Review of relevant literature and the Norwegian national policy development

The literature on open access varies considerably and covers topics like publishing business models and costs [16–18], policies and prevalence studies [2, 19–21], researchers’ attitudes towards open access [22–24], predatory journals [25, 26] and different impacts of open access [27, 28], particularly academic impact [see 29]. These investigations are primarily empirical, and we have not been able to identify any studies on open access that use institutional theory, the study will therefore draw on the institutionalisation of the third mission. This is relevant as a parallel case in terms of institutionalisation and directly linked to open access in science policy. This relationship is also a topic in the last section of the literature review, where we give a brief account of the development of national open access policies in Norway.

Open access and its conceptual ties to university missions

Open access emerged not only as a response to access problems, but as a response to severe economic challenges for universities. The ‘serial crisis’ refers to the situation in the late 1980s when subscription prices grew while library budgets declined, which created an unsustainable economic situation for higher education institutions and their libraries, forcing them to reduce journal subscriptions [30, 31]. Digitalisation and internet-related opportunities made publishers’ marginal cost drop significantly, and there was hope that this would provide a solution to the serial crisis. However, given the nature of academic journals as a non-substitutable good, restricted access remained the norm.

The serial crisis paved way for the formalisation of open access through a series of declarations known as the Berlin-Bethesda-Budapest initiatives in 2002–2003, of which the Budapest declaration is the most widely cited. The Budapest declaration stated the goal of access to the literature by ‘all scientists, scholars, teachers, students’ which will ‘accelerate research, enrich education’, while other user groups where labelled as ‘other curious minds’ rather than industry or other specific non-academic actors [32].

Open access comes in three main flavours. Gold open access is tied to journals where licences enable free and extensive user rights. Green open access encompasses subscription-based journals where articles are put in a repository with subsequent open dissemination after an embargo of usually 6–12 months. Finally, hybrid refers to single articles in subscription-based journals bought free through article processing charges (APC) for immediate release. These are licenced in an identical or similar fashion to articles in gold open access journals. While the first open access declarations saw green and gold as complementary viable options, hybrid entered the scene later [33]. Furthermore, hybrid has been discouraged for economic reasons. The biggest challenge is double-dipping, where publishers charge both processing charges and subscription fees. Hybrid is therefore ineligible in most financing schemes unless part of agreements designed to push publishers into converting their journals to gold open access.

Open access reflects the ideal-normative approach of Merton setting up five central norms of science: communalism, universalism, disinterestedness, originality and scepticism (‘organised scepticism’ in the first version) [34]. Open access has been viewed as a direct translation of communalism [35], expressing how researchers share discoveries and give up intellectual property in exchange for social recognition [34]. Universalism also denotes the imperative of sharing research results widely. Studies argue for both a download advantage and citation advantage from open access [27], although the latter is connected to the type of open access that the author selects. Generally, green and hybrid open access show larger citation advantages than closed access, while gold open access advantages are less clear and can be disputed [19].

Academic impact is connected to the journals’ role in the reward system of science, where the most prestigious and highly ranked journals are often established and prestigious ‘legacy’ journals, as opposed to comparatively new gold open access journals. Journal rankings and their role in the academic reward system are often considered main barriers for the uptake of gold open access publishing [36]. The uptake of open access varies between the disciplines both in terms of levels and types. The medical sciences have the highest levels of open access, closely followed by the natural sciences, while the social sciences are lower but remain above the levels for the humanities [20]. The different modalities of open access also vary due to different structural and institutional factors, polices and other sociotechnical aspects (ibid.)

Although open access is tied to core research norms, it involves benefits for education–the first mission–since academic texts are key elements in curricula. Open educational resources include not only open access but the wider idea of open scholarship, access to educational opportunities and the adoption of open learning strategies [37]. A UK study found that 38% of journal articles licensed centrally by a collective management organisation and used for teaching purposes, were (some form of) open access, yet still a part of the collective licensing arrangement [38]. Although open access would benefit universities, especially with respect to the articles not licensed centrally, the largest potential is to give students open access to monographs and textbooks [38]. There is also a great potential in supporting sustainable lifelong learning for user groups like medical patients and their networks, health advocates and NGOs [27].

