The way libraries acquire basic content for their readers has been completely upended in the last two decades. I have worked in research library collections and acquisitions through most of that period, from the days of the first subscriptions for electronic journals. The old days were good—or at least we had understood for decades how the rules worked: libraries purchased a book (or journal or microform or other tangible format) and, under the right of first sale in the US Copyright Act, they placed it on a shelf and users borrowed the item and returned it, until/unless it fell apart years later. At that point, the library could purchase another copy, or make a reproduction as permitted within Section 108 of the US Copyright Act. Life in library acquisitions proceeded as in a production shop: orderly and careful.
No more. In this rapid electronic environment, content providers are pressed to enhance and update existing products or to produce competitive new products, with ever-increasing functionality and with great uncertainty about what users will pay for and how much they will pay. At the same time, numerous new producers are entering the electronic marketplace. We are living in an information Wild West, which can put libraries and publishers face to face on Main Street at high noon, often without the third-party subscription agents or book jobbers we used to depend on.
Increasingly, libraries gain access to electronic content with a bundle of sometimes confusing and customized rights of use, often without the benefits of ownership. Those rights are codified in a license, which is notionally a contract between two parties: a willing buyer and a willing seller (in the library case, between it and the information provider). Under license, the institution has those rights that are expressly an agreed part of the contract. As all the stakeholders know, there are pros and cons and serious issues along this new path, which will be with us for many years to come. The expertise that library (and publisher) staff nowadays need in order to acquire materials has been ramped up. In addition, academic authors have jumped into the mix, adding to newness and complexity. Exciting? Yes. Headache-making? That too.
How did we get to this place?
Why are we not using copyright law to govern today’s electronic information transactions? There are a number of reasons. Many authors and publishers feel that copyright law does not effectively address issues specific to this new world, wherein users have the advantage of high-powered copying and instantaneous redistribution technologies. Today, in theory, it would be possible to distribute worldwide thousands of copies of a digital information object in seconds. Copyright does not protect materials in the public domain, yet many publishers are packaging public domain materials in new digital ways and looking to monetize the added value— copyright law by itself can’t help them very much. In these kinds of situations, licenses can help information providers gain some control over and income from electronic materials. And licenses are not as vulnerable to ongoing legislative changes as copyright-protected materials may be.
So we now see consortia/publisher agreements that reach not merely into the millions of dollars but the tens of millions of dollars per year in a single negotiation. Not infrequently, government and funding agencies have become interested in supporting electronic resource negotiations, with the goal of delivering access to all the citizens or researchers in a given discipline, state, or nation. In those scaled-up situations, the pressure to secure a contract is increased. The volume of generated business is attractive to publishers, as is the time saved in not negotiating with numerous individual institutions. Buyers get a better deal. All this is serious business, bearing little resemblance to standard library book or serial purchasing. When I took up an academic position as Director of Collections Development at a major institution in the mid- 90s, one of my first tasks was to review and sign a renewal contract for a major database, at that time delivered via text format (the web had not quite taken hold yet). Totally without licensing experience (though with experience of reading other types of contracts), I was in the same position as most of my colleagues at the time—vastly underpowered for this new assignment. Imagine my surprise as I read the contract renewal terms: “No reproduction may be made from this resource by any means, mechanical or electronic.” I phoned the provider: “Does this mean that if my user copies citations with pen on paper, s/he is in violation of the license?” “Yes,” I was told, “but why not just sign—this will be the library’s fourth renewal.” And thus it began. Happily, the publisher was pleased to discuss what a reasonable substitution would be. I cited snippets from Section 107 of the US Copyright Act (Fair Use) by way of example, and we adjusted the contract. It was an exciting moment—to have the opportunity for serious and productive discourse with the “other side” and to begin a process of mutual education. Unfortunately, not all publishers were as open minded and willing, and some language had been written by cadres of lawyers who had no idea what libraries and users were about. Over the years, many colleagues have spent endless hours in intractable negotiations, starting with lousy user terms that have been changed only with sweat and tears, and sometimes over a few years. These kinds of experiences led a number of organizations to issue useful licensing principles  and model licenses.  Over time, there have been many advances in the real benefits publishers allow for users.
