Periodicals Price Survey 2005: Choosing Sides

One indicator is the sharpened rhetoric that signifies growing consensus about the nature of the ongoing "serials crisis." Librarians are quick now to challenge anyone who suggests that an infusion of new funds from their institutions will solve the problem. Higher education itself is in a funding crisis and in no position to rescue library materials budgets. Nor will the crisis be solved by lower rates of annual inflation for journal titles. For decades we focused concern on annual price increases while base prices for scientific journals, in particular, cumulated into such a mass that the entire scholarly communications system has become unsustainable.

As evidence mounts that the STM journals crisis has weakened other segments of the scholarly publishing market, including book publishers, virtually everyone concedes that change is necessary and that it must come quickly. Our "serials crisis" has, in fact, morphed into what some would call a crisis of public policy, pulling patient advocates, taxpayers, researchers, grant agencies, legislators, and antitrust lawyers into unlikely alliances with academic librarians-all united in pursuit of more open and affordable access to scientific information for the good of society as a whole. These alliances do not exclude scholarly publishers, many of whom welcome the benefits of opened access. Even publishers whose opposition is fixed would have to agree that the open access (OA) movement is pushing the market. But there is little agreement on how or to what extent.

There is no immediate answer to the dilemma. Journal prices are not dropping, and academic library budgets are not rising. OA business models, where publication costs are paid upfront and subscribers have free access, are too new to attract established publishers in great numbers. Legislative initiatives to force authors to archive publicly funded research findings on the web have stalled for the moment in the United States and Britain, although major private research foundations like the Wellcome Trust have begun to mandate web posting of research articles resulting from their grants. The advent of Google Scholar may jumpstart scholars into cooperating with the open access repository movement, but that is speculation at this point. So librarians study use data, correlate use to price, and cut every journal that is not essential, while publishers labor over complicated pricing models looking for a way to maintain revenue in a struggling market.

This year's periodicals price study looks at these and other factors that are reshaping the serials marketplace. Three Institute for Scientific Information (ISI) databases-Arts and Humanities Citation Index, Social Sciences Citation Index, and Science Citation Index-provide the 4,893 titles used in the study. These databases typically reflect the journal holdings of large research libraries. For smaller academic libraries, we include an analysis of 2,759 journals in EBSCO Publishing's Academic Search Premier.

Cost history for the study was pulled from EBSCO's database of 282,000 serial title listings. For practical reasons, the data are limited to prepriced titles (as opposed to standing-order and bill-later titles) that can be ordered through a vendor. The data are current as of February 11, 2005.

While strategies of reform are discussed in distant venues, publishers, vendors, and librarians deal with the frustrations of a changing marketplace that defies all attempts to standardize pricing or practices. Tiered pricing, by which the subscribing institution is slotted by type and size into a pricing matrix, has caught on at Project Muse, BMJ (British Medical Journal), and a number of society presses. Some publishers are trying a use-based pricing scheme that starts with an estimated cost for a bundle of content and adjusts future pricing based on usage over the year. Science uses this approach with its top customers.

Less transparent is the maddeningly obscure "quoted pricing" favored by Wolters Kluwer Health, the Nature group, and many others. The move to customized pricing mirrors the move by libraries to customized bundles of journals. Whether negotiated by one library or a consortium, these complex pricing schemes can turn seemingly simple orders for e-journals into full negotiations, drawing out the renewal process for weeks or months in many cases and stretching the renewal season into the winter months.

Despite years of outrageous periodical prices in some fields, the data on the extent of the problem continues to mount-and continues to shock. An exhaustive study commissioned by Oxford University Press (OUP) and conducted by a British university, Scholarly Journal Prices: Selected Trends and Prices reveals great disparity among the pricing behaviors of 12 prominent scholarly publishers (http://www.lboro.ac.uk/departments/dis/lisu). The report gives five years of pricing data, publisher by publisher and comparatively across six broad subject fields.