Considering benefits for the third mission, the scientific literature is a particular codified output from research work, which potentially is useful for both private and public sector actors. For example, it has been argued that lack of access to the scientific literature for small and medium-sized biotech firms hampers innovation [39], and open access is also desired by researchers in the private sector [40]. However, empirical support for benefits is rather limited. A review of the economic benefits of open science found indications of cost savings with open access, but evidence for economic returns are still considered patchy and diverse [41]. A few studies indicate that openly available articles have an advantage compared to the paywalled literature when considering patents. The U.S. National Institutes of Health’s 2008 open access policy was associated with an increase of 12–27% in patents citing its funded research [42]. Large pharmaceutical companies more often cite open access literature and publish more in open access journals [43]. Country reports from the UK, Netherlands and Denmark conclude that private firms save costs, but otherwise present little evidence for innovation benefits [44–46].

A more principal argument sees science as a public good and scientific knowledge as something that should be free and under little restriction [47]. Public good is a term borrowed from economics [48] and is also commonly applied in the context of open science [49–52]. For example, Dalrymple maintains that ‘Scientific knowledge in its pure form is a classic public good. It is a keystone for innovation, and in its more applied forms is a basic component of our economy’ [47, p. 35]. Although the question of whether open access will increase private sector innovation is both interesting and, in our view lacks empirical investigation, we proceed with the observation that policies have adopted the argument that access to the scientific literature is important for different forms of innovation among private and public actors.

Policies and funder mandates are considered the main drivers of open access the last decade [2], where the devil hidden in policy details influences their effect [21]. When, how, and how much of the literature is made open is connected to opt-out mechanisms, mandatory depositing and whether incentives and research evaluations contain open access points [53].

While the idea that the academic literature should be freely available is broadly shared, there is substantial disagreement on the best approach to achieve open access [2]. Policy preference may conflict with autonomy and other key academic values; the push for immediate open access by strong mandates, as in the case of Plan S, may exclude subscription based journals from the researchers’ options [54]. This is therefore seen as an infringement on academic freedom [13].

Studies of open access polices have mostly been concerned with what constitutes an effective policy and their performance [e.g. 21], rather than institutional responses and practices. A study that moves beyond quantitative effect-analysis is a Dutch report that found a top-down induced growth in open access driven by national policies [55]. An Australian investigation, noting that there is no unified national policy, found that half of the country’s 40 universities had no open access policy, and those that had, exhibited a wide variety in their approach in terms of definitions, depositing time and sanctions [56]. Institutional polices were introduced over a period of 17 years with some references to generic benefits to society. Furthermore, references to the two major Australian funders were found in all 20 policies, with a high degree of compliance (ibid).

Institutionalisation of new missions and practices in universities

We have seen that the adoption of open access in science policy is intimately connected to beliefs about benefits to the wider society. Policy documents argue that open access will lead to better utilisation of research-based knowledge in society, both in Norway [57] and internationally [9, 58]. The EU policies in particular frame open access and open data as fuel for the innovation engine, ascribing a commercial third mission logic to open access. The third mission is therefore not only useful as an example of the possible institutionalisation of new tasks in universities–it has become an important rationale for open access which connects to an increased interest in university-industry relations and how universities may contribute to economic and social development.

This interest has produced investigations of how universities introduce and implement new practices. A frequent framing is Clark’s [59] investigation of ‘entrepreneurial universities’ that have transformed themselves and developed significant support of the third mission [cf. 4] of innovation and societal engagement. Noticing an increasing gap between demands to universities and their capacity for response, Clark found that there are several aspects to successful institutional transformation. This includes the universities’ ability to integrate new activities with existing ones, and their use of new activities to protect the academic heartland of collegiality, autonomy, academic freedom and achievement. The academic heartland thus conceptualises traditional academic values, which normally are sought protected when confronted with new demands, which can make changes more likely to fail.

Although Clark’s and other’s empirical investigations have shown that many universities manage to institutionalise new practices without threatening the heartland, severe challenges have been identified. Some universities, even when they are heavily engaged in various forms of societal engagement, have struggled to meet policymakers’ expectations of third mission activities [7, 8, 14]. Responses to external expectations have often failed to become integrated with existing practices and strengths, and therefore remained peripheral or in tension with education and research missions. Often new external demands are poorly suited to higher education institutions’ unique profile [6, 14]. The blame is often put on external policies rather than institutional agency, for example that policies push some aspects of the third mission, commercialisation in particular, in favour of others that may engage all disciplines and staff members [7, 8, 14]. Another challenge is that uniform demands (‘one size fits all’ policies) to a heterogeneous sector and competitive funding based on problematic indicators create pressures that universities struggle to integrate effectively [6]. This leads to unique responses because each ‘university is the product of a distinct process of social, economic and intellectual development’ (ibid. p. 738). Universities are thus subject to isomorphic forces [cf. 60] or heterogeneous institutional logics [e.g. 61] that need to be resolved in one way or another [6].

