Background

In the midst of the ongoing transition to open science, issues of open access (OA), open data, citizen science and other related matters engage universities and many others more than ever, globally as well as locally. In 2021, UNESCO adopted a recommendation for open science, which it pointed out as crucial to reach the sustainable development goals. Within the European Union, the issues have been high on the agenda ever since the Council conclusions were adopted in 2016 on the transition to open science (UNESCO 2021; Council of the European Union 2016). In 2023, new Council conclusions were decided upon under the Swedish chairmanship (Council of the European Union 2023). Finally, in Sweden, governments of different political colours, have highlighted open science as a means to reach both higher quality of research, higher research impact and increased collaboration and innovation.

In the following, I will mainly focus on one aspect of open science, namely on open access to scientific publications. Let me first attempt to frame this discussion. It is important to remember that the open access movement was researcher-driven from the start; it aimed at reclaiming the rights to the researchers’ own results, to spread them openly to other researchers and to society at large and not least also, through openly sharing both publications and data, to enhance the reproducibility and general quality of research. Since then, politicians, in Europe as well as more recently in the United States, have largely adopted the open access vision. Fundamentally, there is nothing wrong with that – in general, academia tends to be pleased when politicians listen to researchers – but there is a risk that researchers’ memory may sometimes be too short and that this might lead to a backlash, where researchers instead argue that open access policies and goals are envisioned and driven by politicians and are fundamentally hostile to research needs. Swedish politicians, through several research bills, have set ambitious goals for the transition to open access. This has led to universities, which in Sweden are mostly state authorities, striving to steer in this direction, which in turn has led to some researchers claiming that this is a threat to research quality, among other things through the growth of predatory open access journals.

In Sweden, the goal set by the Parliament was to have reached 100% open access in publications as early as 2021 – while, in 2023, Sweden has reached approximately 82% OA. In 2026, Sweden is supposed to have reached 100% open access when it comes to research data. This of course exerts high pressure on universities to accelerate their work to make the transition happen. The National Library has had its specific government directives since 2021, to promote and coordinate the work of introducing open access to scholarly publications, which includes submitting a comprehensive survey, analysis and assessment of the national work with open access to scholarly publications, and, since 2022, also including the task to set up a national policy framework. In addition, through the 2020 research bill, which since then has been adopted by the Swedish parliament, higher-education institutions are called upon to advance their work, primarily by helping the Swedish Research Council and the National Library of Sweden in their respective directives to coordinate open science by contributing material. It is clear from this government directive, but also from previous attempts to point out the role of universities and university colleges, that Swedish governments, regardless of their political colour, tend to underestimate the role of higher-education institutions as the central actors in the transition to open access, and that this has contributed to the problem of allowing costs to increase. Universities and their researchers are both drivers in the transition and bearers of the increasing costs. Hence the need to collaborate nationally and internationally.

The Swedish Approach: The Bibsam Consortium

The Bibsam consortium was founded in 1996 in order to coordinate the deals of Swedish universities with publishers and to reduce costs by sharing parts within bigger deals. The National Library of Sweden administrates it. To date, 93 organizations are participating in Bibsam. All universities and university colleges are members, but also a number of government agencies, including research funders and research institutes. Bibsam works as an opt-in consortium, with separate agreements where participating organizations have the choice to join or not, and only in the first case pay its full costs. To date, the consortium has 44 agreements in all, 28 of which are transformative agreements and four pure OA agreements. The turnover is approximately €48.7 million. Bibsam is governed by a steering committee, which is chaired by a university president from one of the research-intensive universities. The Swedish Association of Higher Education Institutions appoints all the members, except for one member representing the smaller libraries, which do not belong to any higher education institution.

The steering committee of Bibsam has decided on the preconditions for its work, which might also be called a policy. When entering into negotiations with publishers, the Bibsam consortium thus has had three major preconditions during the transformative period: (1) immediate OA to all articles published by researchers affiliated with participating organizations; (2) continued reading access for those organizations to previously subscribed content; (3) a sustainable price model that enables the transition to an OA model and thus allows for redirected revenue streams. Contract renewal must include an OA provision, and if no acceptable OA provision is offered, the contract renewal will be limited to one year or even cancelled, as was the case with Elsevier in 2018.

