The Evolution of Publishing Agreements at the University of Michigan Library

Taking as an example an open-access journal with a single editor, this article discusses the various configurations of rights agreements used by the University of Michigan Library throughout the evolution of its publishing operation, the advantages of the various models, and the reasons for moving from one to another.

One of the main reasons library-based publishing operations have been formed is in response to dissatisfaction with traditional publishers, which are frequently vilified for obtaining nearly exclusive rights to scholars' work and producing expensive products, thereby hindering authors' use of their own work and impeding broad and affordable access by readers. In response, library-based publishers have aimed to publish more cost-effectively and provide fairer terms to authors than traditional publishers, especially by allowing authors to retain copyright, granting to the publisher only those rights necessary for publication. This grant of rights or license sometimes happens using a click-through agreement when submitting a manuscript through software like Open Journal Systems (OJS) and sometimes happens by signing a contract. This license is often non-exclusive, meaning the author can grant similar rights to another party besides the library-based publisher.
The University of Michigan Library's publishing operation obtains agreements in writing. As the publishing operation grew from a few staff members (the Scholarly Publishing Taking as an example an open-access journal with a single editor, this article discusses the various configurations of rights agreements used by the University of Michigan Library throughout the evolution of its publishing operation, the advantages of the various models, and the reasons for moving from one to another. Office) to a multi-department staff (MPublishing) and later an operation fully integrated with the University of Michigan Press (Michigan Publishing), the approach to rights management with authors and editors has evolved along with the organization's thinking about these questions. Taking as an example an open-access journal with a single editor, this article discusses the various configurations of rights agreements used by the U-M Library throughout the evolution of the publishing operation, the advantages of the various models, and the reasons for moving from one to another.

FIRSt GeNeRAtION: MeMORANdA OF UNdeRStANdING
When the Scholarly Publishing Office was first created, journals, bibliographies, and other material were accepted for online publication as opportunities arose. Many had already been published in print, and some of the journals continued to publish in print even after partnering with the library. Each project had unique features, and no standard publication types had yet emerged. What was especially unclear was the division of labor between the library and the publishing partner-in the case of a journal, the editor.
To clarify this relationship, a memorandum of understanding was drawn up. It included a description of what files the editor would provide to the library, and what the library would do in return. It was usually written as a letter from the head of the Scholarly Publishing Office to the editor but not signed by either party. SPO staff did not have these reviewed by staff of the university's general counsel, seeing them, incorrectly, as nonbinding agreements.
The library accommodated such journals' production workflow and file formats where possible in digitizing back issues and publishing new issues online. Since these journals already used agreements with their authors, the library's publishing operation only sought a single agreement with the journal editor, not with each author. The standard author agreement was perhaps reviewed to verify that it included rights to publish online, but the responsibility for collecting these agreements lay with the editor. Furthermore, the single agreement with the editor was always non-exclusive: editors were free to make their content available through other channels, both during and after any relationship with the library. Given this arrangement, it didn't make sense for the library to enter into an agreement with each author just for the version published online by the library.