Should we prefer copyright or license?
Copyright and licenses (contracts) share certain characteristics. Both accept the existence of the concept of intellectual property, where rights include those of the property owner and of the user and or purchaser. They are also very different. Copyright law is the law of the land. It varies widely across national borders, changes from time to time, and tends to be high in principle and low in specificity in certain crucial points. US copyright law, for example, famously guarantees the right of “fair use,” outlining high-level generalities about what criteria we should use for determining whether a given use is fair. Licenses are transaction-specific, and in the US contracts are governed at the state level.
As time passes, we may be able to develop copyright laws that dispense with the need for e-resources contracts, though I am skeptical. And this may not be the best outcome. We may be most successful when the law is paired with thoughtful, well-written contract language, firmly grounded in copyright principles. A license also includes important provisions that are not copyright-related: agreements about pricing and other business terms, content inclusion, who are the customers, and much more. The license defines every aspect of the business arrangement, such as what users can do with the property; where, when, for what costs; and what both parties commit to in the deal.
A good license also makes clear the conditions under which it is to be enforced, e.g., specifying the jurisdiction in which legal action would be taken. But it’s worth emphasizing that library license agreements have rarely—perhaps indeed never—been made the basis of litigation among major parties.
Licenses can restrict rights granted by copyright (undesirable from the library point of view), can incorporate copyright definitions and principles (such as interlibrary loan or fair use), and can clarify and even extend rights granted by copyright. (If a license fails to address a specific reader right, copyright then provides the answer by default.)
These days, it’s all about rights
These days everyone’s interested in his or her rights. Authors often want to hold on to copyrights rather than transfer them outright to publishers; publishers want to keep control over use, future use, and revenue; and libraries insist on gaining the rights to use materials broadly—in numerous ways for teaching, scholarship, research, and collaboration with other libraries. Library users demand the right to download, share, and re-use information. And universities are increasingly seeking to become globally visible and to influence the economics of the industry by asserting ownership of the works of faculty and staff—or at least controlling the character of outlets that their colleagues may use.
While not at the core of academic business, commercial authors are nonetheless important contributors to newspapers, magazines, trade books, and other materials that libraries make available and readers need. Over recent years, many of these authors have pushed back at some practices of their publishers. See, for example, New York Times Co. v. Tasini, which was finally decided in favor of the authors by the US Supreme Court.  In this case, members of the National Writers Union brought successful suit against five major publishers, charging copyright infringement when the freelance authors’ previous works were licensed for re- use in electronic databases without explicit permission or payment. Through recent court actions against the HathiTrust, the Authors’ Guild and two other plaintiffs asserted, unsuccessfully, that authors’ rights had been infringed via the Google Books Project.  Without a doubt, as a result of these and other actions, author/publisher contracts have been closely scrutinized by both sides and revised so as to give both what they need in a digital age.
The motivations of academic authors, particularly those writing journal articles, are as intense as those of commercial authors, though generally less from financial need. In the majority of situations, academic authors wish to assure that their works are widely available, distributed, and re-used. Those who in the solely print environment automatically signed standard publisher copyright transfers now have different requirements and expectations. These days, an author-reader of “traditional” copyright release forms is likely to observe that the publisher, in effect, requires a transfer of all rights, leaving the author possibly unable to use the work freely in the classroom, or to post publications on his/her own website, in an institutional repository, in a subject repository elsewhere, or in a mandated site (NIH, for example), to name some common situations. Open access, with a growing number of assorted mandates (by institutions, funding agencies, or governments), and with strong organizational impetus (SPARC) and ARL , for example) and personal advocates, has also had a powerful effect on the old copyright transfer form.
To be fair, it is possible—and some publishers do this—to use a copyright transfer form that gives back to academic authors all the rights they might possibly want (at a given point in time), while leaving the actual ownership with the publisher. Such publishers observe that, in the event of dispute, it is more effective to retain large sets of rights, which enable them to take action against abusers. However, where an academic author is willing to manage his or her copyrights, the position of ownership is the stronger one.
