Introduction

Twenty-five years ago, one of the easiest ways to keep abreast of the latest professional issues was to subscribe to journals and thereby have ready access to information within an arm's length. As subscription rates increased and libraries made copiers readily available at minimal cost per page, however, the number of personal subscriptions decreased. At the same time libraries watched subscription costs skyrocket to the point that they now face a "serials pricing crisis" (Bot, Burgemeester & Roes, 1998). The crisis is the result of libraries not only being unable to subscribe to new journals, but also being unable to continue subscribing to the same number of journals to which they had previously subscribed (Odlyzko, 1999a). Increasing familiarity with computing technology and the appeal of having low-cost scholarly articles available at one's fingertip at home and in the library are powerful motivators to pursuing electronic publishing; and indeed, "upstart" electronic journals have begun to spring up (MacKie-Mason & Riveros, 1997).

Increasing familiarity with computing technology...

To remain competitive, commercial publishers and scholarly societies are producing electronic journals at an ever-increasing rate; and although scholarly electronic nursing journals still number under ten, hundreds of journals are being produced in other disciplines (Sparks, 1999). Despite the fact that subscriptions to nursing print journals usually are less costly than non-nursing journals, institutional subscriptions are at least twice those of individual subscriptions, and certainly, by any measure, neither is inexpensive!

The question of whether quality electronic journals can be produced at lower costs than print journals is an issue that is undergoing vigorous debate. The purpose of this article is to present an overview of the economic implications of electronic publication of scholarly journals in general. Although the literature contains virtually no information regarding the economics of nursing journals, there is a wealth of information available about non-nursing journals that is instructive to those interested in electronic publishing in nursing. The costliness (or lack thereof) of electronic journals in relation to print journals will be addressed, and ways to recover the costs of publishing electronic journals will be suggested. Also addressed will be predictions for the future of this new means of providing scholarly information.

Comparing Costs

The issue of whether electronic journals are less expensive than print journals is a murky one that can lead to comparing apples and oranges unless attention is given to details of the comparison.

The issue of whether electronic journals are less expensive than print journals is a murky one that can lead to comparing apples and oranges unless attention is given to details of the comparison. For instance, for some publishers, an electronic journal is the electronic form of a print journal. Using computer-based technology, the print document is converted to digital format and stored, and then is distributed on demand over the World Wide Web. Costs of the additional electronic version of a journal for publishers are about 30% more than the print journal alone (Regier, 1997). Shirrell (1997) has suggested that a possible solution to these add-on costs might be to combine the production of the two at an early stage rather than converting the finished print form to electronic. He suggested that data could be converted to a common database language, and then the same database could be used to produce both the typeset and electronic versions.

Other publishers produce print-only and electronic-only journals. Comparing costs among these journals would seem to be straightforward, but the question of whether the publisher is a for-profit or nonprofit business such as a university press or a scholarly society affects the outcome of the comparison. For example, Odlyzko (1997a) reports that subscription costs of print journals vary widely among commercial publishers, universities and societies; and at least in economics, journals published by societies are least expensive. An informal review of the subscription rates of nursing journals leads to the same conclusion.

Another type of nonprofit publisher of electronic journals is that of scholars themselves. Odlyzko (1999a) states that the number of electronic journals published by scholars is growing rapidly, and these journals often are available at no cost to the user. To sort out how it is possible to publish free electronic journals, and to predict whether this practice is likely to continue, it is necessary to explore and compare the costs of producing print-only and electronic-only journals. The exploration and comparison must include consideration of categories of common publishing activities and categories of costs direct and indirect costs.

Categories of Common Publishing Activities

Regardless of the finished product, publishing a journal requires a certain sequence of activities; and each activity has associated costs. One category is the acquisition of the manuscript. This category includes authors' time for writing the article, and editors' and referees' time in making the decision to publish the article. In most cost comparisons this category is ignored since these groups of people generally do not receive compensation for their efforts (Bot et. al., 1998).

