National Writers Union 2

National Writers Union
256 W. 38th St., Suite 703 New York,
NY 10018

www.nwu.org

March 5, 2021

Senator Thom Tillis, Ranking Member
U.S. Senate Judiciary Subcommittee on Intellectual Property 113 Dirksen Senate Office Building
Washington, DC 20510

Senator Patrick Leahy, Chair
U.S. Senate Judiciary Subcommittee on Intellectual Property
437 Russell Senate Office Building

Washington, DC 20510

Re: Senator Tillis's discussion draft, Digital Copyright Act (DCA) of 2021^{Footnote 1}

The National Writers Union welcomes the opportunity to submit these comments on the discussion draft published by Senator Tillis of the Digital Copyright Act (DCA) of 2021. We would also welcome an opportunity to meet with Senator Tillis, Senator Leahy, other members of the Subcommittee, and/or members of their staff concerning the issues raised in these comments.

The National Writers Union (NWU) is an independent national labor union that advocates for freelance and contract writers. The NWU includes local chapters as well as at-large members nationwide and abroad. It works to advance the economic conditions of writers in all genres, media, and formats. NWU membership includes, among others, book authors, journalists, business and technical writers, website and e-mail newsletter content providers, bloggers, poets, playwrights, editors, academic writers, and other media workers.

These comments address Sections 3 (orphan works) and Section 20 (group registration) of the discussion draft, as well as three other issues important to our members and other writers which are not addressed in the draft but which we believe should be included in any omnibus copyright reform bill: reform of 17 U.S. Code § 203 on termination of licenses or assignments of rights, moral rights of authors of written works, and a Public Lending Right (PLR).

Our primary concerns with the discussion draft pertain to Section 3, "Limitation on remedies in cases involving orphan works." If the discussion draft or other legislation containing these or similar provisions is introduced and considered by the Subcommittee, we request an opportunity to testify at hearings on any such bill in opposition to this proposed section and on behalf of authors of works which are likely to be deemed "orphan works."

The fundamental problem with the orphan works provisions of the discussion draft is that the definition of an orphan work includes many works which are being actively exploited and generating revenues for their authors in ways that would be undermined by authorizing "usage" (copying and/or republication) in the manner provided in the draft bill.

This is not a minor defect that can be remedied by tweaking the definition of an orphan work or the requirements for a "diligent search" for rightsholders. The definition of an orphan work and the proposals to authorize exceptions to copyright for such works are based on a fundamental misconception about how we authors earn money from our work.

Those who want statutory authorization to copy or republish orphan works without permission, and in certain cases without remuneration to the authors, premise their entire analysis of orphan works on the assumption that if a work is being commercially exploited, it will be possible for anyone who wishes to republish or copy the work to identify and contact the holder of the rights that they wish to exercise. This has been taken for granted, but it's not true.

As discussed in more detail below and as we will testify, if given a chance, at any hearings on this portion of the bill, many written and graphic works are being commercially exploited in ways that do not make it possible for would-be users who want to republish or copy those works to identify or contact the holders of the rights that those users want to exploit.

These modes of exploitation and the revenue streams from them are, increasingly, commercially significant. Some writers make all or most of their income from orphan works.

Not all authors wish to make ourselves identifiable or findable. Some writers have good reasons not to want to be identified or found. Some writers are bound by contractual commitments that prohibit them from identifying who holds certain rights to their works.

Writers are not, should not be, and cannot lawfully be – pursuant to U.S. international treaty obligations – required to identify ourselves or make ourselves findable as a condition of copyright protection or of the ability to obtain redress for infringement of our copyrights.

It is understandable that would-be republishers of our work, including librarians and archivists, are unaware of many of our business models and revenue streams. Many of our business models and revenue streams don't involve libraries and archives. There is no reason to expect librarians or archivists to be familiar with the ways writers make our living.

It is unfortunate, however, that those – primarily some librarians and archivists – who have coined the "orphan works" meme have persisted in their false claim that the works that they would define as orphan works are (necessarily or by definition) not being exploited, despite that claim having been contested and disproved by authors of orphan works for many years.

Nothing in the definition of an orphan work excludes works that are being actively exploited. The claim that an orphan works exception to copyright or to remedies for copyright infringement will not harm writers who are actively exploiting their work depends entirely on the unsupportable premise that exploitation necessarily makes the rightsholder findable.

