On a superficial reading open access is intended to do no more than what it says on the can: provide an internet-based scholarly communication system in which research is made available sans paywall – in other words, a system offering improved accessibility over the traditional subscription system.
On a deeper reading, however, we learn that the OA movement was a response to the unsustainably high costs of the subscription system and that it was based on a conviction that open access would be a more cost-effective way of sharing research – in other words, a system offering improved affordability.In addition, it was argued, open access would be a more transparent way of doing things than the subscription-based system.
Essentially, the argument went like this: If researchers paid an article-processing charge (APC) every time they wanted to publish a paper (rather than librarians paying the costs of publishing by purchasing subscriptions to large bundles of journals courtesy of the so-called Big Deal), then not only could research papers be made freely available to all, but authors would be able to make price-based decisions when choosing where to publish.
This price transparency, argued OA advocates, would introduce market forces into scholarly publishing that are absent in the subscription system. It would also allow new open access publishers to enter the market with lower-priced products, which would help drive down prices.
In short, OA advocates promised that open access would not only provide greater accessibility but a more cost-effective scholarly communication system, thereby solving the affordability problem that has long dogged scholarly publishing. And to achieve this, they said, transparency is key.
Transparency is key
Transparency is key because in order to make price-based decisions buyers need to be able to compare prices. While APCs allow this, Big Deals do not, because with the subscription system researchers have no idea whatsoever what costs are involved, and librarians (who buy on their behalf) do not have a published price list to work from and do not know what other librarians are paying for their Big Deals, since publishers insist on non-disclosure agreements (NDAs). In such an environment pricing is opaque and everyone bar the publishers has to fly blind.
As OA developed, however, it became apparent that most researchers do not have access to the necessary funds to pay APCs, so libraries have again had to start acting as intermediaries. In doing so, however, they have found the task of managing hundreds of APC payments on behalf of researchers difficult, time-consuming and expensive. For this reason, they have struggled to cope.
At the same time, European governments, funders and university leaders have become increasingly impatient at the time it is taking to achieve widescale open access.
These two things have led to the emergence of the OA Big Deal. Here agreements are signed with legacy publishers that combine bulk journal subscription fees (as with traditional Big Deals) plus bulk OA publishing fees so that authors can publish without personally having to pay APCs. Those librarians and university leaders signing these deals have therefore come to view the OA Big Deal as the best way of transitioning to a fully OA publishing environment. And while today the OA Big Deal is more of a European issue, it looks set to become the model of choice elsewhere in the global North (also here).
As we shall see, however, there are good reasons to doubt that this strategy can provide a satisfactory outcome.
Strikingly, it is the most vocal critics of legacy publishers and their prices (librarians and national university associations) who are promoting these deals, either because they fail to understand (or accept) the implications of what they are doing, or because they have been mesmerised by the EU’s rash and unthought-through commitment to make all European research freely available by 2020.
It is also concerning that the negotiators of these OA Big Deals appear to have little appetite for transparency. What these agreements consist of, what they cost, and what kind of value for money they offer (or don’t offer), therefore, is generally unknown to anyone outside the small group of people taking part in in the negotiations.
Hard to describe
Commenting on these deals last year, the CEO of OA publisher Hindawi Paul Peters said, “Describing exactly what these things are is quite difficult, both because … none of these deals looks like another one of these deals – they all have slightly different models. Universities are getting coupons they can use on things, or discounts, or everything is all bundled together.”
He added, “And apart from their being difficult to describe because of the variety, they are hard to describe because they are not public. Most of these things are covered by NDAs and the only things we hear tend to leak out via the grapevine, so it is hard to describe what is happening.”
Contributing a further level of confusion, these agreements have given rise to a rash of ambiguous euphemisms – including “transition agreement”, “offsetting arrangement”, “Read and Publish” agreement and, more recently, “Publish and Read” agreement.
What then can we say about these OA Big Deals? How do we go about differentiating them? And what are the likely implications of their use?
