Click here for the Dutch version of this article
The new ‘big deal’ with publisher and data company Elsevier was meant to deliver more open access, at a lower cost. At least, that was the intended outcome of the Dutch universities when negotiations started early 2018. Having learned from the previous ‘gold deal’ with Elsevier, and with the added benefit of growing international transparency of publishing deals, the institutes’ stance had hardened compared to 2015 when the previous deal was struck.
Recently, Elsevier, the Dutch institutions and research funding organizations finally presented the new agreement, combining publishing and data services. It is a contract that raises the question whether these initial conditions were met in a meaningful way.
In summary, the deal boils down to Elsevier offering Dutch (corresponding) authors open access publishing options in nearly all of its scientific journals. However, a number of journals from the Cell and Lancet families have been excluded from the deal, for now. Additionally, both sides agreed to work towards the creation of infrastructure for research data and information, and to enter into ‘open science’ projects. All of this comes at a price of € 16.4 million per year.
Going by headlines in the national newspapers, one would get the impression that the Dutch are making a giant step forward on the path to open access and open science. But is this really the case? ScienceGuide asked experts and (co)negotiators and scrutinized the fine print of the contract. As it turns out, parties have agreed on very specific definitions of open access and open science, with vague articles in the agreement to underpin them. Agreements that are at odds with earlier statements on open science and on rewards and recognition.
Lack of clarity
The aim of a contract is usually to provide clarity on what has been agreed. During negotiations, parties try to get their mutual obligations as clear as possible, and then confirm those in a written contract, ultimately in the hope there will never be a need to fall back on it.
This is definitely not the way the contract between Elsevier and the Dutch science representatives reads. An important reason for this is the fact that in the contract, two types of services are tied together. On the one hand, the contract contains a – by now – relatively common ‘read and publish’ deal for academic publications. On the other hand, it introduces a unique ‘open science and services agreement’ that is aimed at executing projects with (meta)data, for example to support management and policy decisions.
Ever since the plans to contractually combine these two activities were leaked, this approach dubbed ‘tying’ has attracted criticism and push-back from the academic community and a number of universities. Critics have argued that the two types of services do not necessarily need to be part of the same contract, and that linking them makes it less clear what exactly is being paid for.
In response to these criticisms a short break in the negotiations was announced late December. Parties agreed to a transition period of four months, with an accompanying Memorandum of Understanding. The Dutch knowledge institutes would use this time to task a working group with formulating the principles that would provide a framework for sharing institutional (meta)data. The move was intended to assuage the criticism and fears in the academic community, and enable the negotiators to continue the talks with a clear goal and framework.
Establishing this working group turned out to take longer than anticipated. Its first meeting was held only days before the outbreak of the corona crisis in the Netherlands and thus the question rose whether the first of May first was a realistic deadline. Both parties made it known that they would stick to their agreement. Additionally, the VSNU originally stated that the mandate and composition of the working group would not be made public, but when this did happen after all, it became clear the mandate was decoupled from the contract negotiations. The aims of the working group and its tasks were “broadened”.
Having missed the deadline by just a few weeks, a deal was presented on the ides of May. Remarkably this happened before the working group could even present its guidelines. It is unclear why both parties were intent on rounding up the negotiations, but it is clear that the contract has demonstrably suffered in clarity. The two parties are joined in marriage, but the prenuptial agreements are long and vague. A reluctant ‘I do’ at best.
Regarding the publishing part of the contract, Elsevier has clearly made a huge concession compared to the company’s previous stance on open access. Ever since Plan S was announced in september 2018, Elsevier has actively sought for publishing solutions that would prevent journals from going full open access. This deal offers open access for articles published by Dutch institutes and universities in at least 95% of Elsevier’s 2500 plus journal titles.
It is, however, a very specific form of open access: gold open access. This means that research institutions simultaneously pay the publisher to read and to publish. The two go hand in hand. At the same time the journals are free to keep running a closed access model, meaning that publications by corresponding authors from countries and institutions without such a deal still end up behind a paywall.
Also, Elsevier has not yet been moved to give up its remaining flagships: the Cell and the Lancet. These journal families are explicitly excluded from the agreement for the time being. On close examination the contract does state the intention to include these, now described as ‘non-full open access titles’, in the deal: “both parties will work diligently together during the time of this Agreement [until 2024, red.] to include such journals in order to support the 100% Open Access goal”.
In their joint press releases, both parties emphasize the contract is in line with Plan S. This is also the view of the Dutch Plan S champion, Johan Rooryck: “With this, the Netherlands really is frontrunner regarding the percentage of open access articles. The deal is sufficiently transparent given the current principles of Plan S for transformative agreements.”
One day before the contract with Elsevier was made public, the coalition behind Plan S announced that from 2022 onwards, publishers must provide transparency on the prices they charge for various parts of the publication process. “Regarding the requirements for transparency of Article Processing Costs/Charges that will be in place from 2022 onwards, the deal is not transparent. We hope that by that time, Elsevier will have converted”, Rooryck adds.
A new convert to the open access gospel?