The mission of Bibsam to reach the target of immediate open access has been quite efficient. The main way has been for the payment streams to be re-channelled from the financing of read licence agreements to the financing of publishing costs within transformative agreements, that is: read-and-publish agreements. During the past eight years, from 2014 with the first new open access agreement, until today, with – as mentioned above – 28 transformative agreements plus four agreements with purely OA, there has been a considerable development. However, unfortunately, not only has the number of agreements increased but so have the costs, as read-and-publish agreements may bring costs both for reading and for publishing.

Chasing the Double Dipping

When first taking office as chair of the Bibsam consortium in 2016, I soon became aware of the fact that there was little or no control over the total publication costs at the universities. Researchers paid for publishing in a number of journals and universities paid the subscription fees for the same journals. Moreover, researchers of course had already invested their time and resources not only to do the research but also to act as peer reviewers, serving on editorial boards or as editors and so on. Much has been said about this strange commercial model, where both researchers and universities become losers in a system where the profit margins, not least for the big publishing houses, are only increasing. Since then, much has happened – but still not enough, because the fundamental problem remains largely the same.

At the time, however, the first priority was to gain better control of our total costs. The subscription fees were already well known. But as the consortium decided to also strive for control of the publication costs (the article processing charges, or APCs), we chose to gather information in several ways. Through an initiative by the Swedish Association of Higher Education Institutions, we started to collect information on APCs paid. This took place in parallel with the development that a number of universities decided to pay all APCs from central funds within the university, having first signed the Berlin Declaration for Open Access and later also the Open Access 2020 Declaration. However, as mentioned, the money flows from public research to publishers have traditionally been uncontrolled, with no transparent overall picture of the total costs, which is why in 2019 the National Library of Sweden was tasked with annually collecting and presenting the total cost of scholarly publishing. A first overview showed that scientific publishing costs (not including administration) were close to €49 million annually. This means that approximately 1% of public research funding was set aside to pay for scientific publication at Swedish universities in 2019 (Kungliga biblioteket 2019). Today, it is close to 1.5%.

In 2018 and 2019, there was a substantial reduction in subscription expenses, due to the Bibsam Consortium’s termination of the Elsevier agreement at the end of the first half of 2018 (approximately €11.5 million in total). At the same time, costs for transformative agreements have become a growing item of expenditure; they increased by 50% between 2017 and 2018 and just over 25% between 2018 and 2019, and this has only continued.

Due to these measures, however, the Swedish research community has good cost control today. Therefore, knowing where the money goes and how, it was time for the consortium to start confronting the publishers in the negotiation processes with the double dipping that was taking place. In fact, the separate funding streams, with both subscription agreements and agreements concerning OA publications in hybrid journals, benefit the system, with publications behind paywalls. To break this vicious circle, the funding streams must be redirected towards full and immediate OA. However, there is enough money in the system as such.

At one of the first high-level meetings with one of the largest publishing houses, in which I took part, the company had their OA person and the person responsible for their subscription deals seated at the same table at the same meeting for the first time. They strongly maintained that they only did so because Bibsam had explicitly required it and that these two branches of the company had absolutely nothing to do with each other. Moreover, they told us that it was not on the agenda even to discuss the two within the same framework, and even less so to reach an agreement where both cost streams would be included. Since then, we have indeed come a long way. At that time, it was unthinkable to negotiate both publication costs and reading costs within the same agreement. Today, it has become standard. Plan S and later Coalition S have been quite decisive in accelerating this process, as in making a number of pioneering agreements in different countries, which are now being realized. Transformative agreements were initially seen as the privileged way to move towards a full-scale transition or transformation, flipping from subscription deals to pay-for-publish only.