One obvious problem with the OA Big Deal is that it allows large legacy publishers to lock their high prices into the new OA environment, while marginalising and excluding the new-entrants that were supposed to disrupt the market. Unless something changes, therefore, the affordability problem will only be perpetuated.
And as Peters points out, it is not even clear what these things are. When I took to Twitter to ask Bernhard Mittermaier (head of the Central Library at the Jülich Research Centre and a member of a team negotiating these kind of deals in Germany) for guidance on the term Read and Publish he replied: “To the best of my knowledge, the term ‘Read & Publish’ was introduced by Royal Society of Chemistry, though others had the concept in place earlier.”
Under this model, he explained, “An institution pays for reading (subscription fees) and can publish all its corresponding author articles as hybrid open access. This comes with no additional costs for the authors and for the institution; everything is paid upfront as a lump sum (adjustments according to the actual number of articles might occur in the following year or the next contract). Another name for that is ‘offsetting contract’.”
By contrast, he added, the model that he and his colleagues in Germany are working towards with Project DEAL is what they have dubbed ‘Publish and Read’. With this model, he explained, “We strive to pay according to the number of publications and read at no additional costs. Rationale: Publishers are paid for publishing. One can read the articles elsewhere as well. And there is no reason to pay them for both.”
This website describes the Publish and Read model as one in which all the articles by authors in eligible institutions are made open access on publication under a CC BY license (the Publish component) and eligible institutions have perpetual access to the complete e‑journal portfolio of the publisher in question (the Read component).
It seems fair to say that the Publish and Read approach is more assertive than most OA Big Deal models. And that presumably explains why it has led to a long-running standoff between German universities and Elsevier.
In the Netherlands by contrast, a less aggressive approach has been adopted. Essentially, explained Dutch OA publishing consultant Jeroen Sondervan on Twitter, the Dutch have been pursuing a Read and Publish model. This is being done under the aegis of the association of Dutch universities (VSNU).
But even if the more aggressive German approach were widely adopted, it would still see legacy publishers embedding themselves and their high prices into the new OA world, while elbowing aside OA publishers like Hindawi and PLOS. For this reason alone, some feel that the OA Big Deal is fundamentally flawed.
And as indicated, the current landscape is confused and confusing. Even those seeking to negotiate these deals appear sometimes to be hazy as to what exactly an OA Big Deal consists (or should consist) of.
On 15th March, for instance, the Swissuniversities (Rectors’ Conference of the Swiss Universities) published a document announcing that, as part of a national open access policy intended to make all publicly-funded research freely accessible by 2024, Switzerland plans to negotiate OA Big Deal contracts with publishers. These, it said, would consist of agreements in which [Google translation] “universities finance the publication costs and pay a fixed price for reading and downloading published articles and to pay for works without incurring additional costs for Swiss universities. This will replace the classic subscriptions whose prices vary according to the journal.”
This makes it sound like Switzerland currently buys individual journal subscriptions, which seems unlikely. But what is puzzling here is that when one enquires as to whether the Swiss are looking to emulate the assertive German approach, the more modest VSNU approach, or some other approach, one gets the impression that the negotiators either do not know what strategy they plan to pursue, or wish to keep it secret.
So, another characteristic of the OA Big Deal is that it tends to perpetuate the secrecy inherent in the traditional Big Deal. The problem is that if the details of these agreements are kept secret how will any negotiating team be able to make price-based decisions, or facilitate the emergence of market forces? They won’t know what publishers agreed with other groups, and so will find they are still be flying blind. Likewise, how will taxpayers know what has been purchased in their name, and at what price?
It would seem, therefore, that OA Big Deals will simply replicate the lack of transparency endemic to the subscription Big Deal, and so fail to address the affordability problem.