Thus, a concession from Elsevier, but at what price? At the start of the negotiations, the Dutch negotiators, through Koen Becking (president of Tilburg University) made it known they were setting the bar higher compared to the deal from 2015: “If we don’t get more open access for less money, there will be no deal”.
Under the new contract, the institutions collectively pay €16,4 million per year (from 2021 onwards), compared to €12,2 million (reference point 2018) per year under the previous contract. With roughly 6000 articles published by researchers at Dutch institutes in 2019 this boils down to an estimated €2700 per article.
Other ‘transformative agreements’ that Dutch institutes have struck with other publishers ended around €2000 per article for large parties like Springer Nature and Wiley. Here, too, the agreement is that institutions pay the same amount as previously for subscription costs, on the condition that all publications of (corresponding) authors in the Netherlands will be open access.
However, due to the ‘unique’ nature of the contract, no true comparison can be made with other agreements. Not only because various Elsevier tools and platforms are also included in the contract, but especially because of the arrangements around what has become known as ‘Professional Services’. The market value of the ‘open science’ component is, after all, unknown.
A research platform company
So, what are these Professional Services exactly? On this issue the wording of the contract is vague, and the agreements again are phrased in a suggestive rather than decisive manner. For one, the phrasing in the agreement speaks of “Examples of the Professional Services” followed by a number of possible “pilots” that “could be” carried out.
The universities and research institutes have to opt-in, rather than opt-out of such projects: “Parties have agreed to enter into certain pilots for Professional Services to Institutions, at the sole discretion of each Institution to participate,” followed by many more articles concisely detailing the regulation of the pilots. A dedicated new governing body, installed by the institutions, has to oversee the pilots (without Elsevier’s involvement).
The examples mentioned include a pilot on the aggregation of research metadata aimed at producing insights into which ‘research outputs’ are produced by research grants from ERC, NWO, ZonMW and other funding organizations. Another suggestion is a VSNU-pilot concerning the collaborative effort of Dutch institutes to move towards a new system of recognition and rewards. None of these pilots have been agreed upon yet, which raises the question why they are mentioned in a formal contract in the first place.
The pilots raise additional concerns beyond the previously voiced reservations. For one, the proposed pilots seem to be at odds with the intention to move away from bibliometrics for evaluating research(ers). Additionally, generating and validating this data is primarily a task for the institutions. Consequently, any responsibility for the correctness lies solely with the institutes.
When plans for combining the “Professional Services” and open access first emerged last November, it was met with considerable criticism. Many were cautious about entering into an exclusive contract with a company that already owns a large part of the (digital) research infrastructure, pointing at the risk of a vendor lock-in. Founding father of Plan S, Robert-Jan Smits, also had previously warned about such deals concerning research (meta)data during a public event in September 2018. “Let’s make sure that what we let happen to our research articles, we do not let happen again with our research data”, he then stated.
Although the current contract does not obligate institutes to deal with Elsevier in such projects, it does raise the company to the status of ‘preferred partner’: the default option. Additionally, the platform company managed to define some areas where they have been granted the initiative.
Undesired platform effects
In a recent blog post, scientific director and professor of Science, Technology, and Innovation Studies at CWTS Sarah de Rijcke questioned the portrayal of the agreement with Elsevier as an “open science deal”. “It is evident that Elsevier has quite a lot to gain from this deal. What is in store for them is a unique research intelligence infrastructure, because it is not only a national-level arrangement, but on top of that, the information that goes in is also validated by the institutes and research funders themselves.”
De Rijcke, who is a member of the aforementioned working group on the principles, hails the inclusion of a set of principles in the contract, but is also worried. “I am not persuaded by the contract, and still find it disconcerting that this deal may effectively transfer crucial means to influence Dutch science policy to a monopolistic private enterprise.” The worries about ‘undesired platform effects’ warned about have not been taken away, she states.
The primary question raised by De Rijcke is the obvious one: why the rush? “It would have been far more preferable if the research institutions would have first formulated general principles for collaboration with private parties, and only then had started to engage in projects and look for third-party interest.” She points out that the deadline for the open consultation was already on June 8th [now postponed until June 17th] and the working group was operational only since the beginning of March.
Why exactly the deal that had been in the making for the better part of two years needed to be finalized in such a hurry, remains unclear. What is clear however, is that both parties enter into an exceptionally long commitment when compared to other publishing deals: five years. This could be an effect resulting from the very same fact that negotiations with major platform and publishing companies take up a considerable amount of time and energy in the first place. Some of those involved in the negotiations complain that negotiating with this market leader distracts from the countless other agreements with often smaller publishers, that are important as well.
Without a doubt, the share of open access articles by Dutch (corresponding) authors will rise in the coming years. And perhaps €84 million is a fair price to pay. But any judgement on this is impeded by the basic fact that the two services are tied together. The resulting contract echoes the reservations of academics and research institutes, but does little to provide them with actual clarity or a sound legal basis to fall back on. For Elsevier, the deal ushers in a new era for a platform company as a result of a home game.