In Sweden however, we have accelerated the process for open access to the extent that we are one of the world-leading countries today (ESAC n.d.). This, of course, has certain advantages – having reached approximately 80% OA in 2022, Sweden is at least quite close to reaching the national goal. However, this position also contains numerous challenges, not least when it comes to cost control. We have always argued that there is enough money in the system. The main issue concerns what it is that we pay for. The transformative agreements have already reduced the total costs to a certain extent, as the last Elsevier agreement shows. A continuation with APCs and subscriptions in two separate funding streams would have led to considerably higher costs. In that case, the reading agreement plus the list-price value for publishing would have amounted to €22.5 million, whereas the read-and-publish agreement in 2022 totalled €14.3 million. Still, costs generally continue to increase. Why is that? For a number of years, Bibsam has argued to Swedish politicians that the period of transition would be very expensive for universities, as they would have to pay for read-and-publish agreements, thus still risking increasing the costs in a short-term perspective. The vision in the long run, however, was to bring down the costs once the transition is a fact and the whole system has been flipped. What we have seen so far, though, are still mostly increasing costs even for transformative agreements once the two pillars, publishing and reading, have been established within the same agreement. It is thus necessary to move further ahead.

It is clear that Bibsam managed to substantially bring down the costs when first changing from subscription deals and separate APCs to transformative agreements including both publish-and-read, but now there seems to be nothing more to gain from continuing those agreements. Rather, they risk turning the transformation, which presupposes a change, into ‘a permanent transformation’: a fixed state where the publishers can continue to increase the costs for both reading and publishing. If publishers might not be very eager to move forward, this is not hard to understand; rather, the state of transformation becoming permanent would benefit their interests. Such a development would not only threaten the very concept of transformation but also undermine the motivation to flip the system. At this stage, several large publishing houses honour both the principles and the goal of reaching OA, still claiming that the needs of the researchers steer the process. In the meantime, their own profit margins are maximized, at the expense of the universities’ research funds – in the Swedish case, publicly invested moneys, which go directly to private commercial publishing companies. The reason why this model can still be maintained is, to a large extent, the existing model for research assessment, where the publishers offer their services to uphold research quality for the academic community, which instead should be a responsibility of academia itself.

An article − written by a number of researchers from the network Open Science Community Sweden and published in the Stockholm daily Svenska Dagbladet in the spring of 2023 (Nilsonne et al. Reference Nilsonne, Dahlgren, Eklund, Danielsson, Carlsson, Innes-Ker, Nordström and Willén2023) − also argues in the same direction, i.e., that the situation becomes more and more untenable with double dipping and increasing costs, as Swedish universities and libraries today pay more than half a billion Swedish crowns for subscriptions, while at the same time also paying for publishing. Therefore, they challenge the Bibsam consortium and encourage it not to renew the agreement with Elsevier. The money, they argue, should instead be invested in open infrastructure, which would allow for immediate OA publishing. Furthermore, they argue that journals hiding their research results behind paywalls have played out their role. In the era of the internet, research should be immediately distributed and not be communicated according to the outdated principles of the printed press. The possibilities of rapidly and freely spreading results outside of journals are still quite limited as the high-level journals continue to give prestige to careers, especially those of young researchers: ‘The research community therefore goes totally out of sync as obsolete publication models are supposed to lead us to the research front’ (Nilsonne et al. Reference Nilsonne, Dahlgren, Eklund, Danielsson, Carlsson, Innes-Ker, Nordström and Willén2023). The authors further argue that it is not technology, but attitudes, that need to change, with adequate quality control and better ways of the assessing scholarly merits of researchers, where they are assessed not by the prestige of the journal where they were published but rather by the quality of their actual publications.

European Initiatives

In the Council conclusions prepared under the Swedish EU presidency, the problems were clearly addressed and taken into account. In the conclusion, the Commission is encouraged, ‘in the context of ERA policy action 2, to propose measures to remove barriers to access to and reuse of publicly funded research results and publications and data for research purposes at EU level.’ Furthermore, member states are invited ‘to update their national open access policies and guidelines as soon as possible to make scholarly publications immediately openly accessible under open licences and to make research data FAIR’ (Council of the European Union 2023: 7).