Wikipedia defines publishing as:
“The activity of making information available to the general public....Traditionally, the term refers to the distribution of printed works such as books (the ‘book trade’) and newspapers....With the advent of digital information systems and the internet, the scope of publishing has expanded to include electronic resources, such as the electronic versions of books and periodicals, as well as micropublishing, websites, blogs, video game publishers, and the like.”
More important is their description of the publishing function:
“Publishing includes the stages of the development, acquisition, copy editing, graphic design, production...and marketing and distribution....” 
Long-time academic publishers have made or are making the digital transition, often very successfully. Additionally, the last two decades have seen rapid development of digital communications technologies and tools that entice new entrants into the publishing arena. There are start-ups that deliver formal, for-pay journals (author or subscription funded—many being open access) and databases; university initiatives; collaborative works; and self-published “grey literature,” to name some. A recent high-visibility topic is that of “Libraries as Publishers.” Libraries in academic settings, in response to user needs (students or faculty), their own needs (digitizing collections, e.g., to preserve them or make them accessible to users), or those of their universities (visibility) are now creating sites that range from informal “attics” to carefully curated materials to partnerships on campus (with university presses or IT centers), and wider partnerships. A few are even able to provide publishing services in response to RFPs from outside organizations.
All types of academic publishers, old or new, share similar desires: get the work out and find sustainable ways to support it—whether through charging fees to writers or libraries; or re-allocating institutional resources such as time, tools, and expertise; or finding backers. All want to be sure of the right to produce and make available their work; none want the work to be abused, whether financially or morally. Today’s variances are in investment to be made in acquisition, writing, editing, production, and marketing. Formal publishers are likely to do all these things; most library publishers will do some of them but not all; individuals may do even fewer. However, all share a desire to control their product, protect their investment, and be seen as responsible agents. Licenses can offer the most satisfactory way to achieve those goals.
Libraries and their end users
Types of licenses
Shrink or click
The commonest forms of end-user licenses include some that libraries prefer not to go near—but are part of many everyday lives. These are what we call contracts of adhesion: “take it or leave it” licenses, e.g., the “shrink wrap” licenses that took their name from the protective coating on the boxes in which software may be delivered.
Such licenses are presumptively agreed to when the purchaser rips open the wrapping. These have been largely supplanted by the “click through” licenses that pop up as an online dialog box prior to installation or use. We see these less frequently in institutional settings. They may be legally binding in the US, but there has been no definitive legal tests of that. When encountering such language today, librarians generally go back to the publisher and insist on negotiating something that meets current professional standards. Similarly, the individual one-user license is not a practical reality for library settings where resources will get serious use.
Most commonly, as individual institutions or in groups, libraries seek to negotiate “site licenses,” tied to the physical or virtual facilities of the institution(s) and its/their constituents. While these may notionally be limited to the campus of a university, all such licenses nowadays recognize the practices of institutional users working off-site and connecting to the resources through proxy servers or virtual private networks. As well, site licenses generally recognize that branch libraries, branch campuses, and campuses abroad exist; how publishers choose to charge for these, or not, is often a critical part of negotiations. A vexing piece of negotiation is finding a way to measure the quantity of use and to match price to that quantity. For example, one measure of reference has been “historic spend,” wherein the publisher assesses how much an institution had previously spent on its print resources; another measure might be the FTE count of some or all of the user population—faculty, faculty plus students, students alone, faculty-students-staff, or some combination. Such practices are increasingly challenged by institutions offering large distance/online learning programs, whether for tuition-paying (and thus FTE-countable) students or for the global public (as in a MOOC). In general, libraries aim to retain the right to define (according to institutional measures) and authenticate users into the system and resist such cumbersome practices as providing publishers with regularly updated lists of authorized users. (Outside the US, the consortium idea has led to near-national site licenses. For example, it has proved possible in countries such as the UK and Canada to gain terms effective across much of the tertiary education sector of a nation. In such cases, government or significant foundation funding may have kick-started the arrangements.)