In an in-depth historical study of Stanford University, Colyvas & Powell [5] show how an interplay between top-down and bottom-up internal initiatives contributed to the institutionalisation of technology transfer practices over time. They use this empirical analysis to set up a conceptual framework for understanding three interrelated dimensions of such processes. First, institutionalisation concerns (changes in) organisational structure, practical action plans and ensuring reproduction of the new activity through learning, training programmes and dissemination. Second, legitimacy refers to how new standards develop, how norms of what is appropriate behaviour are enforced and how boundaries to existing practices are remade. Third, taken-for-grantedness implies gradual changes in practices and roles and the emergence of new categories. Colyvas & Powell furthermore indicate three states for each of the dimensions: low, medium and high. A low state can be related to symbols and vocabularies drawn from external sources, medium is represented by institutional-specific ways of talking about the issue in question, and high is when a rich, local language becomes widely accepted.

We initially argued that the ‘openness’ of open access may be more in line with research norms. However, tensions can still emerge at the intersection of research, the third mission and open access, not least due to how the outside pressure in demands is put on universities.

The development of national open access policy in Norway

A core interest in this study is how universities respond to government expectations, which comes in the form of policies on open access. One of the first official statements on open access in Norway appeared in a 2005 White Paper, framing it as a response to the increasingly high subscription fees for academic libraries, which could hamper access for researchers, particularly in developing countries. The ministry applauded the introduction of open access journals and repositories for green open access [62]. In 2008, the Ministry asked the interest organisation of universities and colleges for advice on how to implement open access, which led to the establishment of a collaborative infrastructure project [63]. Future policies were therefore based on the advice of higher education institutions with important contributions from the universities to the initial national open access policy.

A 2009 White Paper on science defined open access as one of nine central policy topics and argued for the importance of this for researchers, firms and the general public [64]. Even though open access was seen as advantageous, it was not unequivocally pushed: ‘in principle, the government believes that all public research should be openly available, unless other considerations prevent it’ [64. p. 125] and ‘researchers should, however, not be required to deposit articles in open repositories if publishers do not allow this.’ [64, p. 127]. These reservations challenged the general point made that open access largely could be realised without compromising academic freedom and copyright [64]. The infrastructure project was made permanent and assigned to a newly established organisation Cristin (Current Research Information System in Norway). The Research Council Norway (RCN) issued a policy the same year arguing for open access so that ‘the private sector … easily can access published scientific results’ [65]. This policy and the 2009 White Paper served to establish the idea of open access; they also highlighted green open access and saw depositing in repositories with embargo as sufficient.

Following a regular four-year cycle for White Papers on science policy, the 2013 White Paper reaffirmed the principles of open access, largely with the same arguments [66]. Again, academic freedom was given priority to considerations of access. The policy also condoned ‘the taxpayer argument’–an argument that it is unreasonable for publicly financed research output not to be openly available, as the taxpayer has already financed it [66]. The same white paper gave Cristin the responsibility for developing indicators for open access similar to other science policy indicators. In 2014, RCN revised its open access policy and included funds for gold open access publishing.

The Ministry of Education and Research launched its national guidelines for open access in 2017 [67]. These were based on a report from an expert group of researchers [68] with a goal of full open access by 2024. The guidelines affirmed Norway’s continued role as a ‘cautious forerunner’ in the pursuit of open access for the benefit of research, the wider society and innovation [68]. They prescribed gold open access as the main ambition but allowed for green open access without specifying a maximum embargo. This signalled a shift in preference for immediate open access, including hybrid when part of a transformative arrangement. Depositing was made mandatory for all articles, and the ministry maintained that when the system ‘allowed for it’, only deposited articles would be considered eligible for the results-based funding system, thus connecting open access to future financial incentives.

