This commentary proposes a model for research institutions/libraries and large commercial publishers to consider adopting as a means of enabling wider access to research articles from subscription journals. This model is called a Read & Let Read deal (R&LR).
R&LR is intended as a direct alternative to the transformative agreement, the “predominant mode for institutional funding of open access publishing,” in which libraries shift payment “away from subscription-based reading and towards open access publishing” (p. x; Hinchliffe, 2020).1 The principle that informs R&LR is that open access is good but not at any cost, just as paywalls may be a hurdle to readership but not if pre-paid at fair prices.
In R&LR, payments from research institutions/libraries only go toward subscription-based reading, in a novel way, intended to provide maximum readership for users inside and outside the subscribing institution without necessarily causing loss of revenue to a partnering publisher.
R&LR
- •
Read: A research institution/library will prepay a publisher a base amount each year according to the total number of articles that institutional users downloaded during the previous year, multiplied by two, which covers the base amount (as described) plus any additional downloads made above the base amount during the coverage year.
- •
Let Read: Any downloads unclaimed by the institution during the coverage year are donated in the following year to any user on the Web.
- •
Updated valuation: All subscription-only articles cost $0.50 to download.
Every Title
The publisher will be asked to make their entire portfolio of articles eligible to prepayment at a half-dollar valuation, regardless of the journal. This must be insisted upon so that the institution/library can discontinue participation in mechanisms that may purposefully or incidentally reinforce notions of journal prestige based on a brand or algorithm. Some methodologies may value articles from certain journals at considerably more (or less) than half a dollar, but the institution/library will pay one flat article cost. It will be in the interest of a publisher to provide a wide catalog to maximize the base amount (owing to download volume) for the contracted institution/library.
In Practice
To better understand this deal, imagine an R&LR between Elsevier and the University of California (UC) system. RELX (parent company of Elsevier) estimated that UC researchers download over 11 million Elsevier articles per year (Marti, 2019). Some hold skepticism of publisher-provided download metrics (Wood-Doughty, 2019), but, if we use this figure for the sake of argument, we come up with a total package of $11 million,2 which would buy 11 million article downloads for UC researchers during the next year3 plus 11 million downloads for anyone on the Web during the following year.4
In Comparison
The UC and Elsevier $10.7 million contract announced in 20215 includes a way for UC libraries to subsidize article processing charges (APCs) on UC-authored works at an average cost of $2,449, according to one estimate (Anderson, 2021). If UC swapped this arrangement for an experimental R&LR, they would be asked to pay $400,000 more than their present deal, flip no articles to open access on the publisher site, and would need to justify to their board of regents why half of a multimillion-dollar contract is not immediately benefitting UC. Where is the upside?
Assume that UC will eventually pay out $5.5 million on APCs, and then compare what $5.5 million could buy under each deal in terms of gained reader access outside of the UC firewall:
- •
UC and Elsevier 2021: A total of $5.5 million spent on $2,449 APCs opens 2,246 UC-authored articles published by Elsevier to the global scholar system.67
- •
R&LR: A total of $5.5 million spent on $0.50 articles gives global researchers 11 million opportunities to download any Elsevier-published article of their choosing.8
In addressing access barriers for global researchers, the 11 million “just-in-time” articles that R&LR buys would be quantitatively more impactful than 2,246 “just-in-case articles” bought under a pay-to-publish model. When researchers download an article through a library or from the Web, they may not care whether it is open access or paid access but rather whether they have any access at all. By creating 11 million annual download opportunities, an institution like UC would ensure a likelihood that every researcher who wanted access to a UC-authored article in an Elsevier title would be able to get it.9
A strongpoint for R&LR is cost-efficiency. However, it is certainly possible that a single UC-authored article could be downloaded from Elsevier more than 4,898 times. Under the R&LR $0.50-per-article valuation, if one single UC-authored article were downloaded 4,899 times, UC would effectively pay $2,449.50, which is more than the $2,449 article processing fee; in those cases, paying the APC would be the more cost-effective route. But, in order for R&LR to work, the publisher will need to share usage data with the institution that will be making their annual payments based directly on that usage. With this data, the institution could choose to identify which articles appear on trajectory to pass the point of cost-efficiency under R&LR and prioritize those for green open access archival in order to provide a second method of access to the article that would not count toward the R&LR download total.
