College & Research Libraries—perhaps the most eminent journal addressing academic library issues in the United States—recently published an article that offers a clear and forceful summary of mainstream opinion about the transition from printed to electronic library materials that is currently underway. The author’s first two of five recommendations deal specifically with this transition: "complete the migration from print to electronic collections," he advises, and "retire legacy print collections" (i.e., discard paper copies). I stress that such recommendations are squarely within the mainstream of professional library opinion today; it is not my desire to criticize one librarian or one article. Yet before libraries decide to act further on such recommendations, there is more to understand about the economics of scholarly publishing.
Until about 1960, not-for-profit entities, such as scholarly societies and academic institutions, were the primary publishers of academic work. Theodore Bergstrom, economist at the University of California-Santa Barbara, reveals the changes that have occurred in one field, economics, since that time. According to Bergstrom’s research, 30 English-language economics journals existed in 1960, and almost all of them were published by not-for-profit entities. By 1980, the number of journals had risen to 120, with about half of these published by commercial entities. By the year 2000, there were 300 journals, and two-thirds were published by commercial entities (Theodore C. Bergstrom, "Free Labor for Costly Journals?" Journal of Economic Perspectives 15.3, 2001).
These two developments—an increase in the number of journals and an increase in the commercial publishers’ share of the market—can be traced across many academic disciplines. The increase in the number of journals is a result of the expansion of higher education since the 1960s, the increase in research funding, particularly in the sciences, and the increased emphasis on research and publication in all areas of higher education.
But what accounts for the increase in the proportion of scholarly journals now in the hands of the commercial publishers? One reason widely acknowledged is that commercial publishers have been better at innovation, identifying (and perhaps creating) new markets, and launching new titles. When higher education expanded quickly in the 1960s and 1970s and more money became available for research, it was mostly the commercial publishers, not the scholarly societies, that moved quickly to meet the new demand for publishing outlets. A second reason is that commercial publishers have actively encouraged the scholarly societies and other not-for-profit entities to relinquish their journals or at least enter into publishing contracts with them. Lacking the technical expertise, the capital, and the economies-of-scale of the commercial publishers, many scholarly societies and others have found such offers congenial, particularly as journal delivery began to move online.
Thus, the educational and research sectors have benefited in some ways from this transition to commercial publishers. But for some time now, economists and others who have studied the scholarly journal market have noted some disturbing facts. A study in 2001 found that the price of a given scholarly journal published by not-for-profit publishers was on average 50-75 percent lower than the price of a journal of comparable quality published by commercial publishers (Mark J. McCabe, "The Impact of Publisher Mergers on Journal Prices: Theory and Evidence" Serials Librarian 40.1/2, 2001). Another study conducted in 2001, of economics journals only, revealed that "cost per page" of commercially published titles was 6 times higher and "cost per citation" was 16 times higher than not-for-profit titles of similar quality. This researcher noted that while just 5 of the 20 most-cited economics journals were published by commercial publishers, these publishers "are absorbing the lion’s share of library budgets" (Bergstrom).
Case, 2001; from www.library.uu.nl/staff/ savenije/publicaties/ticer2004.htm
Prices also tend to increase after mergers and acquisitions. Industry consolidation is most notable in the scientific, medical, and technical fields, which also have the highest journal prices. Mark McCabe, an economist at Georgia Tech, has established that "a firm’s portfolio size [i.e., number of journals published] was positively related to journal prices, and that past mergers were associated with higher prices." When Reed Elsevier, one of the very largest scientific publishers, bought Pergamon, a smaller publisher, the average price of the former Pergamon journals rose 27 percent and the Reed Elsevier journals rose 5 percent. The merger of West and Thomson, two publishers of legal material, resulted in a 30 percent increase in the company’s post-merger prices. This occurred despite a review by the U.S. Department of Justice, which mandated the company to divest itself of certain "content overlap" in order to avoid anti-competitive practice.
One might argue that although prices have risen sharply, they fairly represent the cost of doing business. But the evidence does not lie along these lines. First, the profits of these publishers over the past many years have been in the 15-40 percent range, indicating that revenues greatly outstrip costs. Second, the publishers of other scholarly journals have not raised their prices as rapidly, yet presumably have similar costs. And third, the price of scholarly books, as opposed to journals, has not risen as rapidly, but again we presume that book publishing has similar costs.
Why is it that in a market economy, such outsized prices, and therefore outsized profits, have been occurring? One reason is that the actual consumers of these scholarly journals usually are not the ones who pay for them, and are thus not price-aware, much less price-sensitive. The primary consumers of a scholarly journal are its editors, referees, authors, and readers. While each scholar, by choosing to participate in a given journal, is implicitly choosing to support that journal’s business model as well, librarians have generally been ineffective at bringing this fact to scholars’ attention and helping them see how it directly affects their own self-interest and the well-being of scholarship as a whole.