Another category is the preparation of the "first copy" of the manuscript. First copy costs include such things as copy-editing, which actually is a process that involves the return of the copyedited manuscript to the author and final copyediting of the author's response that goes to the "typesetter" (Fisher, 1997). This process is the same for print and electronic journals. The next step of producing the first copy, however, differs. For a print journal, this step involves typesetting the manuscript (creating the page lay-out and preparing the plates), and for an electronic journal, it involves formatting and inserting links (Fisher, 1997; Regier, 1997). Other categories of costs include mass reproducing, distributing, and administering the journal (Regier, 1997). According to Regier, proponents of electronic publishing focus on the perceived cheaper costs of producing and distributing electronic journals and ignore the "numerous new costs" encountered with electronic publishing. The new costs referred to are those of "system administration, content cataloging, tagging, translating codes, tagging them, network charges, computer and peripheral charges, and additional customer service" (p.3).

Categories of Costs

When determining which form of publishing is costlier, attention must be paid to activities categorized as direct and indirect costs. The latter includes overhead, shared, and other indirect costs.

Direct costs. Direct costs are those that are generated solely by the journal, and therefore, must be recovered totally by the journal through subscriptions or other sources of funding. Compared to indirect costs, they are the easiest to identify.

Indirect overhead costs. Indirect overhead costs are those that are allocated by publishers to their journals across the board according to a rule-of-thumb policy decision. That is, publishers may assign overhead costs to a journal based on: a) production, according to the number of issues produced; b) circulation, according to the number of subscribers; and c) marketing and administrative categories (Fisher, 1997). In addition, Fisher states that electronic journals incur overhead that is allocated according to the filesaver and network costs, and staff time of the journal.

Indirect shared costs. Indirect costs are referred to as shared costs when the publishing of a journal occurs within an environment of broader activities. In such an environment, many of the journal costs can be assigned according to a ratio of specific resource use and total use within the broader organization (Bot & Burgemeester, 1998). An example of shared costs is the Online Journal of Issues in Nursing (OJIN), which is published as one component of the broader web site of the American Nurses Association (ANA). Much of the costs of producing the journal for online publication are shared costs incurred by the activities of the ANA technical staff and equipment required to produce the web site. Similarly, much of the first-copy costs are shared costs borne by Kent State University (KSU) through the effort of the OJIN editorial staff who are KSU faculty members, and by use of university equipment to produce the first copy.

Other direct costs. Other costs may be included as indirect costs when certain publishing activities and services are contracted out to other individuals, companies or organizations. Such "outsourcing" has been reported particularly in relation to maintaining access to the World Wide Web for electronic journals (e.g. Fisher, 1997).

Is Electronic Publishing Less Costly?

In the last few years, several conflicting reports have been made that addressed the issue of the costliness of electronic publishing. The need to determine the kind of costs being assigned becomes apparent when an explanation for the conflicting conclusions of these reports is sought.

A cost comparison analysis. In 1997, Fisher reported a detailed comparative analysis of the direct and overhead costs of an electronic journal and a print journal. Both journals are published by The MIT Press. The electronic journal was the Chicago Journal of Theoretical Computer Science (CJTCS), a new journal that had published 244 pages in 10 articles. A 244-page issue of the print journal Neural Computation (NC) was used for comparison. Direct costs were presented as both total production costs and costs per page. In both total production costs, which included copyediting, proofreading, composition, printing and binding; and in composition cost per page, the print journal was 291% more expensive. The largest differences could be explained by comparing the costs of printing and binding, which was zero for the electronic journal, and of composition, which was 11% more expensive for the print journal.

When overhead costs were compared, however, the results were the opposite. Fisher reported total overhead per subscriber for CJTCS, the electronic journal, as $44,358, and costs per page as $182; but total overhead per subscriber for NC, the print journal, was $3,301 and costs per page were $14. Most of these differences could be attributed to the smaller number of subscribers for the new electronic journal compared to the established NC, and for the general, administrative, and production costs that were included in the overhead by the publisher. She also acknowledged that overhead costs were "affected by the small amount of content published in CJTCS over the course of 18 months of overhead costs compared with NC which published 12 issues over the same period of time" (p.6).

A cost analysis project. A detailed study of both the direct and shared costs of developing and publishing an electronic journal was made by Bot and Burgemeester (1998). The electronic journal, Journal of Comparative Law (EJCL), was developed by library and computer staff, and by faculty of the Tilburg and Utrecht Universities. A model for assigning costs was developed which contained the following direct cost categories: a) web site development; b) web site design; c) web site programming; d) organization maintenance; and e) miscellaneous.