In its comments to the draft Council conclusions before adoption, LERU – the League of European Research Universities – stated that:

The global increase of OA articles due to transformative agreements has been quite impressive. However, with this positive picture of the accelerating transition there also follows a reverse side of the coin, as the costs are also increasing accordingly. LERU (2023) again:

LERU also emphasizes that it has become increasingly difficult for OA publishers ‘to agree financial contracts’ and also notes that this leads to a problematic inequity in the system – between those countries which can afford to pay and those which cannot. There is also a problem with fully Open Access publishers with regard to research funders: how to claim eligible costs. While publishers provide flat-rate agreements, funders require individual ‘payments’. This problem is not easy to deal with ‘unless funders are included in the agreement’, as LERU concludes.

Beyond Transformative Agreements

It is clear that Sweden has now reached the point where it is necessary to move beyond transformative agreements, to pay only for publishing and no longer for reading, as the country has almost reached the level where we are supposed, and want, to be. The next necessary step must be to cut the costs, that is, to reduce the total costs for the agreements, which have become an increasing problem especially for research-intensive universities, whose publishing costs are gradually increasing whereas the reading costs are not being cut. Of course, universities should pay for the actual costs related to publishing. But it is also clearly necessary to allow for the public money spent on publish-and-read agreements, where the costs are not transparent and clear, to go back into research, in order to be able to do more research as well as to reach the most important goal for open access. It is important to remember that the goal is not in itself to reduce the costs; it is to be able to share our results openly without hindering pay walls, and thus also to enhance research quality and transparency.

Against this background, the Swedish Association for Higher Education Institutions, in close collaboration with the Bibsam Consortium and the National Library, decided in 2021 to set up the task-and-finish group, Beyond Transformative Agreements. The goal was to propose a strategy for transitioning from transformative agreements to a financially sustainable system that stimulates the ongoing transition to a fully open publishing system. The group contained representatives from university leadership, researchers from different academic disciplines, research-funding agencies and library representatives, including the National Library of Sweden. The idea was also to investigate different publishing routes, and the work was supposed to build on international discussions and cooperation. Finally, it all winnowed down to four scenarios, which are not to be seen as mutually exclusive, but rather as four parallel paths to explore further and to implement in part or together. A final report has now also been submitted, based on these four scenarios.

The first of these scenarios is either not to renew or to quit transformative agreements and/or end agreements with commercial publishing houses, and not to renew agreements containing publication in hybrid journals, as this is – or was – not supposed to be Coalition S-compatible after 2025. Now that Coalition S allows for exceptions, the strategy remains basically the same, but rather emphasizing the goal of immediate open access and stressing the importance of moving beyond transformative agreements, that is, going from publish-and-read agreements to publishing agreements only.

The second scenario would be to establish a national Swedish open platform for publications, or to join forces with the European Union and their platform Open Research Europe (ORE). LERU (2023) argues in this connection that ORE has made a start by providing an alternative publishing platform, but there are still challenges to be addressed before the platform really becomes embedded in European research infrastructures. They suggest that what Europe may really need is ‘the development of an open, inter-connected, publicly owned infrastructure where all parts are inter-connected and speak to the rest’. They also argue that a ‘single pan-European system is not likely to work successfully’, and stress that next step should be to examine and build the case for such a development. Even though LERU here points to a weakness with ORE − from a Swedish perspective as a small nation − joining the European platform would mean connecting to other European countries in the same endeavour. It would also mean a link to the European Union, which throughout the years has become an active driver in the transition to Open Science, rather than Sweden isolating itself by establishing a national platform.

The third scenario would be to further explore avenues for diamond open access, in order to reach full open access without individual APCs or publication fees. The real publishing costs would then instead be covered by research-funding agencies, universities or learned societies, to mention three possible examples. This includes improving the opportunities for migrating researcher-owned journals from traditional publishers to other platforms.

The fourth scenario would be to further explore the rights retention strategy, adopted for example by several universities in Norway and in the UK, where the rights to the publication stay with the author, fully or partly. In connection to this, secondary publishing rights can also be explored. The latter are emphasized in the European Council conclusions on high-quality, transparent, open, trustworthy and equitable scholarly publishing. The conclusion ‘welcomes the introduction of secondary publication rights by a number of Member States into their national copyright legislation, enabling open access to scholarly publications involving public funds’ (Council of the European Union 2023: 7).