Access in developing nations
Outside the US, one heartening movement has seen the growth in developing nations’ initiatives in content licensing. Publishers recognize the high value that their content often has in economically and societally challenged settings, where normal pricing would effectively prevent dissemination and use. Various initiatives have made it possible for researchers and libraries in developing countries to have favored access to important resources at little or no cost. Hundreds of publishers voluntarily participate in initiatives led by various UN organizations under the banner of Research4Life , and individual providers such as JSTOR have also mounted their own initiatives . NGOs such as eIFL and INASP, along with others, work in developing countries on a large scale . The basic terms of licensing may remain essentially the same: what differs is mainly the price. Some programs are free while some initiatives have tiered pricing: free to a small number of least advantaged states, deeply discounted to a group of others, etc.
Who you gonna call?
Even with the growing skills being brought to bear on negotiations, there is need for a still higher level of expertise. Conditions, practices, laws, and regulations change; publishers innovate; and user expectations develop. Various institutions have developed and published “model licenses” that set out in clear form what concrete license language and terms can look like in order to achieve libraries’ purposes. Since 1997, I have coordinated the LIBLICENSE Project, which has three components: (1) a website rich in general resources on licensing, (2) a lively discussion forum where current issues are reviewed in real time, and (3) model license language and software to support creation of new license agreements . The Project’s first Model License dates to 2001, and at the same time it created and made freely available software to allow for do-it-yourself customization of academic libraries’ best practice licensing language to produce contracts for specific situations. There have been various revisions, most recently the November 2014 rewrite —the most ambitious in our history, based on broad consultation with stakeholders. We are at present working to incorporate the wisdom of this document in a new generation of shareware DIY license-writing software. Throughout, this Model License respects and relies on industry standards and best practices.
Current issues in licensing
The work on the LIBLICENSE Model License has arisen from, and in turn sharpened, awareness of newly emerged issues facing those who negotiate and manage academic e-resources licenses. A few new topics include:
- Text and Data Mining (TDM, Content Mining)
During revision of the Model License, this area received the most comment and interest. Users are increasingly interested in being able to reach into datasets of every sort and ask customized, sophisticated questions—more than just “searching the archive.” The more sophisticated users want to be able to pull information from multiple datasets at once, to find correlations and connections that can never be found in one set alone. So research library contracts need language designed to allow for broad and flexible use, without users becoming trapped into enumerating specific cases and asking permission. There should be explicit license rights to engage in TDM for scholarly and educational purposes, to share the results in scholarly work, and to make outputs (effectively, new, derivative datasets) available for use by others. There is a need for arrangements that allow users to download the data directly, rather than depend on a vendor-provided API. (The publishers often resist this, sometimes out of a desire to retain control of the data, but also perhaps to observe and learn from the kinds of queries pursued.) In some cases, publishers have attempted to levy extraordinary charges for the supply of copies: thus, more negotiation is needed. TDM is these days a contentious issue between many publishers and their customers.
- Use in Discovery Systems (“Content Neutrality”)
As sophisticated discovery systems developed by publishers or third parties allow users to reach into their libraries’ content resources for information of interest, it becomes necessary to require publishers to provide to the licensee’s discovery service vendors, on an ongoing basis, the citation and descriptive metadata (subject headings, abstracts, keywords) and full-text content necessary to facilitate optimal discovery. Here, as everywhere, time no longer marches forward but rather sprints, so the new Model License needed to be reviewed in light of the most recent NISO Open Discovery Initiative release and also industry practices .
- Author Rights
There is increasing interest in ensuring that institutionally affiliated authors are able to re-use their own works for scholarly and educational purposes and to deposit their works in institutional or other open repositories. The Model License reads, in part, that institutional authors “shall retain the non-exclusive, irrevocable, worldwide, royalty- free right to use their Work for scholarly and educational purposes, including self-archiving or depositing the Work in institutional, subject-based, national, or other open repositories or archives (including the author’s own web pages or departmental servers), and to comply with all grant or institutional requirements associated with the Work.” We have taken guidance here from the Model Authors Rights language endorsed by the Association of Research Libraries .