The shared costs were assessed as the amount of use by EJCL within the categories of: a) volume of data transfer in both the network and Internet connection; b) server and client storage; and c) staff time in web site maintenance. Shared costs were calculated as percentages of the total use by the universities. Careful determination of the shared costs led to the conclusion that they were so small for EJCL they could be treated as zero in the analysis (Bot et al., 1998). Although the salary of an assistant editor was provided by one of the universities, the salary was included in the total annual cost figure for the journal.

Since the journal was developed during the project and was not fully established at the time of the report, the numbers of issues per year and articles per issue were not yet known. Based on a plan of four issues per year, the annual cost was estimated to be $20,084. Using the estimated annual cost figure, they determined that if an issue contained two to five articles, the estimated cost per article would be $2,511 to $1,004 respectively. Consequently, based on an estimated 400 to 600 "readers/subscribers," an annual subscription charge of $50 or $33, respectively, was projected. Even after including a 30% profit margin that is usual for print journals, the estimated cost of a subscription to EJCL would be $123. After comparing this price to the average $175 "annual subscription price of all law journals (including free journals and journals published by scholarly societies) in the collection of Tillburg University Library," they concluded "that electronic publishing is considerably cheaper" (p.11).

In summary, the answer to the question of whether electronic journals are less costly to produce than print journals may be a qualified "yes". Nevertheless, the costs to both types involve the creation of the first copy and distribution of the journal issue. Depending on how the first copy is produced for electronic journals, that is, if produced from the print version, it may cost more than for print journals. Further, even if the distribution costs are zero for electronic journals, some cost is involved in creating and maintaining the hardware and in administration (MacKie-Mason, Riveros, Bonn & Lougee, 1999).

In the new arena of electronic publishing, determining what the costs are for producing an electronic journal can be more art than science. As noted by Guthrie (1997), the issue of determining costs in fact is rather circular for any entrepreneurial venture and "requires a complicated assessment of interrelated factors. What are the costs for operating the entity? That depends on how much 'product' is sold. How much product can be sold, and what are the potential revenues? That depends on how it is priced. What should be the product's price? That depends on the costs of providing it" (p. 3).

The Economics of Recovering Costs

Costs involved in producing electronic journals must be recovered regardless of the numbers of journal users. Determining the best cost-recovery method, however, can become a philosophical issue considering that some electronic journals such as OJIN are free to the user, and in the past few years, free journals produced by scholars themselves have become numerous. Odlyzko (1997a) points out that technology advances now allow scholars to produce first copies themselves and to distribute their scholarly work without the large publishing infrastructure that is required by publishers of print journals. In fact, even for print journals scholars are being asked to submit their articles electronically after having "typeset" them themselves, so that with only a little more effort it is possible for them to produce entire journals. MacKie-Mason and Riveros (1997) state that scholars are willing to do this because they are motivated to work in the "service of scholarship" rather than in the "service of profit" (p. 2).

In a discussion of publishing economics, however, MacKie-Mason and Riveros contend that for-profit publishing is a value-adding business...

In a discussion of publishing economics, however, MacKie-Mason and Riveros contend that for-profit publishing is a value-adding business, that is, one that adds value to the authoring efforts of scholars. Publishers provide "many services to authors and readers" with one of the most important being "the marketing function. It is quite evident that scholarly journal authors are not writing for direct cash compensation, but to obtain readership (from which indirect compensation may follow). Good scholars are good at research, not at finding readers" (p. 2). Consequently, they believe for-profit publishing will not be overtaken by free publishing; rather both will coexist competitively. Nurses who desire to enter the electronic publishing field would do well to consider the economics of cost recovery. Following are workable strategies for cost recovery.

Competition: Applying an Economic Principle

MacKie-Mason and Riveros (1997) argue that an important principle of economics involving competition can be applied to the electronic publishing situation. In general the principle affirms that because there are low barriers to entering electronic publishing, competition will drive down returns and insure that publishers do not make excess profits. They will make profits only enough to cover costs and receive a return on capital investments. In a competitive market "consumer plus producer surplus" will be maximized. This means that authors, readers and publishers will receive a maximum value for the price paid; "readers get the best combination of quality, price and quantity possible given the technology and costs of production, and authors get the best readership possible" (p. 3). Whether or not this optimistic view of the electronic publishing environment can be supported realistically will be explored below.

