Last week, the University of California system terminated its license with Elsevier. There has been a great deal of attention to California’s efforts to reach a Publish & Read (P&R) agreement. The what-could-have-been of this deal is interesting and important. But I wish to focus today on the what-no-longer-is of scholarly content licensing, focusing on the big deal model of subscription journals bundled together on a single publisher basis for three to five year deals. In the eyes of major libraries in Europe and the US, the value of the big deal has declined. As a result, we are moving into a new period, in which publisher pricing power has declined and the equilibrium price for content and related services is being reset. What is the principal culprit? I will maintain today that we must look in large part to what publishers call “leakage.”
The University of California was extremely effective in framing the negotiation as being about the desire for a pathway to open access. We have seen in other efforts to negotiate P&R deals, such as that led by Projekt Deal in Germany, that Elsevier is unyielding in insisting that it must take a global perspective on the economics of this transformation.
From their Pay It Forward study, California negotiators were aware that the UC system would be responsible for a far higher share of publication costs than of subscription costs were the same global publishing cost basis to flip to open. California negotiators nevertheless took the negotiating posture that it should experience “reduction in costs” from Elsevier for the move. While Elsevier was unwilling to provide a P&R model that met UC’s price goals, it certainly would have been willing to offer a continuing subscription big deal to UC.
In choosing not to move ahead with a continuing subscription deal at all, California has made an extremely strong statement. While its negotiating position was to seek a P&R deal, in the absence of such an agreement it felt comfortable to walk away from Elsevier. Whatever its strategic goals, California like Germany and Sweden before it clearly felt that it can do without the subscription bundle at all.
Of course, Elsevier is quick to argue the value of its offerings. It argues that the number of subscription articles published is growing faster than any proposed price increases for subscriptions, lowering the cost per article published. And it notes that usage continues to rise, driving down the cost per article downloaded.
Against these arguments, California, through its cancellation, has nevertheless maintained its position unambiguously. It does not need ongoing journal subscriptions through ScienceDirect. Put another way: A major customer’s perceived value in the product offering has declined. Elsevier apparently no longer has the pricing power it once could assert.
The source of the value decline is no mystery. Joe Esposito argued more than a year ago that “Sci-Hub is an unacknowledged reserve army prepared to enter the battle with publishers,” noting elsewhere in the piece that time “is not on Elsevier’s side.”
But Sci-Hub is not alone. Sci-Hub is one of a series of services through which content is “leaking” out of publisher sites through to users. While some of these sites are illicit and pirate, others like SSRN and institutional repositories are accepted parts of the ecosystem, and still others are like ResearchGate, whose intentions vary by observer.
As a result of this combination of services, of whatever legitimacy, publishers are noting that a decreasing share of the usage of a given article is coming from their own content platforms, which are the primary vehicles through which they monetize the library channel.
Some of the larger publishers have attempted to generate estimates of the amount of leakage they are experiencing. While these models are rough, and no one wants their name associated with their estimate, I have heard estimates that suggest publisher usage numbers could be at least 60-70% higher if “leakage” was included in addition to their on-platform usage statistics. This includes “green” options through a variety of repositories (including some that are operated by publishers themselves in addition to library and not-for-profit repositories), materials on scholarly collaboration networks, and through piracy. The share of leakage among entitled users at an institution with a license is probably lower than this estimate, but it is likely well in the double digits.
I am in no way arguing against green models. Indeed, publishers have largely become comfortable with green open access. I am simply observing that these percentages are beginning to add up.
Publishers analyze usage data carefully. Some report it as part of their financial reporting. For example, RELX reports that Elsevier’s global article downloads grew from over 900 million in 2017 [PDF page 14] to 1 billion in 2018 [PDF page 14]. Let’s assume that these downloads represent organic growth and do not include the additional traffic derived from the acquisition of bepress, which took place in this period. Allowing that ScienceDirect traffic has been growing globally, is it possible that emerging markets account for a far higher share of this increase than do mature markets? Might some major research universities even be seeing subscription article usage plateau or decline? There is much here we do not know. Universities are negotiating, in any case, based on their own usage and associated perceived value, not the global aggregate.
The implication is this: publishers license their subscription content for on-platform use at premium prices and usage globally may well be growing; however, also globally, a significant portion of usage now occurs for free, illicitly or not. The key question is how this all adds up at the institutional level. In individual negotiations, these patterns cannot help but raise questions about what the equilibrium price should be for this content.
Having terminated its journal licensing relationship with Elsevier, California is undertaking to ensure that it provides alternative forms of access for its researchers. Its recommendations include a number of services and tools from which its users can choose, depending on the nature of their needs. Most of the alternatives are tools and repositories designed to access “green” versions of articles. In addition, as has become the pattern in such documents, it makes sure to reference Sci-Hub, ensuring that brand recognition grows, while making clear not to endorse it. Interlibrary loan and document delivery are also offered as alternatives, and, while they probably will not be the greatest source of articles to California users, they are a key part of the calculus.
California offers various forms of interlibrary lending, which often requires labor costs, not only for the borrower but the lender as well. I had heard some concerns from other libraries that California will thereby be, in one way of thinking, free riding on its peers in a way that is not sustainable should others wish to follow California in cancelling the big deal.
In my view, document delivery is the more important factor. Urgent needs for journal article fulfillment, such as those that a clinical provider might face in a patient care situation, are among those that have most deterred libraries from cancelling big deal licenses. For these urgent needs, California is offering document delivery.