Well-intentioned and worthy
Let’s be clear, those negotiating OA Big Deals are well-intentioned people and their goal is a worthy one: They want to engineer a transition to a world in which all research is freely available to everyone, which is surely “a good thing”. It is also important to note that they are keen to avoid the disastrous mistake the UK made in trying to take a leadership role in open access by agreeing to pay publishers more money to provide OA.
The UK has signally failed to achieve any meaningful transition to OA while enriching publishers further. It is also now clear that the UK approach is financially unsustainable. This is all too evident if one reads the recent Monitoring the Transition to Open Access report. As the Times Higher noted, “In 2016, a sample of 10 UK universities paid £16.1 million for subscriptions to seven of the biggest publishers, up a fifth since 2013, according to a new Universities UK report tracking the growth of open access publishing, released on 5 December. Yet these universities also spent £3.4 million on article processing charges (APCs), the fees required to publish an article open access, up from about £750,000 three years previously”.
European negotiators have therefore vowed that they will provide no new money for OA Big Deals, although whether they will succeed in this remains open to question. In any case, even if they do manage to avoid the British trap, it is highly unlikely that the costs of scholarly communication will come down as a result of these agreements. Legacy publishers are never going to voluntarily agree to deals that will lead to a fall in their profits, and by their very nature, these deals cannot unleash the kind of market forces that would be needed to compel publishers to lower their prices.
Moreover, even if the OA movement gives up one of its primary objectives (a lower-cost scholarly communication system) how will anyone not involved in the negotiations know whether any OA Big Deal does or does not involve new money and is good value? As Peters pointed out, aside from those directly involved in negotiations, no one knows what is being agreed.
There is, therefore, growing concern in the research community that those agreeing national contracts with publishers are failing to get a good deal, and enriching publishers at the expense of the research community (and thus the taxpayer). There is, for instance, now some disenchantment with the UK organisation charged with negotiating with publishers (Jisc). We shall discuss Finland and the Netherlands in a minute.
To find out what publishers are being paid for their Big Deals researchers have had to resort to making Freedom of Information (FoI) requests – a tactic pioneered in 2014 by Cambridge mathematician Timothy Gowers, and by similarly curious researchers in the US. Their findings have only led to growing anger and scepticism.
It was in recognition of this anger that last September the Association of European Research Libraries (LIBER) published the Five Principles for Negotiations with Publishers. One of these principles (headed Transparency for Licensing Deals: No Non-Disclosure) reads, “Licensing agreements should therefore be openly available. Society will not accept confidential agreements paid for with public money in the form of non-disclosure agreements.”
One question we need to ask at this point is whether OA Big Deals are any more transparent than the traditional Big Deal. This question took centre stage when, on 17th January, Elsevier announced that it had signed a deal with the FinELib consortium based at the National Library of Finland. Like all recent deals in Europe today, the Finnish agreement had an OA component. Yet even though no NDA had been signed, there was a distinct shortage of detail about what had been agreed between Elsevier and FinELib. This was all the more striking given that shortly after the announcement FinELib publicly endorsed the LIBER transparency principles.
When I asked FinELib why so few details had been made public even though no NDA had been signed I was told, “The confidentiality of an agreement is not based on whether or not there is an NDA. At least in Finland the principle of loyalty between contracting parties needs also to be taken into account.”
Needless to say, Finnish researchers were less than impressed with this explanation, and put considerable pressure on FinELib to publish more information about the agreement. This had some success, and in February FinELib agreed to publish more details, including the total price of the deal at the consortium level, the total price per consortium member, plus the text of the Elsevier agreement.
FinELib added, however, that it was not able to publish the annual prices alongside the total price due to a requirement in the Act on the Openness of Government Activities. This, it explained, forbids the publication of “documents containing information on a private business or professional secret, as well as documents containing other comparable private business information, if access would cause economic loss to the private business.”
NDAs, it seems, may not be the whole story here.
At least four problems
The Finnish agreement also graphically demonstrates how OA Big Deals lock in large legacy publishers, to their advantage and to the disadvantage of pure OA publishers and smaller companies.
