Federal Court of Appeal: Netflix Has the Right to be Heard Before Copyright Board

On December 17, 2015, the Federal Court of Appeal released its decision in Netflix, Inc. v. SOCAN in which it granted an application for judicial review from the Copyright Board's decision dealing with a Tariff for musical compositions in audiovisual webcasts. Netflix opposed the section of the Tariff that imposed a monthly minimum fee on free trial subscriptions. The Court found that the Copyright Board breached its duty of procedural fairness with regard to Netflix in disallowing it from participating in the tariff certification proceedings.

Background

The Society of Composers, Authors and Music Publishers of Canada (SOCAN) began filing royalty statements for use of musical compositions in audiovisual works transmitted over the internet in 2007. In 2011, the Copyright Board opened the tariff certification process; Netflix did not participate at this initial phase. The Board suspended the proceedings in order to facilitate settlement negotiations between SOCAN and a group of industry stakeholders. Netflix again abstained from engaging and did not take part in the negotiations. 

Image by Renjith Krishnan

SOCAN and the other parties to the negotiation reached an agreement and SOCAN filed a joint request on behalf of all parties to have the Tariff certified in accordance with their agreement. However, the Tariff's content had changed from the version initially published in the Canada Gazette. The Tariff’s first iteration made no distinction between subscription-based and on-demand services; it also did not provide for compensation in relation to free trial subscriptions. The new version of the Tariff addressed both.

Netflix's interest in the Tariff fundamentally changed at this point. Unlike the signatories to the settlement agreement, Netflix's offerings are subscription-based and it is well known for its one-month free trial. The agreement, to which Netflix was not a party, therefore became a major concern, spurring Netflix into action before the Copyright Board. 

Since Netflix did not take part in the process from the beginning, the Board did not invite it to make submissions on the Tariff as proposed following the agreement; Netflix did so anyway. It argued that the free trials were fair dealing in light of the Supreme Court of Canada's recent case law relating to technological neutrality and consumer research fair dealing (ESA v. SOCAN and SOCAN v. Bell respectively). 

The Board refused to make Netflix’s submissions part of the record. The Board's decision to exclude Netflix was based on: 1) the fact that Netflix raised issues that could not adequately be addressed with the record as it stood and would require additional evidence; and 2) the fact that none of the parties to the proceedings had raised those issues prior. The Board also noted that Netflix had every opportunity to get involved in the proceedings from the beginning and failed to do so.

Despite this decision, the Board chose to commence a new process based upon the Tariff resulting from the settlement discussions, citing exceptional circumstances and the fact that Netflix is such a major player in this space. The Board nonetheless required parties to make submissions based on the record as it stood and without introducing new evidence that did not consist of “noncontroversial facts which shed significant light on the proper course of action". Netflix made its fair dealing argument and requested leave to lead new evidence. SOCAN opposed and the Board refused to allow Netflix's request. The Board ultimately certified the Tariff, more or less as proposed by SOCAN.

Appeal

Writing for the Court, Justice Nadon (Justices Boivin and De Montigny concurring) held that the Copyright Board erred in failing to allow Netflix to make its submissions. After briefly noting that the applicable standard of review was correctness, the Court held that the Board failed to discharge its duty of procedural fairness vis-à-vis Netflix. 

The thrust of the Court’s reasons was that Netflix never had a chance to participate in the certification process relating to the Tariff as approved by the Board. SOCAN argued that Netflix abstained from participating because it chose to rely on other objectors to field its concerns. The Court held that regardless of whether this was the case, even if Netflix had participated at the pre-settlement negotiation phase, its submissions would have been in relation to a proposed tariff that was materially different (especially from Netflix’s perspective) from the one ultimately certified. In that respect, Netflix’s failure to participate from the get go could not serve was a waiver of its right to participate once the Tariff, and stakes, had changed significantly. 

Justice Nadon observed that since the tariff-setting process is one that concerns the industry as a whole, it is that “industry interest”, and not only Netflix’s individual interest, that must be safeguarded. He found this to be a relevant factor in deciding as to whether the Board met its procedural fairness obligations. 

He also took issue with the Board’s assessment of the “ReSound 5 factors”, elements the Board should consider when deciding whether to certify a tariff based on an agreement between the petitioning copyright collective and interested stakeholders. In Re:Sound 5 the Board identified two factors that ought to be evaluated in such a situation: 1) the extent to which the parties to the agreements can represent the interests of all prospective users; and 2) whether relevant comments or arguments made by former parties and non-parties have been addressed.