Competitive Pricing to Recover Costs

To remain competitive in a highly competitive market, publishers would have to price their print journals at marginal costs, which, according to MacKie-Mason and Riveros (1997), may not be enough to recover the cost of the first copy. Electronic journals, however, potentially have lower first-copy costs, and also have opportunities to creatively structure pricing around value-added features such as hyperlinks, search functions, and differential access to articles and issues. In a discussion of pricing structures, MacKie-Mason and Riveros describe two new opportunities as "new service provision" and "price discrimination or nonlinear pricing."

New service provision. Examples of providing new services include such things as hyperlinking to references in an article, journal issue or even to sites external to the journal. Another example of a possible new service is one in which comments and critiques of readers are posted and responded to by authors. A third service that could be offered is one of "information filtering" where reader preferences for information are used to recommend unread articles based on ratings made by others with similar preferences (MacKie-Mason & Riveros, 1997).

Price discrimination. Price discrimination or differential pricing is a potential value-added feature that is based on the fact that readers seldom read a journal cover-to-cover; rather articles and even portions of articles such as the abstract and references are selected out. MacKie-Mason and Riveros (1997) state,

the key to differential pricing is to find observable characteristics (of users, or their uses of content) that are correlated with their willingness to pay.... such as user type (education, business, subscriber or occasional, determined from authentication records); use type (immediacy of access to current information, volume, full documents or components); and quality (display resolution; text-only or images; plain text or formatted). (p. 5)

In nursing subscribers might desire, for instance, information related to their clinical interests.

The PEAK project. Differential pricing based on the different value placed on journal content by readers was part of an experimental research project entitled Pricing Access to Knowledge (PEAK) conducted by researchers at the University of Michigan, and by Elsevier Science publisher (Bonn, Lougee, MacKie-Mason & Riveros, 1999; MacKie-Mason, Riveros, Bonn & Lougee, 1999). The project was focused on extracting additional value from the journals published by Elsevier Science and making them available electronically to 12 participating institutions (10 universities and 2 corporations). PEAK allowed "full text search and retrieval of the entire body of Elsevier content" (p. 2). Three different types of subscriptions were structured: traditional, generalized and per article. The traditional subscription provided unlimited access to a journal; the generalized subscription included unlimited access to a set of articles; and the per article subscription allowed access to an unlimited number of articles, but for a higher per article price than either of the other two (Bonn et al., 1999). The latter two subscription choices allowed users to choose among all the journals without having to subscribe to each one. Of the three choices, generalized subscriptions were the most popular (MacKie-Mason et al., 1999).

Several preliminary conclusions were drawn as the project was ending including: a) the numbers of articles accessed was affected by the cost to the user in terms of money, time and effort; b) most articles are read only occasionally, and "even the most popular articles are read only a few times" (p. 12); and c) the most recent articles were read the most frequently. Finally, MacKie-Mason et al. stated "the most important finding is that access can be expanded through innovative schemes like the generalized subscription while maintaining a predictable flow of revenue to the publisher" (p. 12). Descriptive data were not included, and so it is unknown if some of the users were nurses. Nevertheless, these conclusions are as likely as not to be applicable to nurse readers.

Other Ways to Recover Costs

Instead of subscriptions, non-profit organizations may use other strategies to cover costs.

Instead of subscriptions, non-profit organizations may use other strategies to cover costs. Walker (1998a) has described the experience of the Florida Entological Society (FES) to post electronically all articles published in the Society's refereed journal, the Florida Entomologist (An International Journal for the Americas), simply by storing articles as Portable Document Format (PDF) files. Since 1994, the articles have been posted on the web server of the Florida Center for Library Automation (FCLA) at no charge as a convenience to FCLA users. Recently the FES has added a search service to the articles at minimal cost to the Society, and has begun to scan and post all issues since the beginning of the journal, in 1917, on the Web. The $12,000 cost for the latter project is being borne by industry and the University of Florida.