The Beyond Transformative Agreements working group delivered its final report, ‘Charting Sweden’s path beyond transformative agreements – analysis and proposals for strategic direction’, in September 2023 (Association of Higher Education Institutions 2023). The main conclusion of the report is that ‘it is vital for control of scholarly publication to reside in the research community, while also emphasising the need to reduce publication costs.’ Its primary recommendation, therefore, is for the Bibsam consortium

After having proposed a number of initiatives and actions in line with the four scenarios already mentioned, the group finally concluded that (Association of Higher Education Institutions 2023):

As early as 2019, a common group had been established with representatives from the research funding agencies as well as from the Bibsam consortium, with the aim of exploring the redirection of funding streams, but its active work was put on hold to wait for the final report from the Beyond Transformative Agreements group. Since the preliminary conclusions of this group have become public, the common group has taken up its meetings in order to establish a joint initiative in this direction. In line with the primary recommendation in the fall of 2023, the group has now agreed in principle to co-fund agreements with fully OA publishing houses, using a step-by-step approach with increased participation from the funding agencies, although the details concerning each agreement still have to be decided. This means that the research-intensive universities will not be alone in bearing the costs for the transition to OA, which is also fully in line with the 2020 research bill, later decided by parliament, which calls for research funding agencies and universities to work together to advance the transition.

Concluding Remarks

There are several voices from within the community of researchers that see the need to enhance the transition even further, as already mentioned above in connection with the article by Open Science Community Sweden (Brembs et al. Reference Brembs, Huneman, Schönbrodt, Nilsonne, Susi, Siems, Perakakis, Trachana, Ma and Rodriguez-Cuadrado2023). Its authors argued, more radically, that academic journals need to be replaced by a more modern solution. They maintained that not only affordability, but also functionality and replicability in science are at stake. The solution that they propose (Brembs et al. Reference Brembs, Huneman, Schönbrodt, Nilsonne, Susi, Siems, Perakakis, Trachana, Ma and Rodriguez-Cuadrado2023: 1) is one that may not only

Finally, they also argue for

As a follow-up to that article, Nilsonne (Reference Nilsonne2023) emphasized in the Stockholm daily Svenska Dagbladet the arguments for a new publication model. He also underlined that this needs to go hand in hand with a new model for research assessment, which has been taken into account by CoARA, the Coalition for the Advancement of Research Assessment, a joint initiative by the European Commission, Science Europe and the European University Association (CoARA n.d.). The initiative may be joined by universities, funding agencies and other research organizations, and the idea is to shape and develop a new assessment model together. In Sweden, the Swedish Association for Higher Education Institutions has joined the initiative, as well as several universities and funding agencies. It is important to closely follow the development of CoARA, both on a national and a European level. And the need for close collaboration within the academy is only increasing. In the words of Nilsonne et al. (Reference Nilsonne, Dahlgren, Eklund, Danielsson, Carlsson, Innes-Ker, Nordström and Willén2023): ‘Through working together, the academy can reclaim control over publication and research assessment, and deliver us from the private companies’ paywalls, whose profit interests will not be in the service of humankind’ (my translation from Swedish).

It is thus indeed a primary goal for academia to reclaim control over its own research results. The role of the Bibsam consortium, however, is not primarily to turn away from negotiations, at least not as its first option. The role of the consortium, following its general policy, is to conduct negotiations with the aim to cut costs maximally and to move away from publish-and-read to publish-only agreements. However, if this strategy turns out to be impossible, the consortium may at any time opt for turning down specific agreements. But this should not be dictated from the outside; it has to be decided by the universities together, just like when the previous Elsevier agreement was at first turned down.

Therefore, it has also become increasingly important to emphasize, again and again, that issues of open access are no longer a responsibility for the libraries, as they used to be for many years, but indeed a strategic question for university leadership. University leadership is responsible for the strategic priorities and the costs of the universities. It is necessary to maintain and, indeed even to a greater extent, to gain control also over the costs for publications. The goal should be to create a sustainable financial model that will allow universities to choose the right priorities for the future, both to enhance research quality and to promote the open sharing of results and data, in the service of humankind.