Some interesting factoids: Elsevier has the highest overall median price, based on its entire portfolio of journals. Cambridge University Press has the lowest. Elsevier also has the highest median price in each of the six subject fields, though Kluwer and Sage come close to Elsevier's median price in the social sciences and humanities. Sage achieved the dubious distinction of highest overall rate of price increase between 2000 and 2004 (94%). Librarians looking for comprehensive cost/value analysis for journals in biomedicine will find a wealth of data in the report.

Impact factors are a common measure of the quality of a particular journal based on the number of times its articles are cited by researchers in the field. The implication is that the higher a journal's impact factor, the higher the quality of the journal and the more the publisher can charge for a subscription. The OUP study looked for that correlation for biomedical titles and failed to find it, i.e., the priciest titles were not necessarily generating the highest impact factors. The study found a stronger correlation between number of pages and price. These kinds of data are essential tools in the current serials market. More are needed. Impact and usage data are changing the way librarians discern the value of journal subscriptions, and they are voting with their renewal lists.

Journal cancellations, particularly print duplicates, are epidemic in American libraries. A Publishers Communications Group survey in the spring of 2004 reports that 84% of respondents said they cancel print when electronic is available. Theoretically, print cancellations should help a publisher's bottom line unless advertising revenue is affected. The same can't be said for cancellations of e-journals. Publishers must have cringed when OhioLINK, the prototype of shared virtual libraries, announced in February that it will cut electronic titles in 2005 and 2006. Somewhat counter to the American experience, much of the rest of the world continues to take print. European libraries favor print because of an anomaly in the VAT structure that protects print, but not online, subscriptions from taxation.

Emerald Publishing gets this year's award for the most egregious breach of ethics. The publisher formerly known as MCB University Press has for nearly 30 years been republishing articles without notification that they were duplications-560 duplicated articles in 73 journals at latest count. Philip Davis of Cornell, who discovered the duplications, found articles were at times published simultaneously within journals in the same discipline, many of which carried hefty price tags.

As Davis points out, this practice raises serious questions for the libraries that unknowingly purchased the duplicates and for the peer-review process that may have been compromised by editors who had a commercial stake in the company. It will be interesting to see if this is an unwelcome anomaly or merely the first shoe to drop in a scandal that, either way, does the commercial publishing sector no good.

Proponents of open access experienced disappointment in November when the British government declined to enact recommendations from a parliamentary committee that would have put Britain into a leadership role in the OA movement. The committee had endorsed principles of open access and proposed strategies to get scientific information into the hands of the public. Various groups in England are working to implement some of the recommendations, but the government continues to waffle on the core issues raised in the report. Apparently they have chosen not to upset wealthy STM publishers, some of the largest of whom are headquartered in the UK.

A similar setback was experienced in the United States. Last summer Congress asked the National Institutes of Health (NIH) to develop a mandate to get its grant-funded research findings into PubMed Central (NIH's open archive) within six months of the article publication date. Powerful alliances for and against the proposed policy were formed, and arguments continued through the fall. A watered-down proposal, announced in February, removed the mandate for grantees to deposit articles and essentially lengthened the embargo period to 12 months for articles that are deposited, with some exceptions. Again, the commercial publishers seemed to win, aided mightily by a vocal core of society publishers that saw the proposal as a threat to business. While the new NIH policy is better than no policy, the fear is that it will make the 12-month embargo the de facto standard for some time to come, unnecessarily stifling the timely exchange of scientific information.

Meanwhile, the European Commission (EC) has begun a study of scientific journal publishing that will try to reconcile concerns about access with the economic interests of publishers. In announcing the report, the EC indicated that Europe leads America in producing scientific publications by a good 10%, yet the publications of its authors are cited less frequently. The EC sees a possible link between access and impact and apparently recognizes that the roots of change may lie in the market itself. The report is due later this year. Interestingly, the British Office of Fair Trading has stated it is waiting for the EU report to decide how it can best monitor the market that it pronounced two years ago as "not working well."