Specifically, the open access “pilot” built into the FinELib agreement gives researchers a 50% discount if they opt to publish open access in a group of Elsevier journals. Finnish researchers were quick to point out these journals are predominantly hybrid OA, not pure OA. Consequently, rather than hasten a transition to OA, they complained, the deal will embed hybrid OA in the Finnish publishing landscape for the foreseeable future. This will prolong the subscription model (and so delay the transition to OA) and elbow out pure gold OA publishers.
Frustration grew when it was discovered that Helsinki University already offers a 50% discount for publishing OA in Elsevier journals, giving the publisher an even greater advantage over pure OA journals. “Elsevier is now free (and cheaper than full OA) for a researcher,” tweeted one researcher.
It seems to me, therefore, that there are at least four problems with OA Big Deals. First, even where no NDA has been signed, transparency over what has been agreed, and the costs, remains inadequate. Second, these deals often seem as likely to delay as accelerate the transition to open access, and certainly not in the way governments and funders in Europe are now keen to see. Third, they unfairly advantage large legacy publishers. Fourth, these deals are as a result unlikely to solve the affordability problem.
On a more positive note, Finnish experience suggests that if researchers are persistent they may be able to persuade negotiators to provide more transparency that has been possible historically. However, this is more to do with heightened suspicion about publishers than open access. Either way, however, many will feel it is an insufficient reason to warrant pursuing OA Big Deals.
The Dutch approach
I argue that those negotiating OA Big Deals are shy about sharing the details of the agreements they are signing. In this regard it is worth considering the activities of one of the organisations that pioneered the OA Big Deal – VSNU. The association has been active in this space since 2015 and is keen to promote its self-styled Dutch Approach.
When it started VSNU vowed that it was going to take a hard line with legacy publishers, especially Elsevier. “We are willing to pay publishers for the work they do, but Elsevier’s profit margin is approaching 40 per cent, and universities have to do the [editing] work and pay for it”, asserted Professor Meijer, then chief negotiator for VSNU and Chairman of Radboud University Nijmegen. “We aren’t going to accept it any longer.”
As part of the sabre rattling, VSNU threatened publisher boycotts, and began to reject publishers’ offers. In 2017, for instance, it turned down the deal offered by Oxford University Press (OUP).
Nevertheless, it has gone on to agree a number of OA Big Deals, including in late 2015 one with Elsevier. This came after a year-long deadlock (and is currently up for renewal). Nevertheless, when researchers saw the details of the agreement they expressed some disappointment. And there has been rumbling discontent with the Dutch approach since, not least over its lack of transparency. Again, it took FoI requests to obtain more information and in 2016 this saw details of what Dutch universities pay to scholarly publishers become publicly available (Details here; see also here).
Currently, VSNU is engaged in a new series of negotiations. In doing so it again insists that it is taking a hard-line with publishers and indeed on 12th March it announced that it had failed to reach agreement with the Royal Society of Chemistry (RSC). Speaking to ScienceGuide Koen Becking, President of Tilburg University and board member of VSNU said, [Google translate] “We have shown that a ‘no deal’ is also possible. We stand behind our commitment: we only pay for output.”
On the same day (12th March) VSNU also reported that it had renewed its agreement with Springer Nature.
Both VSNU’s announcements consisted of little more than a brief statement with no meaningful detail. While subsequent news of a deal with OUP also included a couple of self-serving quotes, there was little else, and researchers were left scratching their heads. As one perplexed scientist commented on Google+, “The article does not mention the cost of the deal – how much Netherlands pays for the deal per year?”
If the research community doesn’t even know how much money is involved how can it hope to judge the value of any OA Big Deal. And even if we are convinced by VSNU’s robust statements about beating publishers into submission, we are not ourselves able to judge whether taxpayers’ money has really been spent effectively, let alone what exactly that money has bought.