The orphan works provisions of the discussion draft would harm writers, especially those of us who are already marginalized, vulnerable, or threatened with harassment or retaliation by those who dislike our work. Less privileged writers who can't afford to give their work away, and depend on income from their writing to be able to write, would be further disadvantaged.

When someone claims, "We're not hurting you," and the victims say, "Yes, you are hurting us," we think the claim to be doing no harm should be received with great skepticism.

The orphan works provisions of the discussion draft would (1) interfere with increasingly significant modes of exploitation of written work and revenue streams for writers, including those that allow us to maintain our anonymity and protect ourselves from harassment or retaliation; (2) close off some business models and force writers into others not of our choosing (or out of the business of writing), thwarting innovation and further harming both writers and readers; and (3) as applied to works by non-U.S. authors, violate U.S. obligations pursuant to the Berne Convention (with respect to formalities and permissible exceptions) and the WIPO Copyright Treaty (with respect to effective remedies for all cases of infringement).

The orphan works provisions in the discussion draft should be stricken entirely.

1. Rights to many orphan works are being actively exploited.

To understand how orphan works are being exploited and generating revenue for their authors, it's first necessary to understand what works would be deemed orphan works – or, more accurately since rights are often divided, which rights would be deemed "orphan rights."

Particular rights would be deemed "orphaned" if, after a would-be infringer performs a "diligent search," the infringer is unable to locate and identify the holder of those rights.

It should be obvious from this definition that any work that is published anonymously – especially work that is self-published anonymously, so that there is no publisher to serve as an identifiable and contactable anonymizing proxy for the author – is by definition an orphan work.

This proposal, like all orphan works legislation, must be understood as a frontal assault on authors' privacy, on the possibility of making a living from anonymous writing, and on the writer who self-publishes her work anonymously is trying to make it difficult or impossible entirety of the large, growing, and valuable industry of anonymous self-publishing.

A writer who self-publishes her work anonymously is trying to make it difficult or impossible for readers or the public to identify or contact her. There are many good reasons she may choose to do so. But to the extent she succeeds, her works would be orphan works.

Publishing or self-publishing anonymously does not mean giving up the possibility of revenue. Exploiting works commercially, particularly in advertising-supported digital business models for content delivery including web content, does not require rightsholders to engage in transactions with readers, users, or licensees, or to offer works for sale or licensing.

Advertising-supported digital self-publication, for example, is attractive to, and engaged in by, some authors precisely because it enables them to exploit their works commercially without having to disclose their identities or engage in transactions with purchasers or licensees. Only the advertiser or ad broker needs to know where or to whom to send payments, not the reader. And online payment platforms allow payments for downloads or subscriptions to written content to be made and received anonymously, even when there are transactions with readers.

Anonymous digital self-publication of orphan works takes many forms. An author can take subscription payments by Paypal to an anonymous e-email address for her "story of the week" e-mail newsletter, or for downloads from her website of ebooks, articles, or stories. She can self-publish anonymous fiction or non-fiction ebooks distributed through other platforms. Or she can publish her orphan works on an anonymous website that generates advertising revenue. She may or may not wish to be contacted, especially if most of the messages she receives at any address she makes available on her website are hate mail, harassment, threats, spam, or scams.

The emerging and growing industry of digital self-publication and the greater possibility of anonymous commercial publication it affords to authors – who generally needed a publisher as an anonymizing proxy for commercial publishing in print formats – has been a special boon to marginalized, vulnerable, and underprivileged writers, and to readers of their work.

Readers benefit from the availability of these works and the ability of these writers to make a living from their work, which enables them to devote more of their time to writing. We are disappointed that librarians, who are such strong defenders (along with writers) of readers' privacy, have been so insensitive to the implications of their lack of concern for writers' privacy.

Anonymous self-publishers – which is to say, by definition, writers of "orphan works" – include whistleblowers, leakers, writers on controversial or stigmatized topics, and writers who fear harassment or retaliation if they are "outed" or can be identified or located. Such writers may be making their best efforts to ensure that their identities cannot be found by even the most diligent search, at the same time that they are actively exploiting their anonymous work and, if successful, earning a living (or at least earning some income) from that work.

A writer can publish an anonymous family blog, obscuring details of their family and perhaps even what city or state where they live in. By writing and publishing anonymously, they can explore issues that they might not be able to address if they were identified, and do so without invading their family's privacy. They may blog or publish articles or stories about workplace issues, without naming their employer or co-workers. Or they may write and publish anonymous orphan works about their work as a doctor, therapist, or other health care worker or first responder. Many blogs like this are anonymous, for good reason. It is valuable for the public to understand these perspectives and hear these stories, which couldn't be told by name without invading the privacy not just of the authors but of the other people who appear in their stories.