In Justice Nadon’s opinion, the Board was wrong to conclude that a negotiation process which excluded Netflix, the largest player in the space, adequately represented all perspective users’ interests. This is especially true in this case where the provisions in question were seemingly targeted at Netflix, and by extension its user-base. 

Justice Nadon also found that the Board failed to properly assess the second Re:sound factor in that it chose to ignore Netflix’s arguments on fair dealing. Since this factor contemplates non-parties, Netflix’s lack of participation in the proceedings is no excuse for failing to consider its arguments.

The Court also failed to accept the Board’s reasoning that ruling on Netflix’s submissions would have required it to consider evidence that was not on the record. Justice Nadon noted that while this was true, the only reason why the required evidence was not on the record was because the Board failed to admit it. The Board is therefore precluded from relying on a lack of evidence caused by its failure to accept same. 

Nadon J.A. concluded by stating that though administrative tribunals do and should exercise a broad discretion over their procedural rules, the end goal of those rules are to assure that justice is done. In his view, failing to allow Netflix to advance its arguments ran counter to the interest of justice. Denying their right on procedural grounds is therefore inconsistent with the very goal that procedure is meant to encourage. 

Conclusion

This was a tight and well-reasoned decision. From a practical perspective, the Board’s decision did not pass the smell test. The proposed Tariff as published in the Canada Gazette made no mention of subscriber-based services and free trial subscriptions. The parties to the settlement negotiations did not offer subscription-based services or free trials; they were therefore effectively negotiating with Netflix’s chips. To allow this to stand would not only deny Netflix the opportunity to defend its interests, it would have allowed Netflix’s competitors to benefit from a more favourable compensation regime while stifling their biggest competitor’s business model. 

Due to the decision’s proximity to the holiday season, we will have to wait until the New Year to find out whether SOCAN will ask for leave to appeal to the Supreme Court of Canada.

Copyright Board Approves SOCAN Tariffs For Performance Of Musical Works Contained In Audiovisual And User-Generated Media Transmitted Over The Internet

The Society of Composers, Authors and Music Publishers of Canada (SOCAN) proposed two tariffs (Tariff 22.D.1 and 22.D.2) for the collection of royalties for musical works performed as part of audiovisual works transmitted over the internet. Several stakeholders including Facebook, Netflix and Bell objected and made submissions in opposition to the proposed Tariffs. On July 18, 2014, the Board rendered its Decision allowing the SOCAN Tariffs. 

Image by Renjith Krishnan

The royalty rates set in these Tariffs are based on an agreement reached during settlement negotiations between SOCAN and some of the original objectors. The Board addressed the setting of general tariffs based on agreements reached by copyright collectives and users in its decision in Re:Sound Tariff 5. That Tariff deals with the public performance of sound recordings. Before allowing an agreement to form the basis for a tariff, the Board said that it is “generally advisable” to consider: 

(a) The extent to which the parties to the agreements can represent the interests of all prospective users; and 

(b) Whether relevant comments or arguments made by former parties and non-parties have been addressed.

Since the Settlement Agreements here involved some of the biggest players, including Apple, Bell, Rogers and Cineplex, the Board found that the parties to the Agreement were sufficiently representative of the interests of other users. It also found that those objectors who were not party to the Agreement had ample opportunity to make their submissions. The Board also noted that the rates fixed in the Agreement were the fruit of extensive negotiations between skilled and savvy counsel.

Several grounds were raised by the objectors. Among them, Facebook and Netflix’s arguments bear some discussion. 

Facebook argued that it has a strict copyright enforcement policy and has implemented software to make sure that any protected uses of copyrighted works are prevented from being uploaded to its network. Facebook noted that any use of copyright protected works not caught by its software would qualify under the new UGC exception found at s.29.21 of the Copyright Act. 

Facebook also argued that an “audiovisual page impression” (defined in the Tariffs as “a page impression that allows a person to hear an audiovisual work”) should only give rise to a royalty when the work in question is actually viewed or heard. The wording of the Tariffs make it so that once a web page that contains the audiovisual work is loaded, a royalty is payable regardless of whether the work is ultimately accessed or not. 

The Board rejected the first argument out of hand. Whether Facebook’s activities would fall under the Tariffs is irrelevant. The Copyright Board sets tariffs of general application. If Facebook does not feel its activities fall under these Tariffs, it does not have to pay royalties until a court of competent jurisdiction says otherwise. To say that the Tariffs should not be allowed because it does not apply to Facebook is a non-sequitur.

