In looking back at the serials literature of 2004 and 2005, examining past literature reviews to get an idea of where the profession was prior to 2004 and where it might be heading is valuable. Two serials literature reviews have been published in recent years. In 2004, Silva reviewed the serials literature of 2002.1 Noting that a review of literature had not been written since Riddick’s in 1992, Silva did not attempt to fill in the gap between 1992 and 2002.² Instead Silva focused on serials literature of 2002 while briefly comparing 2002 topics of cataloging, electronic publishing, collection management, serials pricing, and serials management with the major themes identified previously in 1992. Silva found that, in 2002, networked information and electronic publishing—which were only just being introduced in 1992—inspired “nearly as much discussion in the serials literature as cataloging issues do.”³ The issue of electronic serials was no longer limited to networked information and electronic publishing; it pervaded all of the themes identified by Silva.

In the second and most recent serials literature review, covering the period of 2000 through 2003, Corbett found that electronic serials dominated the literature.⁴ Addressing a broader time period than Silva, Corbett took a slightly different approach to the review and limited it to the major themes of cost, management, and archiving. Instead of trying to capture all of the articles published within those themes, Corbett identified articles of significance, or that best represented a group of articles on a topic. However, not all major events and research were captured in the refereed literature, and Corbett looked beyond peer-reviewed publications for those.

Picking up where Corbett left off, this review covers the serials literature of 2004 through 2005. The author has chosen to focus on the subjects of cost, management, and access in the literature. This review also looks at how the connections between parties in the scholarly communication process have shaped the literature in those three areas. Like Corbett, this author found it necessary, at times, to look beyond peer-reviewed literature in order to cover major events or research. Not included in this review are issues related to collection development, such as statistics, transitioning from print to electronic collections, and archiving.

In 2004 and 2005, periodical prices rose at about the same rates as in the past several years. According to Dingley’s study of United States periodical prices, inflation rates continued to climb between 2003 and 2004, going from 7.5 to 8.2 percent, but dropped slightly to 6.5 percent in 2005.⁵ As serials prices increased, so too did the serials percentage of libraries’ materials expenditures. In 2003 and 2004, serials expenditures made up 67 percent of materials expenditures at Association of Research Libraries (ARL) institutions.⁶ The cumulative increase in serials expenditures from 1986 to 2004 was 273 percent, an increase of 13 percent from 2003.

In late 2004, Oxford University Press released findings from a study it commissioned on journal price increases among twelve publishers of scholarly journals.⁷ Both commercial and not-for-profit publishers were included in the study, which found that overall price increases, from 2000 through 2004, varied from as low as 27 percent at Cambridge University Press (CUP) to as high as 94 percent at Sage Publications.⁸ Elsevier had the highest median journal prices for its titles, while CUP and Johns Hopkins University Press were in the bottom three median prices for every subject in which they publish.

In an article that evolved from a presentation at the 2004 North American Serials Interest Group (NASIG) annual conference, Siar, Schaffner, and Hahn examined, from the perspective of a librarian and a not-for-profit publisher, three types of current pricing models: variations of the traditional subscription, tiered pricing, and consortial pricing.⁹ Traditional subscriptions have been for individual print titles, while variations on the traditional subscription now include subscriptions for individual online titles, and either print or online with the other format for free (or at an additional, although reduced, fee).

Consortial pricing allows a group of libraries to leverage their collective bargaining power to obtain a reduction in subscription rates. Consortial prices generally provide groups of libraries with access to a larger number of titles than could be achieved individually. In the tiered-pricing model, differential prices are based on the size of the subscribing institution. This model has been common for electronic resources, but it has recently crept into practice for individual print and electronic journals. In a survey of tiered-pricing models in 2004, Hahn found that several scholarly society publishers adopted tiered pricing for thirty-one journals.¹⁰ To understand the consequences of tiered pricing, Hahn conducted sensitivity analysis of tiered-pricing models and found that, for journals with a subscriber base predominantly in lower tiers, subscribers in the highest tiers would face significant price increases, even when the differentials between the lowest and highest tiers are low.¹¹

Another pricing model featured in the serials literature during 2004 and 2005 was the Big Deal. The Big Deal is a license subscribing to a package or bundle of electronic journal titles over multiple years. As publishers have merged, there are fewer journal bundles available with larger numbers of titles. Typical licensing terms for bundles have caps on annual inflation rates and cancellation limitations. Depending on negotiations, licenses may contain terms on archiving and nondisclosure. The journal bundles often contain titles that a library may not necessarily want and would not have selected if selecting on a title by title basis, and cancellation terms protect these titles during journal cancellation projects. A 2005 survey of ARL members on the topic of bundled journals found that although bundles were common, with respondents having, on average, three bundles with the five largest journal publishers, librarians’ satisfaction with pricing and licensing terms were decreasing as Big Deals were renegotiated.¹²