California’s document delivery provider is Reprints Desk. The basic model, which is similar to a service offered by the Copyright Clearance Center, is that a user can place a request, directly or mediated through their library, for an article. For many publishers, including Elsevier, the document delivery service can provide access to an article with little to no delay. The library pays the service and the service passes along a portion of the fee to the publisher. If the service is configured to trigger when a discovery workflow ends without other forms of access to a licensed or open version of the article, these services can come within striking distance of replicating the user experience of having access to the licensed big deal. Presumably, California has concluded that the inconvenience or delays imposed on article access by the overall set of alternatives — including document delivery for those important urgent needs — does not create an undue inconvenience to their users.
Both standard interlibrary lending and document delivery impose additional costs on the university as demand for them increases, as can be expected following a major cancellation. Depending on volume, interlibrary loan could require some additional labor, both within the California system and at other institutions, while document delivery incurs direct expenditures with the provider. While the demands for the services may not be much greater in the early period following a cancellation, especially in a case such as this one when backfile years remain available through the publisher, over time a greater amount of content will become unavailable and demand for lending and delivery services may be expected to rise. As one of my readers explained to me, costs will grow over time and so, “At a certain point it seems likely that UC will be compelled to pick up subscriptions to individual titles as a cost-control mechanism.” Presumably, California has modeled its expectations for levels of use and has concluded (as did Germany) that the articles that will for the foreseeable future be requested through these services will result in a lower cost than the big deal license that Elsevier was offering.
There is an important likely conclusion here: the inconvenience and cost of providing an alternative access system is not expected to net up to the value of the big deal price. This is important because the per-article prices were deliberately set in such a way as not to undermine the economics of the primary publishers’ licensing businesses. As a result of content leakage through other sources, the amount of content needed by a library through these document delivery services to make the cancellation logic work has likely dropped considerably from what once would have been the case. As part of this overall mix, it seems these document delivery services are something of a game changer for big deal negotiators.
Thus, for all the discussion about open around this latest round of failed negotiations, I believe the first question is about value and price for the subscription business. To be clear, the work of publishing has value, and I respect the importance of the curatorial, editorial, technical, and production functions, among others. Still, my working hypothesis is this: the pricing power among the largest publishers for their big deals has declined and the equilibrium price for journal subscriptions may therefore decline as well.
First, Germany and Sweden, and now California, have used this perceived decline in value to push publishers towards what would ultimately replace a subscription model with a publishing services agreement. In the case of Germany, Wiley appears to have accepted this proposal, prepared to gain market share for its existing open business and even creating a flagship Deal-branded journal that could position it as a leader in showcasing German research, while Elsevier is frozen out, one must imagine, not only as a content provider but also increasingly as a publisher of German scientific manuscripts. California would seem to represent a gold rush opportunity for Wiley, Springer Nature, or another publisher interested in acquiring market share for publishing services among one of the world’s research powerhouses.
Elsevier has a global footprint, and although it appears to be overly reliant on North America for its publishing revenue, even losing what may be its largest North American account may not amount to a material impact on its bottom line. But, California is widely perceived as a leader among American library groups, and the rumblings about what seems possible this week that did not seem likely early last week are hard to miss if you speak with consortial leaders.
For this reason, I am surprised that the strategists at Elsevier did not manage to construct some kind of creative deal, bundling parts of their platforms business or reducing the price of the expiring license agreement, to retain California. Surely there must be a way to escape the logic of P&R negotiations, which Elsevier strategists understandably hate.
Of course, any “contagion” of cancellations will spread slowly across North America, not rapidly, since not all consortia have the same ability as California to speak with a single voice, and moreover as prices adjust to match the new equilibrium. After all, how many lost agreements can Elsevier withstand?
What comes next will assuredly be of interest. While we have seen Germany able to maintain a common position in its negotiations with Elsevier, it remains to be seen how strong the California schools will prove to be. Will any of them choose to license individual journal titles on their own, breaking the big deal but acknowledging the value of some of its contents? Will faculty members rise up in revolt and cause the consortium as a whole or some of its individual university members to take a different approach? Will Elsevier return at some point with a game changing offer? If California can maintain control of its stance to the degree that Germany has been able to do, it will suggest that its position on the value of the journal bundles can hold.
I maintain the belief that leakage is a key factor in reducing the pricing power of the major publishers. With legal action thus far going nowhere fast, and the green cat out of the bag in any case, pragmatists may see only a few prospects for how to address leakage.
First, to the extent that document delivery services are the game changer that some have speculated they may be, publishers may want to understand this particular part of the marketplace better. Not all librarians are equally sanguine about these services and their methods. Publishers may wish to dig into this landscape further and explore options.
Second, publishers are already looking at the syndication model, as a strategic initiative to co-opt content leakage and turn much of it into COUNTER compliant usage that can be applied to the library channel. Several major publishers are actively exploring syndication and considering the trade-offs it involves, and Springer Nature and ResearchGate just launched a pilot. It seems that syndication can help publishers maintain the value of their journals — and thereby preserve the library channel.
For many, leakage will only seem to be a short-term concern as we move inexorably towards a more open future. But brief though today’s transitional period may be, libraries will increasingly leverage this new dynamic to their negotiating advantage. Publishers reliant on the subscription model will need to grapple with leakage — in terms of their platform strategy, related technical issues, product bundling, and negotiating tactics.
I thank Christine Wolff-Eisenberg for encouraging me to write about this topic. I also thank her, along with Rick Anderson, Brandon Butler, David Crotty, Lisa Janicke Hinchliffe, Kimberly Lutz, Mark McBride, and Doug Way, for commenting on an earlier draft.