases there is hence an all important analysis that has to be undertaken of the strength or directness of the nexus between the alleged advantage and the state budget. That nexus has to meet a requisite level of certainty and formality. In the present case the alleged advantage comes nowhere near to meeting the requisite test.

Secondly, and even if the advantage could be said to involve a budgetary debit, the alleged benefit conferred by the introduction of the copyright exception is “inherent” in section 28B; it is merely an incidental consequence of the introduction of a measure designed to meet other legitimate aims and objectives. It is on a par with the sorts of benefits and burdens that much legislation in the commercial or consumer sectors, when introduced, results in. If the Intervener’s case was correct then many pieces of legislation would be capable of amounting to an “aid” simply because they conferred a collateral benefit upon a definable category of natural or legal persons who were in business. In my view the present case is – again – exactly the sort of case that the Court of Justice has become astute of the need to take outside of the State “aid” rules and which it has done through the rule that benefits which are “inherent” in general legislative measures do not create “aid” simply because the benefit might in some loose way be described as revenue foregone by the State.

In my judgment: (a) the alleged link between the advantage conferred and the foregoing of revenues on the part of the State is far too remote, indirect and informal for it to amount to aid through State resources; and/or (b) in any event such advantage as does arise is no more than is inherent in the modification to copyright law brought about by section 28B which is a general legislative measure designed to achieve other policy objectives and for this reason also is not “aid”.

Accordingly, the application under Article 107 TFEU fails.

For the reasons set out in this judgment at Section I this application for judicial review succeeds.

I will hear submissions as to what flows from this conclusion and from the judgment generally. In particular I will hear submissions as to whether any issue of law that I have decided should be referred to the Court of Justice and if so as to the question(s) that should be asked.

I will also hear submissions as to the appropriate relief that I should grant given (i) the nature of my conclusion as to why the Defendant’s decision to introduce Section 28B was unlawful; and (ii) any decision that I might make as to reference.