Addressing the Big Deal from another perspective, a Stanford University survey of members from twenty life sciences societies looked at how scholars have reacted to journal pricing models, including the bundling of e-journals.¹³ The survey authors indicated that profit maximizing by publishers through Big Deals has resulted in cancellation of individual subscriptions and driven scholars to rely on library or other institutional subscriptions. With reduced personal subscriptions, publishers increase journal prices in order to maintain profits and rely more heavily on institutional subscriptions. The survey authors recommended that publishers keep in mind that “libraries are aggregate consumers (for individuals) and if individuals are not willing to pay for journal access at a given price, neither are libraries.”¹⁴ The authors also recommended that libraries might be able to increase their bargaining power by collaborating with users. By communicating and building awareness among scholars of publishers’ pricing practices, libraries may be able to influence where scholars submit and publish their articles.

Support for bundled electronic journals began to diminish in late 2003 and into early 2004, as academic libraries did just what the Stanford survey recommended. Efforts by libraries to increase faculty awareness of the ever-rising journal subscription rates (also known as the serials crisis), the role of the Big Deal in the crisis, and alternative publishing models paid off. In December 2003 and early 2004, faculty senates at several institutions passed resolutions demonstrating support for their libraries’ decisions to refuse the Big Deal.¹⁵

The Big Deal continued to be discussed at conferences and written about in 2004. Speaking at the 2004 NASIG conference, Frazier addressed what he considered to be the biggest problem with Big Deals: the eventual inability of libraries to support journal package cost increases.¹⁶ Following Frazier with another take on the Big Deal, Ebert presented the benefits that the Big Deal can bring to smaller library collections.¹⁷ For smaller libraries, the Big Deal can represent significant increases in their collections. Beyond the discussion of the positive and negative aspects of the Big Deal, an article by Gatten and Sanville in D-Lib Magazine examined how a consortium might go about backing out of the Big Deal, title by title.¹⁸ Armed with usage data for titles across the OhioLINK consortium, the authors identified low-use titles across the consortium and determined that usage data could be used in the future to drop titles from Big Deal collections.

With the growing faculty awareness of the serials crisis that withdrawing from the Big Deal fostered, libraries and their institutions increasingly promoted open access (OA) as a way to provide access to scholarly output and attempt to halt or slow journal price increases. Although OA journals have been around for some time (Regazzi noted that as of early 2004, 81 percent of the titles listed in the Directory of Open Access Journals were founded in 2001 or earlier), momentum for OA as an alternate publishing model built in late 2003 and continued in 2004 and 2005.¹⁹ As interest in OA picked up, many articles on the subject appeared in the literature, including a special issue on the topic in Serials Review.²⁰ The articles in that issue, in addition to articles and reports published elsewhere, focused on the various types of OA publishing; the acceptance of OA with publishers, librarians, and authors; the sustainability of it as a business model for scholarly communication; and analysis of OA costs.

Interest in OA as a publishing model arises from two concerns: rising journal subscription prices and restrictions placed on the use of journal content.²¹ Two basic types of OA publishing models have developed to address the two problems. The first model addresses, or seeks to address, escalating journal prices by switching support for publishing costs from the subscriber to the author, the author’s institution, scholarly society, or funding agency. BioMed Central and the Public Library of Science are two examples of the author or institutional fee-supported publishing model. Many variations of OA are in use: some publications offer selected OA articles, leaving the choice up to the author; other publications offer open access to articles in a set number of months after the article’s initial publication.

The second type of OA model uses repositories and addresses the issue of restrictions on use. Repositories often are hosted at academic institutions and offer authors a place to self-archive pre-prints or post-prints, or both. They can be subject-based, such as arXiv, or limited to the intellectual output of a specific institution. While arXiv—originally a pre-print database for physics—has successfully expanded to include mathematics, nonlinear sciences, computer sciences, and quantitative biology, not all subject repositories that have been created or proposed have succeeded. Weller noted both the success of arXiv and failed attempts in chemistry, psychology, and medicine, and concluded that, “The acceptance and use of preprint databases appears to be discipline specific.”²²

Mark Ware Consulting provided an environmental scan on institutional repositories (IR).²³ The report examines forty-five IRs, looking at the reasons for them, their uses, issues facing them, and publisher attitudes toward them. The report found that IRs are generally small and in the early stages of development. The report cautioned that while e-prints make up about 22 percent of the content in IRs, evidence suggests that they may have little impact on scholarly communication reform, and the number of e-prints could reflect the early stages of IRs.²⁴ The report concluded:

What is far less clear is whether IRs will develop large, interoperable collections of published literature, as hope the advocates of open access. IRs are currently at an embryonic stage with only small, experimental collections of documents, but a clear message from the IRs is that one major hurdle—possibly the major hurdle—is overcoming faculty’s inertia or indifference to self-archiving. It seems possible at present that IRs per se will fulfill a real and valuable function in supporting scholarly communication, research and teaching but that this function will be complementary to scholarly publishing rather in conflict with it. The impact of the wider open access movement is of course another matter.²⁵

In 2004 and 2005, various authors and libraries sought to determine the true cost of OA for institutions, but this did not prove easy. As Holmström noted, “It is difficult for institutions to compare the costs and benefits of subscription-based and open-access-based journals since they charge for their services in totally different ways.”²⁶ One measure used to determine cost for print journals is through cost-per-use data. Adapting the cost-per-use measure, Holmström proposed a method of measuring cost for OA journals by looking at the cost-per-article reading of OA articles. Once the cost-per-use (or cost-per-reading) has been determined for OA journals, then institutions have a method to compare traditional subscription-based journals to OA journals.

Cornell University Library (CUL) also tried to determine the true cost of OA publishing on an institutional level. CUL convened the Task Force on Open Access Publishing in January 2004, and the task force submitted its initial report in August 2004.²⁷ In the preliminary findings, the task force addressed the possible impact on CUL if OA were adopted broadly by Cornell authors. The task force determined that, if Cornell were to move completely to paying author fees in refereed OA journals instead of buying journal subscriptions, OA peer-reviewed journals “would not bring about cost savings for Cornell. In fact, taking into account the number of articles published by Cornell researchers each year and the average cost to publish a single refereed article, CUL would likely see its serial expenditures rise significantly if the library used its current subscription funds to pay for author fees instead.”²⁸ Responding to the CUL report, Gass found the results invalid as the report failed to take into account the role of third-party funding agencies in paying author-side charges.²⁹ “In the United States,” Gass noted, “many research funders like the National Institutes of Health already pay, directly or indirectly, for the publication of primary scientific articles.”³⁰

Addressing the sustainability of the current model of scholarly communication, Bosch outlined seven basic components needed for a sustainable business model for scholarly communication, and examined the OA model in terms of those components.³¹ Bosch concluded that OA is promising, but added that “no business model exists for scholarly journal publishing that makes sense and is going to be sustainable for all participants.”³² Bosch’s conclusion is reinforced by the findings of a report commissioned by the Joint Information Systems Committee (JISC).³³ As part of the report, librarians and publishers in the United Kingdom were surveyed on their preferences in OA business models when given a choice between seven different types. The findings indicated that, while everyone was dissatisfied with the current situation, no one model appealed to both publishers and librarians. A variety of models likely will continue to coexist.

During 2004 and 2005, issues around serials management generated a significant number of articles. Some of the issues related to the topic were shifting collections policies from print to electronic journals, managing licensing and subscription information, and providing and maintaining access. Threaded throughout the serials management theme was the relationship between publishers, libraries, subscription agents, and vendors.

Publishers, subscription agents, and libraries have built a collaborative system, one in which libraries place orders and pre-pay subscriptions to a subscription agent, subscription agents place the orders with publishers and pay the subscription fees, and publishers supply libraries with the ordered materials. This has been a trusted system, but occasionally the system is tested when one member of the system does not live up to expectations. Such was the case with the 2003 bankruptcy of subscription agent RoweCom and its parent company, divine. “This is such a small community, and we are all so dependent on others to complete our respective missions,” commented Geller as she traced RoweCom’s history and troubles to its roots in the F. W. Faxon Company.³⁴ RoweCom had received almost $73 million in prepaid orders from libraries. Instead of placing the orders and paying publishers, the money paid to RoweCom was taken by divine. When the two companies went bankrupt, publishers and libraries were left to deal with the resulting chaos of trying to determine which subscriptions had been placed and paid by RoweCom.

Documenting the efforts of publishers and librarians to resolve the financial mess, Wiegand also offered ideas on how libraries might protect themselves in the future.³⁵ In the end, publishers agreed to fulfill a majority of orders. According to Panos, the lawyer representing RoweCom creditors, “Because publishers agreed to fulfill a majority of orders and thus assumed those claims against RoweCom, currently there are nearly $29 million in library claims and $45 million in publisher claims against divine.”³⁶

In the aftermath of RoweCom, one of the strategies employed by smaller libraries was to bypass subscription agents and order from publishers directly. Doubt about the need for subscription agents was acknowledged by the Association of Subscription Agents (ASA) in a statement made by Rollo Turner, ASA secretary general: “The ASA recognizes that the situation has put considerable strain on the trust that still exists between libraries, agents and publishers.”³⁷ Eventually the strain placed on the relationship among libraries, agents, and publishers eased, for, as Geller noted, “the reality is that this intermediary is vital to the many-to-many relationship of libraries and publishers for all but the smallest library that can handle its own few subscriptions and the largest publisher that can afford to do its own customer service.”³⁸