SecONd GeNeRAtION: AGReeMeNtS BetweeN tHe LIBRARY ANd tHe edItOR
As standard publication types emerged, the library's publishing operation (which by then was rebranded as MPublishing) needed boilerplate agreements that included the best clauses from past agreements to ensure the rights of authors, the journal editor, and the library. Furthermore, library staff wanted to ensure that important clauses not previously included in agreements-notably, an explicit granting of publishing rights by the journal to the library-were included as well.
The practice of the library entering into only an agreement with the journal editor continued. The journal editor warranted that he or she had the right to authorize the library to publish the articles in the journal-that is, that the editor had secured author agreements from all contributors. Since the library increasingly took on journals that had not previously been published in print or electronically, it became increasingly important to offer guidance to the editor on author agreements. The library provided two variants of a model agreement for use by the editor: one in which the author retained the copyright but granted to the editor a non-exclusive license to publish and to grant others (such as the library) the right to publish, and another in which the author transferred the copyright in the article to the journal. The latter was originally devised out of concern that, if the author kept the copyright, the library would exacerbate the orphan works problem by making it harder for readers to track down authors in order to republish the work; however, publishing staff eventually decided that the former agreement was indeed sufficient to cover future uses.
In addition to a warranty that the editor had secured the right to publish all content, the single agreement also included a clause, standard in publishing contracts, that guaranteed that the journal contained no defamatory or libelous material. Furthermore, the editor indemnified the library for any breach of the agreement, meaning the editor would be completely liable for any content published by the library as part of the journal that could lead to a lawsuit. This was a problem for journals bringing back issues for migration to the library's site, for which author agreements could not always be secured. It also left the editor personally liable for actions undertaken in the course of editing the journal. The library recommended that agreements be signed not by the editor personally but by a representative of an organization sponsoring the journal (if one existed). Alternatively, editors were encouraged to incorporate as an S corporation or an LLC and sign as this corporate entity.
Past agreements were gradually revisited to move to the new standard agreements. While Creative Commons had emerged as the preferred method for sharing open-access content, the focus for the publishing operation had always been simply on making content available to read online, without insistence on attaching a CC license. While the first-generation arrangements predated Creative Commons as an organization, once use of CC licenses became common, they were incorporated into the model agreements and single agreement with the journal editor. Originally the Attribution license (CC BY) was used for journals, though as one editor after another balked at such permissive licensing, the default was changed to the Attribution-NonCommerical-NoDerivs license (CC BY-NC-ND). However, as major players in open-access publishing such as Elementa, PeerJ, Wiley Open Access, and OASPA began using CC BY (in accordance with the definition of open access from the Budapest Open Access Initiative), the default was changed back to CC BY, with an understanding that this might lead to a productive discussion with the editor and, if necessary, a modification of the terms of the agreement.

tHIRd GeNeRAtION: AGReeMeNtS wItH edItORS ANd dIRectLY wItH AUtHORS
As the library's publishing operation was fully integrated with that of the University of Michigan Press (with the combined operation rebranded as Michigan Publishing), it made sense to reconcile the different rights agreements in use. The press, as a publisher of monographs, had always made agreements directly with authors. In the case of an anthology, the press would make an agreement with the editor of the anthology, with a brief contributor agreement signed by each author. All of these agreements were kept on file at the press. The press anthology model will be used as the basis for the third generation of agreements for journals. A single agreement will be signed by the editor and a representative of the University covering the journal as a whole, but the library will also require a signed agreement from every author of a journal article granting a license to publish to the university. This agreement could be consulted in case of a dispute, instead of having to rely on the editor's word that the necessary rights had been secured as in previous generations of agreements. More important for the editor, he or she-or the journal's sponsoring organization-would not be liable in case of such a dispute.
However, the story of the evolution of the U-M Library's publishing operation isn't just one of increasing formality and conformance to the model used by the press. The integration of publishing operations and creation of Michigan Publishing also led to an examination of the author agreement used for University of Michigan Press titles. In a new standard author agreement for press titles that debuted in 2013, authors are allowed to keep copyright while granting publishing rights to the press, allowed to deposit the work in an institutional repository after three years, and offered the opportunity to license their work with a Creative Commons license, either immediately or after three years. If they choose the immediate option for a CC license, they receive an advance on royalties. Why do this? Michigan Publishing is committed to taking a leadership role in the expansion of the open-access philosophy to monograph publishing but understands that one of the impediments to author adoption of open access is the risk of losing royalty revenue. The incentive program is designed to nullify this particular concern. While Michigan Publishing believes that, in many cases, open access to monographs will stimulate sales, the advance against royalties serves as a kind of "insurance policy" to authors who would are interested in going OA but don't feel enough data yet exist to persuade them that doing so won't undercut their sales.