The Board also rejected the argument on page impressions. It simply said that SOCAN’s proposed method of calculation was acceptable and consistent with other Tariff 22 classes. 

Netflix argued that its one-month free trials should not be captured by the Tariffs. It based this position on two arguments:

1) The free trial is fair dealing for the purpose of research along the lines of the Supreme Court of Canada’s decision in SOCAN v. Bell (discussed here). In that case, the issue before the Court was whether the sampling of 30-90 second excerpts of musical works constituted fair dealing for the purpose of consumer research.

2) Paying SOCAN for free trials of the service would cause double compensation contrary to the principle of technological neutrality as set out by the Supreme Court in ESA v. SOCAN (discussed here).

On the fair dealing argument, the Board found the analogy between Netflix’s free trial period and the song previews in SOCAN v. Bell to be tenuous. There is a big difference between a low quality sample of a portion of a song and a high-quality, full version of a television show or movie. The Board also declined to look any further into fair dealing as no one (including Netflix) led sufficient evidence to make such a determination. 

The Board was equally unconvinced by Netflix’s technological neutrality argument. It reasoned that there is no technological alternative to Netflix’s free trial period that is or has been used in the offline world: “There is no alternative-technology equivalent to a Netflix free trial. Video stores never offered a free month’s membership with the right to rent as many videos as the customer wanted for no additional charge. Thus, there is no issue with technological neutrality.”

Thoughts

Overall the Board’s decision appears to be reasonable. Its rejection of Facebook’s argument regarding page impressions strikes me as a little dubious. Is there some technological reason why royalties should be calculated this way? Surely it would be more accurate to track the number of clicks received by the media player on the page than the raw number of times that the page is loaded. Some dynamic webpages have auto-refresh features that refresh parts of the page while leaving other parts untouched. This may skew results in favour of more compensation. Still, this outcome seems far from unreasonable. 

Netflix’s argument based on technological neutrality was puzzling. It asserted that allowing SOCAN to collect royalties for free trials would lead to “double dipping” (presumably because Netflix would have to pay a different royalty rate if the free user eventually subscribed). This seems a little simplistic. Surely Netflix is capable of –and certainly is- monitoring which free users eventually subscribe to its service. What is so difficult about subtracting the free trial royalty rate from the higher subscriber royalty rate and paying SOCAN that amount? 

To take a basic example, assume that the free trial royalty rate is $12/year (or $1 for a one-month free trial period) and that the regular royalty rate is 5% of the subscription fee which is $25/month. This would produce an annual royalty of $15 per subscriber.  For a subscriber who takes advantage of the free trial month and then signs up and pays the subscription fee for the rest of the year, all you have to do is subtract the free trial month amount ($1) from the annual royalty ($15) in order to assure that SOCAN is not paid twice for the same period of time (once at the free trial rate and once at the full rate).

While Netflix’s argument was ill-founded, I think the Board may have made a mistake in its reasons. Instead of rejecting Netflix’s argument for the reason set out above, the Board justified its decision by noting that in reality, video rental stores never offered free trials. It reasoned that given the lack of an analogue to Netflix’s free trials in the offline world, the latter’s argument based on the Supreme Court’s establishment of the principle of technological neutrality in ESA was moot. 

The principle of technological neutrality stands for the proposition that copyright protected works should be given the same treatment regardless of the technological medium by which they are conveyed. The Board uses the following example to illustrate: “…since only the reproduction right is triggered when a CD is sold in a store, only the reproduction right should be triggered when a digital album is sold online. The CD is an alternative technology to the digital download.”

The Board is essentially saying that technological neutrality only applies when there is an alternative technology that has been marketed to the public to compare the technology in question to. That need not necessarily be the case. There is nothing in the Supreme Court’s decision in ESA that limits the principle of technological neutrality to comparing extant (or previously extant) business models. This view cheapens the value of this interpretive principle. 

Take the facts from SOCAN v. Bell as an example. In that case, users could access previews of songs by clicking an icon on a service provider’s website. That decision has nothing to do with technological neutrality. However, for illustrative purposes, assume one of the parties wanted to make an argument for more or less compensation based on the principle of technological neutrality. They would have to come up with some analogous technology to the service provider’s website. One potential analogue to the website could be a business method by which music stores send individuals door to door with samples of music they have for sale. The sales representative would play the samples on CD’s or some other physical media; kind of like the Avon lady, but for music. 