- Open Access Reporting and Article Processing Charges (APCs)
Keen interest on all sides in the development of open access publishing leads to new emphasis on usage information. The Model License language now asks providers to report annually on the number of open access articles published and encourages good-faith discussions about subscription fee impacts, with a goal of reducing such fees in proportion to the amount of open access (particularly APC) revenue received. The goal is to manage the economic impact of local open access authorship in a way favorable to the research and library communities and to watch for double-dipping (where publishers, inadvertently or otherwise, charge twice for the same publication). Recent announcements from Jisc regarding their “Total Cost of Ownership” approach  suggest it is possible to reach agreement in contract negotiations .
- Confidentiality and Privacy
The realization that digital data make institutions and individuals vulnerable to loss of cherished security has pumped substantial energy into discussions of these related issues. We speak of confidentiality when it comes to maintaining control over data about the licensing deal and its operations—e.g., over handling of usage statistics, financial terms, and institutionally privileged data. Those concerns can be intense but pale in comparison to burgeoning global concern over information privacy. In academic settings, normal concern over privacy of personally identifiable information extends as well to a need to maintain the integrity of the research process. Privacy concerns are leading, as well, to divergent legislation and government practices around the world, making it harder for information providers to establish a single set of protocols to apply everywhere. The Model License was able to address some of these issues, but much more time needs to be invested in this issue in the future, by all stakeholders.
- Other Issues
Experience has taught libraries to seek additional or updated licensing provisions and we have attempted to address them in the new Model License. For example:
- Perpetual Access
Licenses increasingly include affirmation of right of perpetual access to licensed resources— essentially that access should continue for resources that libraries previously licensed/paid, even if the resource is discontinued, the library cancels its active subscription, or the resource changes publishers. In the latter case, library licenses these days include clauses requiring the transfer of obligation, when the intellectual property managed by one publisher is acquired by another; the Transfer Code of Practice  is the standard here.
- Holdings Lists
Libraries may wish to seek the right to obtain itemized holdings lists annually or on request, in KBART-compliant format . This may be of especial interest in determining content completeness when digitized backfiles, newspapers, or commercial collections are created and then licensed.
- Digital Rights Management (DRM)
This has become a focus of attention in recent years. Libraries have various concerns about the use of DRM, which can restrict otherwise legal copying, sharing, reformatting, or changing electronic information, particularly in purchased e-resources. Not only may DRM impede access to resources that might be normally permitted by copyright law; but DRM also can make it impossible for a library or consortium to exercise its full rights of perpetual access . (DRM tools, intentionally or otherwise, can also be seen as intrusive on the privacy of individual users.)
The new LIBLICENSE Model License aims to be format- neutral, i.e., to be applicable not just to e-journals, but also to other scholarly electronic formats such as books, databases, reference works, AV material, and so on. Special circumstances can apply to different formats and careful negotiations must address those. For example, of significant concern these days are licenses for e-books—whose numbers are now growing, seemingly as rapidly as e-journals did between 1995 – 2005. Some library and user-experienced problems are well outlined by Walters, who addresses “restrictions on viewing, printing, downloading, circulation, and ILL.”  Herman provides an academic user’s perspective . Both of these pieces raise a number of issues that, for better or worse, librarians and publishers must work together to resolve, and licensing must play a key role.
The end of librarians’ licensing labors often comes invisibly, transparently, and wonderfully. A scientist in her laboratory reaches out and finds just the article or
just the dataset that makes a crucial difference in the next discovery that will make the world safer or cleaner or healthier. We know well from experience that when such an “aha” moment occurs, that user may not be aware of the role librarians have played in opening the pathway through which that knowledge has flowed. Nonetheless, librarians know that they have wizardry of their own, and they will do what it takes to maintain those powers. One hopes this magic will happen in an increasingly cooperative world between librarians and the information sector.
(firstname.lastname@example.org) has been working with the Center for Research Libraries as Senior Advisor on Electronic Strategies as of Fall 2011. She is one of the active, founding spirits of the International Coalition of Library Consortia (ICOLC). She founded the NERL consortium and led it for 15 years, through summer of 2011. In particular, she has been a leader in licensing electronic scholarly resources, having created the LIBLICENSE Project, which includes a newly- re-written Model License, whose previous versions were adapted widely by libraries and organizations.