I was alerted to the Springer Nature deal by a librarian who is clearly unhappy about the way VSNU is negotiating these deals, and the abiding lack of transparency that surrounds them. Describing herself on Twitter as an open access advocate, she emailed me to ask, “Would it not be time for some questions towards VSNU on how, why and for what amount?”
Since that seemed like a good idea I took to Twitter, and invited VSNU and Springer Nature, as well as RSC, to answer some questions.
I received no response from Springer Nature.
The RSC did reply and sent me a statement about its failed negotiation with SURF (who act on behalf of VSNU) along with an invitation to schedule a conversation to discuss the RSC’s OA work more generally (I responded by offering to visit their offices).
Unfortunately, the RSC statement tells us little, although it is interesting to learn that (amongst other things) it offered a Read and Publish option to the Dutch universities. If Sondervan is right to characterise the model that VSNU is pursuing as being Read and Publish, why did it turn down RSC’s offer? We do not know. Presumably, the issue was pricing, but without more information we cannot know who is being reasonable here and who unreasonable.
In the end, only VSNU can tell us why it turned down the RSC offer, which is partly why I contacted the organisation. That said, I was mainly interested in obtaining details of the Springer Nature deal. At first, my attempt to do so met with an encouraging response. VSNU replied to my Tweet by inviting me to email my questions to them. So, I emailed over 22 questions. Yes, it was a lot, but how else are we going to know what is going on with these deals if we don’t try to drill down beneath the press releases, beneath the corporate quotes, and beneath the declamatory statements.
VSNU’s Spokesperson and Advisor Public Affairs Bart Pierik responded by saying (fairly) that he would need a little time to answer my questions. I was disappointed, however, when the next day I received an email from him saying, “Considering the fact that we are finalizing some more deals with publishers at this moment (we just published good news about Oxford University Press) my proposal is that we would be glad to make one Q&A in April about all of these deals.”
I was disappointed with this response not just because it would mean a delay before details of the Springer Nature were made public, but because it seemed self-evident to me that if my 22 questions were re-focussed and rolled into a discussion about a bunch of other agreements the details of the Springer Nature deal would get lost. As I say, it is the details of these agreements that seem to me to be vital if we are ever to establish exactly what is going on and whether these OA Big Deals offer any kind of value for money.
VSNU’s response to my questions, plus the fact that it took a series of FoI requests to find out what Dutch universities are paying publishers last time around, left me a little sceptical as to how committed to transparency the organisation is. The desirability of their being transparent seems all the more pressing given that there is some discontent with the way VSNU has been negotiating OA Big Deals. At a Couperin event held in January, for instance, Becking (along with Director of Scientific Information Provision at the Max Planck Digital Library Ralph Schimmer) faced a series of awkward questions and complaints from the audience.
Concern was expressed about pricing, about the dangers of allowing large legacy publishers to lock themselves into the new OA environment, about the consequences of this for pure OA publishers and smaller companies (and for future innovation) and about the future of scholarly publishing if OA Big Deals become the norm.
(In passing we would note that Couperin, the French consortium of academic and research institutions, recently announced that it has failed to reach agreement with Springer Nature for an OA Big Deal, and so researchers in France will lose access to the publisher’s journals on April 1st). [Update 31/03: I have been advised that this was not an OA Big Deal negotiation but (presumably) a regular Big Deal. However, as The Scientist points out, “the consortium was pushing for a reduction in subscription costs to account for the increasing proportion of open-access articles – for which authors pay an article processing fee to publish – in Springer's journals.” It’s complicated.]
The questions and points aired at the Couperin meeting appeared sufficiently compelling that Becking closed the session by saying [at 43m in this video], “I have two take-home messages for myself. That is to communicate better on what we are doing and why etc. etc. And secondly to invest more in new initiatives.”