Second, we can note that the market for a scholarly journal is quite different from the market for a cell phone or a car or any other usual consumer item. A given journal has a given level of prestige, usually developed over decades, built upon the reputation of its editors, referees, and authors, and upon the quality and impact of its articles. It is extremely difficult for one journal to usurp another journal’s status. Bergstrom has pointed out that authors, editors, promotion and tenure committees, and librarians are playing a coordination game: "In a coordination game, each player chooses an action from among several alternatives and each player’s payoff increases with the number of other players whose choice is the same as her own. An equilibrium is an outcome such that given the actions of others, no player could individually benefit by switching to another action. Coordination games commonly have many different equilibria, in each of which all players choose the same action. An outcome can be an equilibrium even though there is another equilibrium that would be better for everyone and which could be reached if all players were to change simultaneously to the same new action."
In other words, talented authors and editors "coordinate" at prestigious journals, prompting more authors to seek to be published therein, in turn prompting libraries to subscribe and make these journals available, in turn making these journals likely to be read and cited more, in turn reinforcing these journals’ prestige and impact. A status quo is entrenched, and, absent a sudden re-coordination at another journal, this state of affairs may persist even when "another equilibrium…would be better for everyone."
House of Commons, Science & Technology – Tenth Report, 2004 — www.parliament.the-stationery-office.com/ pa/cm200304/cmselect/cmsctech/399/39907.htm
Third, the journal market differs from the market for other consumer items in that it lacks a substitution mechanism, one of the features of most markets that help to keep prices competitive. Access to one journal is not equivalent to access to another, even when they have equal prestige and are devoted to the same topics. If important work is being published in Journal A, to which a university’s faculty members want to have access, it will not do for the university library to offer instead the important work being published in Journal B. The professors may indeed appreciate having access to Journal B as well—they want to have access to as much as possible—but one journal cannot serve as a substitute for another in the way that a cell phone available from one company can plausibly substitute for a cell phone available from another. Economists refer to such conditions, in which demand for a particular product is only weakly affected by price increases, as an "inelastic market."
Faculty members, as a whole, wish to have access to as many scholarly journals as possible. Academic libraries therefore attempt to subscribe to as many journals as their budgets permit. This, coupled with the fact that journals, unlike books, exist as publications that extend through time (as a series of issues published consecutively and without anticipated completion), has prompted a rational and highly effective strategy on the part of the large commercial publishers. Mergers and acquisitions have been one part of the strategy, since they increase the size of a given publisher’s portfolio of journals. Another part of the strategy has been to exert pressure on libraries to transition from printed to online journal format. As journals have moved online, the large publishers have created databases that can be marketed as products in their own right, quite apart from each journal title that these databases may include. For the time being at least, libraries are still permitted to subscribe to individual titles, but publishers are now pricing their products in such a way that it is to the library’s advantage, on a price-per-title basis, to subscribe to databases composed of particular sets of journals, or to all of the publisher’s journals, rather than to select title by title. Because libraries desire to subscribe to the greatest number of journals possible within the constraints of their budgets, and pricing is arranged so that very little savings are gained through selective cancellation, we can see why most libraries have accepted these deals, which are known as bundles.
What does journal bundling do for the publishers? First, it replaces old-fashioned subscriptions with contracts. In most cases, bundling deals are offered to libraries as multi-year contracts. As a condition of making the deal, the library is usually locked into a series of annual price increases, laid out in the contract, and prohibited from cancelling any of the publisher’s titles during that period. Contract law supersedes Fair Use and other provisions of copyright law, enabling the publisher to define how the content can be used. (And with any electronically delivered content, it is important to remember that libraries usually pay only for access, not for ownership. Archiving and other long-term rights to the content vary by contract.)
In addition, the price of the journal bundles is not the same for all customers, but is negotiated according to a given library’s apparent ability to pay as revealed by that library’s subscription history and other factors. One library pays one price to gain access to a given bundle of journals, while another library may pay a higher or lower price for identical access. For this reason, bundling contracts may include a nondisclosure clause, prohibiting libraries from discussing with one another the details of the contracts they have signed. Libraries’ negotiation powers are weakened by their inability to assess the market value of the journals they are purchasing.
And finally, bundling gives the publisher a larger share of the library’s overall budget than it may have held in the past, and, as mentioned earlier, it locks in that share for the entire multi-year contract. This necessarily reduces the funds the library has available for buying desirable content from other publishers. Limiting competitors’ market share is one way that a company may prosper. Another way is to increases profits. McCabe has argued that bundling enables publishers to increase profits by exploiting the market inelasticity mentioned above, since the publisher’s more-desirable titles can be bundled with the less-desirable ones. This increases the subscription rate of the less-desirable titles, which often carry higher profit margins.