To the best of my knowledge, this “Avon lady” sales model was never employed by record stores. This should not matter.  This method of distribution is conceivable and therefore ripe for comparison for copyright purposes. 

This hypothetical technological analogue theory has not been addressed in the case law. In ESA, the Court was comparing two modes of distribution (online digital delivery and shelf display in stores) that do in fact exist side by side. The Court in ESA did not expressly state whether the technological alternatives being compared must be ones that not only exist, but have been implemented in commerce. Given the Act’s statement of the s.3 right as the right to produce or reproduce a work “in any material form whatever”, the lack of limiting language in the Court’s decision should not rule out hypothetical technological alternatives concocted for the purpose of comparison. In that regard, the Board gave the principle of technological neutrality a more narrow interpretation than the one expressly prescribed by the Supreme Court. 

Finally, whether this decision is appealed to the Federal Court of Appeal is an interesting and open question. While it appears that the Board’s decision is reasonable, in light of the Majority of the Supreme Court’s ruling in ESA, the decision may be reviewed, at least in part, on a correctness standard.

Authors Guild v. Hathitrust: Book Digitization And Indexing Considered To Be Fair Use (And Transformative?)

Background

On June 10^th, 2014, the United States Court of Appeal for the 2^nd Circuit handed down its decision in Authors Guild v. HathiTrust, a case dealing with copyright and the digitization of print books. The Court found that Hathitrust’s digitizing books for the purpose of enabling research constitutes fair use. 

Image by Renjith Krishnan

Hathitrust is an amalgam of 80 American universities and other institutions. Its mission is to digitize and preserve the print books contained in the catalogues of its constituent members. The Hathitrust Digital Library serves three different purposes: 

1) Search functionality – The digital library lets a user search for terms contained in books included in the database. For books that are currently protected by copyright, the search engine will not display any text unless expressly authorized to do so by the copyright holder; it will merely display page numbers of books contained in the catalogue in which the search terms appear.

2) Enabling individuals with “print disabilities” – Individuals with print disabilities are able to access full text versions of the books contained in the catalogue. Print disabilities includes people with perceptual disabilities such as blindness, but also includes individuals with other physical disabilities that would prevent them from being able to turn the pages of a print book. 

3) Book replacement – The system would allow member institutions to print replacement copies of physical books in their catalogues that have been lost, damaged or stolen (and a new copy cannot be obtained at a “fair price”).

District Court

The Authors Guild and other individual and institutional Plaintiffs sued Hathitrust in the District Court for injunctive and declaratory relief. Hathitrust moved for summary judgement arguing that its digitization project constituted fair use. It also argued that allowing access to full text versions of the books by individuals with print disabilities is protected by §121 of the U.S. Copyright Act (Also known as the “Chafee amendment”). 

The relevant portion of §121 reads as follows:

(a) Notwithstanding the provisions of section 106, it is not an infringement of copyright for an authorized entity to reproduce or to distribute copies or phonorecords of a previously published, nondramatic literary work if such copies or phonorecords are reproduced or distributed in specialized formats exclusively for use by blind or other persons with disabilities.

At the District Court level, the Trial Judge found that all three of the above listed functionalities constituted fair use. The Judge found specifically that the uses were “transformative”. In Campbell v. Acuff-Rose Music, 510 U.S. 569 (1994), the United States Supreme Court determined that when the use of a copyright protected work is transformative in nature, it is more likely to be considered fair.

The Judge also determined that Hathitrust’s argument based on the Chafee amendment was sound. According to §121 authorized entity means: “a nonprofit organization or a governmental agency that has a primary mission to provide specialized services relating to training, education, or adaptive reading or information access needs of blind or other persons with disabilities”. The Trial Judge found that Hathitrust’s members could fit within that definition. 

Hathiturst also made an argument as to the standing of some of the Plaintiffs. The Court acquiesced here as well and struck certain institutional Plaintiffs, including the Writers’ Union of Canada and the Union des Écrivaines et des Écrivains Québécois, from the suit.

Appeal Decision

The Court of Appeal for the 2^nd Circuit reviewed the case de novo. They found that the Trial Judge’s decision as to the standing of certain Plaintiffs was right in law. Those Plaintiffs were not themselves the owners or exclusive licensees of the copyright in the works of their members. 17 U.S.C. §501(b) only allows owners of “an exclusive right under a copyright” to sue for infringement thereof. 