Pursuant to the proposed legislation, a nonprofit organization could copy these works in their entirety and reproduce them, on the World Wide Web or otherwise, without permission or payment. Since the website of a library or archive is likely to rank higher in search engine results than an author's self-published site, the infringer is likely to divert most of the traffic of web visitors that the author's site would otherwise have received, and deprive the author of most of her revenue. On the web, clickstream diversion can deprive rightsholders of revenue they would otherwise have received, even if the infringer is not monetizing the diverted web traffic.

To try to abate this undermining of her revenues, the author would have to take time away from her writing to search the database of "Notices of Use" at frequent intervals for each of the titles (if they have titles, which web pages often don't) of her works. This is an extraordinary and unfair burden to impose on working writers. And to file a "Notice of Claim" the author would have to "out" herself and disclose her address and phone number.

The discussion draft would require a rightsholder claiming infringement to provide an address and telephone number and to send a written "Notice of Claim" to the infringer. But ironically, an infringer would be required to provide only a name, and no contact information, in a "Notice of Use." In many cases, even if a rightsholder learned of an infringing use from a "Notice of Use" the rightsholder wouldn't have enough information to be able to serve a written "Notice of Claim" on the infringer, especially if the infringer has a common name. Infringers wishing to avoid service of a "Notice of Claim" could simply hire an agent with an untraceably common name to act on their behalf and file the "Notice of Use" in the agent's name.

Another major category of actively-exploited, revenue-generating orphan works (or again, more accurately, "orphan rights") are rights that have been assigned pursuant to a contract that contains a non-disclosure clause prohibiting the author from telling anyone to whom she has assigned those rights. In such a case, an infringer who asks the author who holds certain rights to a work may be told that those rights have been assigned to someone other than the author, but that the author is not allowed to say to whom she has assigned those rights.

This sort of nondisclosure clause is increasingly common, including in some especially lucrative types of purchases or assignments of rights. For example, the purchaser of the exclusive right to use a particular written work as the basis for a film or video production or an electronic game may not want rival production companies to know what concept it is working on.

Any number of infringers could obtain the same right to use such an orphan work by paying the same license fee, and could continue that infringement indefinitely once started, as long as the work was combined with other original material (as such an adaptation of a written work to another medium likely would be). The orphan works law would thus effectively destroy the value of the exclusivity of the rights assigned in the original contract.

Even when the author is neither seeking anonymity nor prohibited by contract from identifying the assignee of a right, it is often impossible to identify and locate the author or other holder of a particular exclusive right. Many actively exploited rights are orphan rights.

The proposal contemplates that a "diligent search" for rightsholders would be conducted through sources such as "references" and "databases."

It's understandable that librarians (erroneously) think of their bibliographic databases and references as the definitive sources of information about printed works. But there is no information at all about rights holdings in these databases. They indicate authorship, but rarely contain any author contact information. They do not indicate which rights were assigned, which rights have reverted, and which rights are retained by the author. These sources may indicate contact information for the publisher of one of more editions of the work. But even if the publisher is still in business, publishing a print edition is not evidence of having ever held rights to reproduction in digital formats, or of still holding rights to print publication.

Publication data, even if current, is not data about rights holdings.

Except for the very small minority of published written works that are offered for licensing through a collective licensing agency or "reproduction rights organization," there is no information in any database about who holds which rights to most written works. The only way to find out who holds which rights is to start by identifying, locating, and contacting the author, and then – if they choose to tell you, which they are not obligated to do – tracing the chain of assignments of rights and their current validity, starting with the creator as original rightsholder.

Bibliographies list publishers, but that may not help to find rightsholders. A publisher or former publisher (if they are still in business) may choose to refer rights inquiries to the author, if they still know how to contact the author. But they have no obligation to do so. Once a book is out of print and no more royalties need to be paid, and rights have reverted, a publisher has no reason to keep in touch with the author or refer inquiries to them, especially if the author and publisher parted company in disagreement or conflict.