With libraries licensing e-serials packages directly (or via consortia) through publishers and the aftershocks of RoweCom’s bankruptcy, libraries were questioning the need for subscription agents. Not wanting to be squeezed out, subscription agents were trying to define their role in handling electronic serials. Few articles were published on this subject. One of the few, by Wang and Schroeder, examined the evolving role of subscription agents and outlined how agents could expand their services to provide a service hub offering publishers and libraries support in distributing and managing online serials.³⁹ Several presentations at the 2003 and 2004 NASIG conferences on this topic included “Creatively Coping with Your Subscription Agent’s Bankruptcy” and the two-part session “Helping Manage the E-Journal Forest: Do You Need an Agent Any More?”⁴⁰

Subscription agents were not the only group to come under scrutiny. Davis documented Emerald’s (formerly MCB University Press) republishing of articles without notification from 1975 through 2003.⁴¹ After searching articles in the Emerald database, Davis found evidence of republishing in more than seventy journals and, for some journals, entire issues consisted of republished articles. Republishing without notification goes against publishing guidelines outlined by the Committee on Publication Ethics and accepted by the Association of Learned and Professional Society Publishers. Publishers also have guidelines in place to ensure that original content is being published. Republishing without notification creates confusion in citations and it comes at a cost to libraries that may end up paying for the same content more than once.

In a letter to the editor in response to the article and to a letter by Davis, Howard reported that Emerald conducted a follow-up study of republishing in their journals and found 560 instances of original papers being republished, representing 1.1 percent of the total Emerald database.⁴² Howard noted that, “87 percent of republishing took place in or prior to 1999” and that any republishing after MCB’s name change to Emerald in 2001 was due to author or administrative error.⁴³ Emerald has adjusted its publishing practice, Howard acknowledged, to prevent further republishing errors, and articles republished now are fully attributed.

Electronic journals permeated the major themes of cost, management, and access outlined in this review. The impact of online serials on publishing, libraries, and intermediaries has been tremendous. As libraries began turning away from the Big Deal, the dialogue around OA increased. Repositories and open access journals were promoted as ways to provide an alternative to traditional publishing and, hopefully, as a way to slow journal price increases. The serials literature of 2004 and 2005 shows libraries struggling to cope with providing access to and managing electronic journals. Libraries were implementing new products and services, such as link resolvers, A to Z lists, and ERMs, aimed at providing that control and access. These services and products are provided by e-serials management intermediaries, a new group in the scholarly communication circle. In addition to the new e-serials management intermediaries, traditional subscription agents were redefining their work to include electronic journals and offering libraries services for managing e-journal subscriptions.

A fourth theme interwoven throughout the literature is the connections among all groups in the scholarly communication circle—authors, publishers, libraries, and vendors. The literature review found evidence of those connections at work in the communication and collaborations between libraries and scholars; libraries, subscription agents, and publishers; libraries and ILS vendors; and libraries, e-serials management service providers, and content providers. Libraries communicated with faculty, who are authors and editors, about journal pricing issues, open access, and repositories. Successful communication allowed a number of academic libraries to walk away from the Big Deal, with the full support of their faculty senates. It also generated further discussion and promotion of alternative publishing models.

An integrated library system vendor partnered with libraries to develop a new product to respond to the growing need for a system to manage electronic resources administrative data. Libraries and publishers worked together to find the best resolution to the divine/RoweCom financial crisis. Efforts are underway, with involvement from all parties, to create standards for communicating serials metadata. Hopefully, those efforts will pay off in improved communication between content providers, libraries, and e-journal management services as they work to provide users with timely and accurate information about the title availability and full-text coverage information for those titles.

A few subjects were not well-represented in the serials literature, mainly in the areas of serials acquisitions, and the relationships and communication between libraries and third-party service providers, and between those intermediaries and content providers. The author was surprised that more was not written about serials acquisitions work. With libraries changing collections from print to electronic formats, little was written about how print serials acquisitions work has been affected, how staffing was altered, and whether any activities were discontinued or added to acquisitions work.

Another topic receiving surprisingly little attention in the peer-reviewed literature was the relationship and communication between libraries and subscription agents. A number of conference presentations and a few articles addressed the subject; perhaps more will make their way into the literature soon.

Finally, accurate and timely communication of serials metadata among libraries, e-serials management service providers, and content providers is crucial to supplying library users with the information they need. Little, however, has been written on how often title lists and coverage data are published by content providers, the accuracy of the data when it is provided, and how frequently the information is updated. The need for improved and standardized communication of serials data to various interested parties is evident and will be more important as new tools are developed that rely on the same coverage information.