Norway’s only research council RCN was one of the original signatories of Plan S [1], a policy with a non-negotiable preference for immediate open access. The first version allowed only gold immediate open access publications, a requirement which was softened in later revisions allowing immediate green open access yet remained stricter than the national guidelines. Although Plan S primarily concerned research council funded projects, the policy was launched with a significant political investment from the Norwegian Ministry. A 2018 White Paper signalled a tempo shift and portrayed Norway as a forerunner in the pursuit of open access along with the Netherlands, the UK and Germany, and further aligned Norwegian policy with European polices on open science [69]. Other initiatives have followed, for example a 2019 report on creative commons licences in academic publishing discussing whether demands on the researchers to choose creative commons is at odds with academic freedom [70].

With the Norwegian policy development as background, our analysis will use official university documents, similar to Kitagawa et al. [6], Colyvas & Powell [5] and Wakeling et al. [56]. The main concepts we bring into the analysis concerns how institutional polices and strategies handle open access and its relationship with the three missions and the ‘academic heartland’ and further how this reflects or contrasts with national policy. We explore whether the implementation of open access is a process that can be separated into institutionalisation sub-dimensions that develop into distinct states or stages over time.

The role of open access and the three missions

All ten universities tie openness to research in strategy documents, and they furthermore connect openness to the third mission and societal engagement. However, as the term openness implies, statements are often general, and the strategies do not necessarily apply the term ‘open access’. An example is the Norwegian University of Life Sciences (NMBU): ‘NMBU will contribute to a knowledge-based and open public debate.’, ‘have an open and respectful collaboration with the world around us, both with the public and private sector. [..] ensure that funding is used efficiently and purposefully and for the benefit of the whole of society’, concluding that ‘knowledge must be shared.’ These statements support the general idea of open access, and we find that these passages from the main strategy are quoted and operationalised in NMBU’s open access policy. The Norwegian University of Science and Technology (NTNU), which did not mention open access in its 2011 strategy, stated that ‘[the university] disseminates research results openly’ in the third mission section of its 2018 strategy. Open access is commonly integrated in the context of third mission activities, outreach and science communication also in other universities’ main strategies. The University of Bergen (UiB) underlined societal contact and the role of science in society in their 2011 strategy without invoking open access. In 2016 and 2019, similar strategy statements brought up open access as means of dissemination.

Seven institutions connect open access to the third mission in the sense of innovation and benefits for the private sector. However, innovation is a rather minor part of main strategies, and they often contain general claims that the private sector will benefit from open access. An example is the University of Oslo (UiO), which emphasised open access and innovation in its 2011 strategy. This was followed by a separate innovation strategy for 2013–2015 that included open access, but the claim was downplayed in the next (current) strategy.

However, we also see examples of strengthened links between open access and innovation. The University of Tromsø (UiT) made generic statements about open access and societal contact in its 2011 strategy, while the most recent strategy portrays UiT as ‘a driving force for increased innovation to contribute to business and development in the public sector’ due to a ‘strong culture for dissemination through open channels of publishing’. NTNU’s publishing policy (2014–2020) states, ‘we know that scientific publications are a central source for innovation in the private and public sectors’. UiB’s open science policy (2020) puts forth a similar argument, referring to technological changes and digitalisation and makes a connection between openness in science and open innovation.

Open access is thereby more frequently integrated into statements about the universities’ third mission than statements about research activities (second mission). While there is an innate connection between open access and research, the context of research in strategies is often excellence or international collaboration. Open access is instead found in lower-level strategy documents and internal web pages about research.

None of the main strategies mention open access as supportive of education, while only two general open science policies cover educational resources. NTNU’s open science policy ‘includes all results from research, education and dissemination activities’ and UiB’s open science policy includes a section on open educational resources to ‘facilitate employees’ sharing of educational resources and make educational resources openly available in cases where it does not conflict with professional or legal considerations.’ Eight institutions do include students in their open access policy, primarily through mandating or encouraging students to deposit master theses in the institutional repository.

We also find that although the universities highlight the importance of open access for the third mission and various societal aims (democracy, critical debate, innovation etc.), we find no examples or documentation in strategies or annual reports of the benefits of open access to a wider audience that could lend legitimacy to the policies. Open access accounts (numbers, graphs) in annual reports are presented unsystematically and summarily, if at all. Annual reports are, however, used for other open access purposes, for instance introducing goals and expectations for the next period, (plans for) revisions of policies, or simply stating the importance of open access. We also note an interesting rhetoric twist in NTNU’s narrative in the annual report for 2020, which reports the decreasing shares of closed articles instead of the overall increasing open access numbers.