It's not clear how an infringer could possibly form any opinion "with a reasonable degree of certainty" as to who holds any particular rights to a work, without having consulted the (suspected) author not only to confirm the attribution of authorship but also to find out about current assignments of rights. Contracts assigning rights are private contracts. Even when they don't contain nondisclosure clauses, the parties are under no obligation to disclose their terms.

There is, of course, no database of websites, web publishers, or works published online. A domain name registration must include a name and some contact information. But the domain name owner is not necessarily or typically the owner of copyright in all works published on websites in that domain, and doesn't necessarily know who is. Because (a) many domain name registrants prefer to remain anonymous, and (b) most phone calls, letters, and e-mail messages received at contact information made public in domain name registrations are spam or scams, most domain name registrations include contact information only for the registrar, and neither the name nor any direct contact information for the domain name owner.

Copyright registration is not required for copyright protection, and cannot be required of non-U.S. authors pursuant to the Berne Convention prohibition on copyright formalities. And copyright registration records are of little use in identifying who currently holds which rights to a work, even when copyright in that work was once registered.

Most of the written work being published today is published online, not in print. Registration of copyright in web content, as the NWU has repeatedly pointed out to the U.S. Copyright Office, is prohibitively expensive and time-consuming. In response to a petition for rulemaking by the NWU and other organizations of writers and authors, the Copyright Office recently established a procedure for group registration of multiple short works published online. The Copyright Office made explicit that this procedure is not to be used for registration of websites. But the Copyright Office has established no other procedure for registration of web content. The new group registration procedure is limited to 50 works at a time, while a writer's website may contain hundreds or thousands of pages. Virtually no web content is registered with the Copyright Office, even when it is being commercially exploited and the copyright owners want it protected. Web content is not registered because registration is too expensive and time- consuming, not because these works have been "abandoned" or are not being exploited.

Section 20 of the discussion draft of the Digital Copyright Act would expand the possibilities for group registration of multiple works, but would still be limited to works "first published as contributions to periodicals." Websites and blogs are not "periodicals," so this would still leave authors of web content without any affordable way to register their work.

We welcome the proposal in Section 20 of the discussion draft, but we strongly urge that this section be amended to strike the words, "first published as contributions to periodicals."

To avoid ambiguity as to whether or when a website or portion of a website would be deemed to constitute an "online publication," we also recommend that the deposit requirement for these groups of works in subsection (4)(A)(i) be amended to add the word "website", so that it would refer to "each entire section in the case of a newspaper, online publication, or website."

Even when copyright in a work has been registered, few assignments or transfers of rights are recorded — not surprisingly, since recording a transfer or assignment of rights is substantially more expensive and complicated than registering copyright in the first place. As with initial registration, lack of registration of transfers or assignments of rights does not indicate that those rights have been abandoned or are not being exploited.

2. Orphan works legislation would impose choices of business models and thwart innovation.

Reshaping the markets for written work, especially in digital formats, may be a goal for some proponents of orphan works legislation, but it is a misguided and inappropriate goal.

There is a diverse and changing variety of ways that writers can monetize their copyrights, especially in digital formats. ^{Footnote 2} Many of those modes of exploitation of written works do not require transactions between readers and rightsholders, and do not require authors or rightsholders to identify themselves or make their contact information public. That is a good thing for readers and writers alike, and contributes to the volume and diversity of available writing.

Copyright legislation should encourage diversity and innovation, not try to force digital or print publishing into legacy business models determined by which publication formats and distribution modes are amenable to copyright registration on the one hand, or would result in successfully exploited works automatically being deemed "orphan works" on the other.

For better or worse, advertising, not sales or licensing, has become the primary source of revenues for written work. There are many legitimate concerns with the ad-targeting industry. But defining works monetized through advertising (which does not require the author to identify herself or provide her contact information to readers) rather than through sales or licensing as orphan works will only undermine authors' livelihoods, not reform the advertising industry.

Orphan works legislation should not be used as a back-door mechanism to impose more burdensome copyright formalities on writers, to discourage writers from trying to make a living from their work, to discourage or discriminate against writers who can't afford to write for free, or to expropriate writers' copyrights to their personal backlists – the only thing other than Social Security most freelance writers have to live on in old age – for little or no compensation.

3. Orphan works legislation would violate the Berne Convention and the WIPO Copyright Treaty (WCT).

As discussed above, the NWU believes that the orphan works provisions in the discussion draft would be bad policy. As applied to non-U.S. works, they would also be contrary to U.S. obligations pursuant to international copyright treaties to which the U.S. is a party.