Some of the smaller publishers, including some society publishers, have seen the advantages of the bundling model and are beginning to use it also, but the smaller size of their portfolios make their bundles less attractive to libraries. A Morgan Stanley report in 2002 noted that "Market leader Reed should outperform the market…as libraries trim peripheral suppliers who can’t bundle journals as effectively." To "trim peripheral suppliers" is to cancel subscriptions to the journals published by the smaller publishers. In the words of one commentator: "The journals produced by small publishers may enter a vicious cycle whereby as they lose subscriptions more quickly, the dissemination and circulation of the work published in them is reduced, resulting in a fall in impact factor. As the impact factor drops, their position on librarians’ ‘must have’ lists falls, leading to even greater cancellations, reduced dissemination, lower visibility and exposure, falling usage, further decreased impact, etc., etc. Conversely the inessential, low-impact journals from large commercial publishers will have expanded dissemination through [bundling], leading to greater impact and a strengthened position" (David Prosser, "Between a Rock and a Hard Place: The Big Squeeze for Small Publishers" Learned Publishing 17.1, 2004). Under such conditions, scholarly societies must feel great pressure to transfer their titles to the large publishers.
McCabe points out that if the very large publishers continue their mergers and acquisitions, with one STM (scientific, technical, and medical) publisher gaining access to 50 percent or more of the journals, "effectively the publisher could say, ‘Either spend all of the [library] budget on my 60 percent of the STM content, or you can go and spend your entire STM budget on the 40 percent of stuff out there that is not in my bundle.’… This would be an offer [libraries] couldn’t refuse." When we consider that the actual cost of producing the research is borne by the public, through tax-funded government agencies, or by the very institutions whose libraries are now buying back the results, McCabe concludes: "Let’s be clear: We are talking about a true market failure" (qtd. in Richard Poynder, "A True Market Failure" Information Today, Dec. 2002). McCabe notes that such asituation can occur only because of the transition to online journals, which has made bundling possible. In a print-based market, a library could always select journals to subscribe to on a title-by-title basis, even if the majority of these were held by one publisher. "Once publishers eliminate the print option," he notes, "the temptation to pursue a super merger strategy may be hard to resist."
In sum, "the market for scholarly journals as currently constituted appears to be one in which, because of its distinctive features, price competition is weak or nearly absent; a few dominant suppliers extract huge profits, and few of the ‘self correction’ mechanisms are present if markets are to serve the public interest" (Richard Edwards and David Shulenburger, "The High Cost of Scholarly Journals" Change, Nov.-Dec. 2003). In fact, colleges and universities are caught truly in a no-win situation by the current arrangement of the market. Even if they are able to increase their libraries’ acquisitions budgets faster than the general academic inflation rate, the commercial publishers only raise prices further. This is a logical outcome. As Edwards and Shulenburger observe, "In a market in which demand is inelastic, the reaction to more purchasing power…is simply higher prices." While it is possible for libraries to cancel subscriptions, publishers have calibrated their pricing in such a way that a 1 percent increase in price results in a subscription decline of only 0.3 percent, thus enabling them still to come out ahead. And as shown earlier, bundling contracts make it harder for libraries to cut subscriptions.
What is good for the large commercial publishers is fundamentally at variance with what is good for scholarship. What is good for scholarship is a wide variety of journals from a wide variety of publishers, each one offering valuable work to as wide an audience as possible. Helena Norberg-Hodge, speaking of changes occurring within indigenous communities, remarks that "increasingly, people are locked into an economic system that pumps resources out of the periphery [that is, out of the local community] into the center [the industrial center]…. Often, these resources end up back where they came from as commercial products…at prices [that the original community] can no longer afford" (Helena Norberg-Hodge quoted in Wendell Berry, Sex, Economy, Freedom & Community, Pantheon, 1992). Is this not an apt description of the changes occurring in academic communities also?
Those in academe who find this trend disturbing offer various possible solutions: open-access publishing, renewed support for university presses and scholarly societies, improved understanding and management of authors’ copyrights, and so forth. My own view is that the success of these efforts will be partial at best. The transition from printed to electronic journal delivery, currently underway in our academic libraries and urged upon us by almost all of what we read and hear, has been an essential step toward the increased market dominance and profits of the large commercial publishers. Electronic delivery, making use of certain inherent features in the nature of academic journals, has enabled the creation of new products (bundles and databases) and sales of them through contract, which have in turn weakened the smaller publishers and the libraries that buy the products. This is true regardless of the merits of electronic delivery as measured by other criteria (shelf-space savings, ease of use, and so forth). If this argument is correct, then our acquiescence in a complete transition to electronic journal delivery without probing these economic realities is an inadequate response to the conditions we face.