Despite rumors to the contrary, the OA movement remains a powerful catalyst for change. The number of journals in the Directory of Open Access Journals stood at 1,463 in February, double that of a year ago, with substantial numbers of peer-reviewed titles in fields like biology (61), chemistry (40), general medicine (164), neurology (31), public health (58), geology (22), philosophy and religion (48), education (110), and computer science (45). An ISI study found that the open access journals it tracks for impact are doing well, even when compared with very well-established traditional journals. As other studies of OA vs. toll-access articles emerge, indications are that OA literature will exceed toll-protected literature in both citations and downloads.

There are signs that commercial publishers are willing to experiment. OUP switched to an open access business model for Nucleic Acids Research, a top-rated journal. A number of hybrid OA experiments are underway that give authors a choice about when and how to make their articles free and open on the web, usually but not always based on an author's willingness to pay up front. Blackwell's Online Open service and Springer's Open Choice program are two early examples.

Cell Press is representative of another type of hybrid. Starting in January 2005, it is offering free access to the content of its e-journals once they are 12 months old. HighWire Press and others have been doing this for a long while, but it's the first time an Elsevier journal with the cachet of Cell Press has done so. These initiatives seem designed to keep authors within the folds of the traditional publishers rather than lose them to emerging journals that are fully open access.

One of the reasons the UK and U.S. proposals drew such fire from publishers is the fear of losing revenue if authors are allowed (or, worse, required) to post their articles in disciplinary repositories after they are published. Over 90% of publishers now allow authors to post articles on homepages or in institutional repositories. If authors took advantage of these rights, theoretically most scientific research articles would be available to anyone around the world for free, with minimal delays for embargoes imposed by some publishers. Since authors have largely not bothered to do that, publishers consider the risk to be minimal. Disciplinary repositories, like NIH's PubMed Central, may appear more threatening because they encourage direct and timely exchange among researchers. In physics, however, where arXiv has been a hugely popular repository for years, publishers have coexisted nicely with an open access repository, with no reported attrition in subscriptions caused by the access to preprints and postprints.

Publishers in other fields, however, aren't buying the idea that sales will remain stable or grow when the information is available on the web for free. Nature, for example, seems to have been spooked by the NIH proposal. A day before the expected NIH announcement, Nature imposed a six-month embargo on its authors, who had long enjoyed a liberal policy of self-archiving whenever they wished. Google's entry into scholarly communications may make these discussions somewhat moot if the added exposure drives up both use and publisher revenue.

Google Scholar, now in beta test, aims to deliver scholarly content to new audiences by crawling repositories and web pages where articles have been posted before or after formal publication in a peer-reviewed journal. Researchers may have been slow to adapt to the notion of self-archiving their scholarly output on the web, but Google may entice cooperation because of its proven success in reaching worldwide audiences.

The majority of STM publishers already open their metadata to crawling by Google and a score of other third-party web services because they recognize the value of the increased exposure for their content. CrossRef Search, a well-established linking service founded and owned by STM publishers, and Google Scholar are already collaborating. Beginning this month, Google Scholar will give top billing to a publisher's official link whenever multiple versions of an article appear. Publishers are already seeing new revenue from pay-per-view requests that come directly from the web, rather than through library-sponsored databases. It will be interesting to see how the presence of the ubiquitous browser affects the traditionally conservative habits of scholars and publishers who, up until now, have had a rather exclusive relationship.

Libraries are buying scholarly journals in a market dominated by a small number of publishers, the result of a decade of consolidations in the industry. STM publishers are scattered around the globe, as are their customers, providing a natural hedge against the kinds of currency fluctuations that used to wreak havoc with library budgets. Given a U.S. dollar that was weak against the pound and the euro for most of the year, for example, last year's price increases for European titles were lower than expected. Publishers may have been more concerned about cancellations than currencies. U.S. publishers also seemed to show restraint when they set prices for 2005. We may see a trend emerging of overall increases in the 7%-9% range, rather than the 9%-11% of past years. On the other hand, if the dollar continues to weaken, look for price hikes to return to the higher range for 2006.