Another service that is now offered by the FES is the posting of a PDF file to the author of an article. As opposed to print-published reprints, the supply of electronic reprints is never depleted. Walker (1998a) reports that this service costs $3 per page, and the costs are taken out of the $45 per page publishing charge that is paid by the authors. While not a common practice in nursing, such page charges often are assessed by non profit societies in other disciplines, and help keep subscription prices down while supporting other services to members (Walker, 1998b). Walker views page charges as a way to insure free net access to journals in the future.

The Economic Future of Electronic Journals

Odlyzko (1997b) has stated "technological predictions are notoriously hard to make correctly" (p. 1). Nevertheless, all predictions for the future of publishing include a recognition that electronic publishing is here to stay. In fact, Day (1999) has observed that "there is a growing feeling in the Internet age the printed journal may not be able to survive as a primary means of scholarly communication for much longer" (p. 2). A variety of opinions are offered, however, on the pace of change to a predominately electronic environment. Some are predicting a long time span will be required, while others are predicting rapid change.

Predictions of Change

A prediction of long term change was made by Whisler and Rosenblatt in 1997. They believed that there was a reluctance of individuals and libraries to change to electronic dissemination of scholarly information, and this would mean that capital investments in the technology needed to deliver such information also would be made slowly. Consequently, they predicted that electronic publishing would continue to evolve "for at least the next twenty years" (p. 14), and in "the near-term future, pressure to maintain both print and electronic dissemination [would] be great" (p. 1) .

A counterpoint to the question of reluctance is found in the results of a survey conducted by Rusch-Feja and Siebeky (1999) within the Max Planck Society, a German academy of science that contains 84 research institutes. Among the journals available to the respondents were the 1,100 electronic journals from Elsevier Science, 412 electronic journals from Springer Press, and 174 electronic journals from Academic Press. The 1042 respondents (a response rate of about 16%) indicated not only a high acceptance of electronic journals, but also "an unwillingness to return to print versions only" (p. 1). They indicated that if libraries had to reduce holdings, giving up print subscriptions and retaining electronic journal subscriptions would be the preferred cost savings. The advantages of electronic journals were listed as being able to obtain them more quickly than print journals, being able to access them from the desktop, and the ability to do full-text searching.

Brown (1999), however, conducted a similar survey with scientists at the University of Oklahoma and reported slightly different results. The purpose of the survey was "to understand how scientists are responding to the changing methods of information dissemination" (p. 1). The majority of the 49 respondents (a response rate of 61%) indicated they wanted to be able to continue to access print journal articles from the library's collection, but they were eager to expand their access to databases of electronic bibliographic information. The problem with such a "maintain-the-status-quo" attitude expressed in this study is that it does nothing to address libraries' serial pricing crisis, and ignores the opportunities offered by access to electronic journals. Odlyzko's (1999b) complaint that it "is the tremendous inertia of the scholarly community, [that] impedes the transition to free or inexpensive electronic journals" (p. 3), seems particularly relevant!

Schloman (2000) surveyed 619 faculty in schools of nursing concerning the credibility of print versus electronic journal publications in tenure/promotion reviews at their institution. Of the respondents, 47.6% reported the print and electronic publications have the same credibility in tenure/promotion reviews at their institution, while 51.8% believe print publications are more credible.

The Coexistence of Subscription and Free Electronic Journals

Publishers are rapidly moving their journals online, and probably for awhile there will be a transition period when print and electronic journals will coexist. Soon, however, only a few will remain in print-only format (Odlyzko, 1997b, 1999a, 1999b). Within the next decade, Odlyzko (1997b) predicts journals will exist only in electronic form. Likely two systems of cost recovery will exist at that point, one that includes subscriptions, site licenses, and "pay-per-view"; and one that allows free access (Walker, 1998b).