But if the Couperin meeting has encouraged VSNU as an organisation to commit to greater transparency it is not immediately evident. Why did it not provide greater details when it made its recent announcements? And why did it seek to refocus and (I assume drown out) at least some of my 22 questions? I am also not aware that anything beyond the sparse announcements made in early March (and the quote from Becking in the ScienceGuide piece cited above) has been released to the public in order to clarify what VSNU has agreed with Springer Nature and OUP (and failed to agree with RSC). The organisation also seems inclined to ignore requests for information from others.
In light of my growing scepticism, on receipt of VSNU’s email I took to Twitter again to share my disappointment. “Information on the new Springer Nature/ VSNU OA Big Deal is sparse, to say the least,” I tweeted. “VSNU agreed to answer my questions, but after seeing them decided it does not want to answer any questions on its negotiations until April.”
VSNU’s Bart Pierik responded: “This is bogus (and you know it). With many negotiations going on, we offered to make one complete Q&A, as to offer all relevant Information in one place – your website.”
My response to Pierik is here.
But what about Springer? Why did it not respond to my tweets? Again, I cannot say. However, when I later suggested that the European Commissioner for Competition might be advised to look into the growing number of OA Big Deals, I did finally get a response from the publisher – in the form of a comment from Robert Boissy, Springer Nature Director of Institutional Marketing & Account Development, Americas. He said, “So at the end of the day the OA availability is less important than the path taken to the OA availability? I respectfully disagree.”
In thinking about the implications of Boissy’s comment it occurred to me he is saying that all that matters in the transition to open access is the end-point, not the process. But surely how one arrives at an end-point often determines the nature of that end-point. Boissy appears to be saying that accessibility is all that matters, not affordability, and not transparency. So long as the world gets OA it doesn’t matter what the taxpayer has to pay for it, or whether what is paid (and what that payment buys) is made public.
Boissy’s comment would also seem to imply that it does not matter if legacy publishers are allowed to dominate the new OA landscape.
Yet history shows us the consequences of allowing large publishers to dominate scholarly communication. Despite claims that the internet levels the playing field, the arrival of the Web has seen ever greater consolidation, coupled with a routine lack of transparency, and a consequent cycle of ever-rising prices.
As a paper published by PLOS ONE in 2015 called The Oligopoly of Academic Publishers in the Digital Era explained, “Combined, the top five most prolific publishers [now] account for more than 50% of all papers published in 2013. Disciplines of the social sciences have the highest level of concentration (70% of papers from the top five publishers), while the humanities have remained relatively independent (20% from top five publishers).”
As I see it, the new OA Big Deals can only accelerate the process of consolidation and push out (or swallow up) the (inevitably smaller) pure OA publishers. The latter, let’s recall, were supposed to disrupt the scholarly publishing market by making it more competitive, and driving down prices.
What happens when we get to the point where scholarly publishing is dominated by not five but just one company – let’s call it SNEW. Would it really be good for the research community? Would not that company be able to hold researchers (and the taxpayers that fund them) hostage?
Meanwhile, legacy publishers are moving into the workflow, analytical services, institutional repository, electronic notebook and data services areas of scholarly communication. Every part of the scholarly infrastructure is now threatened with appropriation and domination by large commercial publishers, with Elsevier leading the pack. Surely every OA Big Deal signed assists in this process of consolidation and appropriation, as does every agreement where the costs and terms are not released to the world. This is public money being spent on these companies, people. Is it really being spent wisely?
In short, the world really needs to know what is going on. I told VSNU I would get back to them in April. But on reflection, I wonder whether there is sufficient value in undertaking a Q&A if one allows the interviewee too much control over the questions and how and when they are answered. VSNU has my 22 questions. It is clearly free to answer them or not as it desires. Either way, it is now two weeks since I emailed them over, and two weeks since I proposed to RSC that I visit their office (for which I have yet to receive a response).
So, I will end by again inviting VSNU to answer my questions. By doing so they can help shine a light on this somewhat crepuscular corner of scholarly communication and demonstrate that affordability and transparency are just as important as accessibility.