While the membership of the NWU consists primarily of U.S. citizens and/or residents, we are mindful of our counterparts around the world whose work is published in the U.S., especially online. Because of the dominance of the web hosting industry by U.S. companies, writers from around the world publish their work on web servers in the U.S. or on multinational "cloud" platforms with some of their mirror servers in the U.S.

With respect to works by non-U.S. authors and works first published in other countries, exceptions or limitations to copyright, such as those for orphan works proposed in the discussion draft, are permissible under Article 9 of the Berne Convention only "provided that such reproduction does not conflict with a normal exploitation of the work and does not unreasonably prejudice the legitimate interests of the author."

Conflict with "normal exploitation" is part of the so-called "three-step test" for compatibility of national legislation or regulations with the Berne Convention.

Any inquiry into the compatibility of orphan works legislation with the Berne Convention must therefore begin with an inquiry into the normal modes of exploitation of the works which would be deemed "orphan works" under the proposed legislation.

Unfortunately, such an inquiry has yet to be conducted. Proponents of orphan works legislation have been willfully deaf to the protests of authors of orphan works and have persisted in pretending that inability to find the rightsholder is evidence of non-exploitation. As discussed above, this assumption is unwarranted, and often simply wrong.

That leaves it to Congress to conduct its own fact-finding concerning the normal modes of exploitation of orphan works. As an organization some of whose members earn all or significant parts of their livelihood from exploitation of orphan works, the NWU would be happy to consult with you and/or your staff and to provide testimony at hearings on this issue.

We believe that, once any serious consideration is given to the types of works that would be deemed orphan works and how they are now being exploited, it will be clear that the provisions in the discussion draft would conflict with normal exploitation of these works and unreasonably prejudice the legitimate interests of their authors.

Article 14 of the WIPO Copyright Treaty (WCT) provides that "Contracting Parties shall ensure that enforcement procedures are available under their law so as to permit effective action against any act of infringement of rights covered by this Treaty, including expeditious remedies to prevent infringements and remedies which constitute a deterrent to further infringements."

We appreciate the enactment of the Copyright Alternative in Small-Claims Enforcement Act of 2020 (CASE Act) in late 2020. However, there is no guarantee that infringers will consent to have claims related to orphan works heard by the Copyright Claims Board to be established pursuant to the CASE Act. Writers may still need to bring Federal lawsuits against infringers.

The cost of Federal litigation is prohibitive, in almost all cases, for individual writers. The prohibition on recovery of attorneys' fees for claims related to orphan works amounts to a denial of any effective enforcement procedures in those cases, in violation of U.S. obligations, pursuant to the WCT, to make such remedies available against any act of infringement. The WCT does not permit derogation from remedies for infringement of orphan works.

For all of the reasons above, the NWU opposes the orphan works provisions in the discussion draft. We will be obliged to defend the rights of our members and other writers in the U.S. and around the world who earn all or part of their living from the creation of orphan works by opposing any omnibus copyright bill that includes these or similar provisions.

Despite the range of issues addressed in the discussion draft, several of the issues that are most important to our members and other writers, and that we have raised in the past with Congress and the Copyright Office, are not addressed in the discussion draft.

We believe that any omnibus copyright reform legislation should include the following:

• Reform of Section 203 of the Copyright Act (17 U.S. Code § 203) on termination of licenses or assignments of rights, which is well-intentioned and much-needed, but which has been almost entirely unused and has failed to serve its intended purpose, both because of the excessively long time before assignments or licenses can be terminated and because of the difficulty for authors in tracking down and serving notice on the current successor-in-interest of a license or assignment of rights, especially if the original assignee or licensee has gone out of business (a "zombie publisher" that holds "orphaned rights" that the author wants to recover) or has undergone mergers, acquisitions, and /or restructuring.
• Moral rights of authors of written works, especially the right to attribution.
• A Public Lending Right (PLR) funded by the Federal government and applicable to library lending of written or graphic work in print and digital formats.

We would welcome an opportunity tom work with you and your staff on legislation to address these concerns, either as part of an omnibus bill or as separate bills.

Thank you for the opportunity to comment on the discussion draft. We look forward to the opportunity to work with you and the Subcommittee on copyright legislation.

Sincerely,

Larry Goldbetter, President
Edward Hasbrouck, Co-Chair, Book Division

National Writers Union
256 West 38th Street, Suite 703 New York, NY 10018
<http://www.nwu.org>