Moving to the fair use argument, the Court of Appeal agreed with the Trial Judge that all three functionalities constituted fair use. The Court went through a rather detailed description of the fair use doctrine and its position of central importance to copyright law. 

The Court then undertook the four-factor analysis set out in §107 of the Copyright Act which sets out the following considerations for determining whether a use is fair:

(1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes;

(2) the nature of the copyrighted work;

(3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and

(4) the effect of the use upon the potential market for or value of the copyrighted work.

Each of the three functionalities was submitted to this four-factor analysis. Dealing first with the search functionality, the Court found this to be a fundamentally transformative use of the work. For the use of a work to be transformative, the Court writes that there must be more than a mere repackaging or republishing. 

The Court of Appeal compares the search functionality with other instances of transformative use culled from the jurisprudence. It found that “Full‐text search adds a great deal more to the copyrighted works at issue than did the transformative uses we approved in several other cases”. One of the case pointed to was Prince v. Cariou (Discussed here), a case in which an appropriation artist used a photographer’s photos to create a collage type work. In that decision, the 2^nd Circuit found that the use was transformative, even though the artist Richard Prince was operating in the same medium as Patrick Cariou, the photographer. 

The Court of Appeal more or less brushes off the second fair use factor as being inconsequential. They found that factors three and four favoured a finding of fair use.

In addition to being covered under the Chafee amendment, the Court found that the making available of full-text versions of the works for people with print disabilities constitutes fair use. This view was expressly endorsed by the U.S. Supreme Court in Sony Corp. of America v. Universal City Studios, Inc. Though the Court of Appeal found that the second factor militated against a finding of fair use, they found that factors one, three and four cut in favour of it. 

Finally, with regards to the preservation of books, the Court of Appeal felt that the matter was improperly dealt with by the District Court. They vacated the Trial Judge’s order and remanded this issue for a technical reason – namely that the record did not show that any of the Plaintiffs owned copyright in books that have been or are currently affected by this functionality. Since the U.S. Constitution only allows Federal Courts to hear actual “cases and controversies”, the Plaintiffs did not have standing with reference to speculative claims.

Thoughts

This decision is particularly important in light of the coming decision in Authors Guild v. Google, a case dealing with Google’s own book digitization project. One important factual difference between this case and the Google Books case is that when an individual searches for text in Google, the search engine will actually display snippets of copyright protected works containing the search terms. The Hathitrust Digital Library merely supplies the page numbers of the books in which the search terms may be found. This, however, seems far from dispositive given the Court’s emphasis on the transformative nature of full-text searchability of books. 

On that point, I have to disagree with the 2^nd Circuit’s reasoning. While enabling the full-text search of entire catalogues of books is certainly innovative and useful, I am not sure it is transformative. When one thinks of a transformative work, one thinks of a remixing or mashing-up of existing content to create substantially new content. In the Cariou case cited by the Court of Appeal, the artist Richard Prince took photographs taken by Cariou and altered them in a visually obvious manner by superimposing other images onto them and adding visual effects. 

When Hathitrust (or Google) digitizes a book, the substance of the book remains exactly the same. While there is a format-shift taking place, there is no transformation of the work on a substantive level. Furthermore, even if there is a transformation, it is not being carried out by the people ultimately making use of the copyright protected work. 

While it carries no weight in the U.S., the Supreme Court of Canada’s decision in Alberta (Education) v. Access Copyright found that the relevant perspective for determining whether a dealing is fair is that of the end user of the copyrighted work. That case dealt with the ability of teachers in Alberta to photocopy portions of literary works in order to give their students reading and research projects. Access Copyright argued that because teachers made several copies of copyright protected works (one for each student), the dealing was less fair. The Court disagreed. It wrote that the relevant perspective was that of the student (the individual engaging in the protected activity, private study) and not the teacher who was merely acting as an intermediary. 

Applied to the current case, the 2^nd Circuit asked and answered the wrong question. Instead of deciding whether Hathitrust’s digitizing is a transformative use of the copyrighted content (Which I think it is not), they should have instead asked if users interacting with the books in this novel way –i.e. searching the full-text of the book in order to find specific terms- is a transformative use of the book (which I think it is). I am therefore of the opinion that the 2^nd Circuit reached the right result for the wrong reasons; and they did so employing an ill-fitting analysis.