Subscriptions, site licenses and pay-per-view. Odlyzko (1999b) believes that when print journals are eliminated, publishers' costs will be reduced by 20 to 30 percent. Whisler and Rosenblatt (1997) caution, however, that:

there is no evidence that the emergence of electronic journals will change the fundamental economic problems in the cycle of scholarly communication in the short term, at least with respect to commercial publishers. The basic premise of these publishers is that they must both protect their current revenue base and secure guarantees to cover future inflation and increases in content. (p. 5)

Consequently, publishers are likely to structure innovative payment plans that include subscriptions, site licenses and pay-per-view features (Walker, 1998a). Actually, pay-per-view would be the only feature new to the present system, and would mean that those who do not subscribe, or are not affiliated to institutions that pay by subscriptions or site licenses, could access articles on a pay-per-view basis. Walker points out that the problem with this system is that it will fail if anyone is able to access electronic journals without being assessed a fee by one of the three methods. "Toll-gate" assessment would be impossible to enforce, however, given the motivations of scholars to provide free access to their work and the impossibility of preventing them from doing so given the ease of Internet distribution.

Free access. A rather revolutionary idea that has captured the interest of scholars recently is that of free access to scholarship dissemination.

A rather revolutionary idea that has captured the interest of scholars recently is that of free access to scholarship dissemination.

Two proponents of the idea, Day (1999) and Walker (1998a, 1998b) have traced back the beginning of publishing scholarly journals to 300 years ago when societies of scholars sought a way for members to inform each other of their scholarly work. The approach to the dissemination of scholarly information developed by commercial publishers of today, however, perpetuates "a 'Faustian bargain' made between author and commercial publishers, whereby authors trade the copyright of works in exchange for having them published" (Day, 1999, p. 3).

A proposal for a visionary project entitled E-biomed would return to scholars the ownership of their work. The idea for E-biomed was introduced to the scientific community in 1999 by Harold Varmus, Director of the National Institutes of Health. Varmus (1999b) stated the impetus for the proposal was the desire to "shape the future" of electronic publishing. He and his colleagues envisioned that E-biomed would be like "an electronic public library to which people go whenever the mood takes them to see what is available in the scientific literature" (p. 2). One of the assumptions of the proposal is that:

what scientists are about is generating results, typically paid for by the public or private funders; the scientists' objective is to get results of the research seen by as many people as possible. We have no interest in making money by publishing our results. We want to get them out where everyone sees them. It's good for our careers [and] it's good for the development of our science in the community....(p. 3)

Predictably, the proposal met with an outcry of protest from many in the publishing community. For instance, Delamothe and Smith (1999), ended an editorial in the British Medical Journal (BMJ) with the statement: "We may find that we lose our jobs or have to accept reduced salaries if the BMJ Publishing Group becomes unprofitable because of proposals like that from the National Institutes of Health" (p. 5).

In response to the concerns and suggestions, changes were made in the proposal, and on August 30, 1999, Varmus announced that PubMed Central was to be established in January, 2000. This Web-based repository will "archive, organize and distribute peer-reviewed reports from journals, as well as reports that have been screened but not formally peer-reviewed" (Varmus, 1999a, p.1). Changes from the original proposal included the timing of article submissions. Originally articles were to be added to the repository upon acceptance by journals, but after acknowledging the "concerns of publishers about financial consequences of rapid submission," the timing of submission of articles was changed so now that it could occur anytime, including after publication. Delamothe (1999) stated that publishers would determine how long the time needs to be from publication to submission to insure that subscriptions continued to be "financially viable." He noted, for instance, the content of an issue of Molecular Biology of the Cell will be offered to PubMed Central two months after publication. The shortfall expected by publishers might be addressed by collecting page charges from authors.

Summary

In conclusion, current numbers of electronic journals in nursing indicate a "low base level" in comparison to other disciplines, but "expanding interest and potential" (Murray & Anthony, 1999, p.160), and the current economic environment of nursing electronic journals mirrors that of the broader environment. In the short-term, for-profit and free models will continue to co-exist. Free access likely will prevail in the long-term, however, because as Walker (1998a) states, "authors and institutions will want it," scientific societies will provide it," and "commercial publishers will be forced to follow" (p. 4).

Author

Karen W. Budd, PhD, RN
E-mail: kbudd@kent.edu

Dr. Budd is an Associate Professor at Kent State University and an Associate Editor of Online Journal of Issues in Nursing. As a member of the editorial team from the initial discussions where the idea of an electronic journal was conceived, she has brought her extensive experience with Boards of Trustees and finance committees of various professional, university, and community organizations to the tasks of economic planning and marketing the journal.

, B. (2000). Nursing faculty and scholarly publishing: Survey of perceptions and journal use. (Currently under review).

---