Conclusion
The benefits of R&LR outweigh those of transformative agreements. Flipping paywalled research to open access research is a noble goal, but paying APCs to do so is an inefficient method toward universal access that ultimately benefits well-resourced institutions, thereby perpetuating inequality in the research system. Flipping paywalled journals to a model in which neither the reader nor the author pays is a positive step, but, as long as scholarship continues to specialize in subdisciplines, commercial publishers will continually initiate new titles to accommodate author demand. Building out new journals on noncommercial infrastructure should be our long-term goal. However, while large numbers of scholars still use commercial publishers, the focus of research institutions/libraries negotiations with such publishers should be on cost efficiency models that create downstream benefits and minimize downstream harms.
Under R&LR, publishers (like Elsevier) could continue to receive a level of income similar to or higher than what they presently enjoy, which should make them amenable to this plan. Researchers affiliated with a subscribing institution (like UC) would be granted all the research articles they want from a partnered publisher without delay, which should keep a participating campus satisfied. Furthermore, researchers not affiliated with a subscribing institution would gain a likelihood of access through one of the (potentially) 11 million prepaid article downloads put onto the Web annually, which addresses the most immediate concerns of global researchers. If 11 million prepaid article downloads would be the result from one R&LR alone, just imagine if there were two.101112
Notes
- Institutional pay-to-publish models only apply to articles when the “institutional affiliate is the corresponding author” (p. 1; Ghamandi, 2020), which may be “good for the groups who sign them but bad for the overall system” (p. 2; Jester, 2021). [^]
- Based on 11 million downloads (per RELX estimates) at a cost of $0.50 apiece, adding up to a base amount of $5.5 million. Multiplying this base amount by two, the total deal becomes $11 million. [^]
- With access to tap into an additional 11 million downloads, if uses run over the base amount. [^]
- Subtracting any UC downloads that ran over the base amount in the prior year. [^]
- https://web.archive.org/web/20210316193706/https://osc.universityofcalifornia.edu/uc-publisher-relationships/uc-and-elsevier/ [^]
- 5.5 million divided by 2,449 is 2,246, rounded up. [^]
- 2,246 articles represent 9.73% of the 500,000 articles that Elsevier publishes each year (https://www.elsevier.com/__data/assets/pdf_file/0005/1095953/Fast-Facts.pdf). [^]
- Because R&LR applies to all articles from a publisher, readership will extend to UC-authored works in which the UC author is not the primary contact, thereby making R&LR of even greater benefit to UC. [^]
- Every time a UC author was cited by a global researcher, thanks to the access of R&LR, the UC author would be able to take heart in the fact that the decision to cite was based on merit rather than an inequitable advantage based on the ability of an author or institution to pay an APC. [^]
- These are just the primary benefits. There may be secondary benefits, such as plugging data leakage lost to piracy, alleviating workloads of interlibrary loan offices, and lessened time explaining journal subscription churn to campuses. [^]
- To be clear, this model is not under consideration anywhere at the time of writing. [^]
- The author thanks Kevin Hawkins, University of North Texas, for early feedback. [^]
References
Anderson, R. (2021, March 25). Six questions (with answers!) about UC’s and Elsevier’s new transformative deal. The Scholarly Kitchen. https://scholarlykitchen.sspnet.org/2021/03/25/six-questions-with-answers-about-ucs-and-elseviers-new-transformative-deal/
Ghamandi, D. (2020, January 3). Uncensored scholarly communication fragments. Humanities Commons. doi: http://dx.doi.org/10.17613/xmj2-zd67
Hinchliffe, L. (2020, April 8). Seeking sustainability: Publishing models for an open access age. The Scholarly Kitchen. https://scholarlykitchen.sspnet.org/2020/04/07/seeking-sustainability-publishing-models-for-an-open-access-age/
Jester, A. (2021, April 22). Discomfort with gold OA. Commonplace. doi: https://doi.org/10.21428/6ffd8432.913aef43
Marti, D. (2019, July 18). Commentary: UC Library’s hasty implementation of a complex plan means academic researchers lose. CalMatters. https://calmatters.org/commentary/2019/07/uc-vs-elsevier/
Wood-Doughty, A., Bergstrom, T., & Steigerwald, D. G. (2019). Do download reports reliably measure journal usage? Trusting the fox to count your hens? College & Research Libraries, 80(5). doi: https://doi.org/10.5860/crl.80.5.694