Table 2 shows the presence of any reporting of numbers on growth and prevalence of open access in annual reports, either in the form of tables, graphs or writing. We have not included reporting on other aspects related to open access, for example financial aspects (costs of open access) or information campaigns.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Timeline of annual reports.

Marked cells signify whether the report incorporates any numbers on the growth and prevalence of open access.

https://doi.org/10.1371/journal.pone.0273091.t002

Eight institutions have revised their open access policies since 2009, primarily stating a preference of gold open access, even if the policies default to green open access and have opt-outs for green. Furthermore, all universities require their employees to deposit research articles in the institutional repository, with three institutions supplying opt-out-possibilities.

There are interesting differences in compliance measures. UiO and the University of Stavanger (UiS) have chosen a formal approach by incorporating mandatory depositing in employment contracts following considerations tied to academic freedom and intellectual property rights (IPR). The other universities have not put this aspect into employment contracts.

Over time, we find that open access policies use formulations we interpret as stronger with more insisting language. For example, the NTNU policy of 2014 stated that ‘knowledge ought to be easily available’, while the 2021 revision replaced ‘ought to be’ with ‘must’. Similarly, the former UiB policy ‘asked’ all employees to deposit a version of scientific articles, while the current policy is more insistent.

In some cases, there is a shift in the responsibility to fulfil open access obligations. NMBU’s 2014 policy stated that ‘knowledge about open access is to be develop through information and training aimed at university employees’, while the 2019 version declared that the researchers are responsible for acquaintance and compliance with open access regulations.

Four institutions have implemented incentives. UiT and USN have incentives directed at the individual researcher since approval of sabbaticals, tenure and promotion will take open access into consideration. UiO and UiB have anticipated the forthcoming arrangement in the national guidelines and as such place negative incentives at the department level. Scientific articles are not eligible for the national reimbursement program unless they are deposited, regardless of open access type. While this primarily has economic consequences at the institutional level, trickle-down arrangements that allocates funds back to the departments incentivise local action [78]. UiO has further announced that open science is one of the elements which should ‘change the routines for evaluation of promotion and tenure’.

OsloMet, one of the forerunners on open access in Norway, introduced incentives for depositing as early as in 2010, but later removed them as a part of the institution’s merger process in 2018. We do not know whether this is due to legal aspects in the merger process, or whether incentives no longer were deemed necessary and depositing behaviour was seen as institutionalised. The latter may be a reasonable assumption, since OsloMet had over 95% of all articles deposited in its repository as early as 2010 [79].

Six institutions discussed open access and the boundaries to academic freedom and intellectual property rights (IPR) in their documents. All of them prioritised academic freedom and legal considerations with respect to ownership of texts and emphasised the individual’s journal selection rights. This is mostly statements about granted rights; Nord University typically confirmed that ‘as a researcher, you have the right to choose publishing outlet when publishing your work’.

We summarise the timeline of the broader integration of open access at the universities in Fig 1, which illustrates if and how the institutional strategies integrate open access generally and/or with either of the three missions. We use colour codes for three levels of open access engagement (low, medium, high) in strategies, the same colour codes are used to signal whether policies have evolved in terms of language use and the use of mandates. In addition, we indicate whether polices prefer immediate gold open access.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Summary of the development of the ten Norwegian universities’ strategies and policies.

Summary of the development of the Norwegian universities’ strategies and policies. Dashed vertical line indicates when National guidelines were introduced in 2017. Colour grading is used to indicate level of emphasis of open access with grey (low), light green (medium) and green (high). Letters signify G = General open access, E = Open access connected to education, R = Open access connected to research, 3M = Open access connected to the third mission.

https://doi.org/10.1371/journal.pone.0273091.g001

Our preliminary summary is that all Norwegian universities have incorporated and strengthened their engagement in open access throughout the period. Open access is primarily seen as part of general third mission activities and often tied to science communication. Benefits of open access are not documented in annual reports.

Cross-institutional comparison and alignment with national policy

All institutions refer directly to national documents issued by the ministry. References and quotes primarily invoke the national guidelines, but also the 2013 White Paper. Eight institutions have revised their open access policies, inspired by the national guidelines, which is stated explicitly. For example, UiB stated in its 2020 policy that ‘the government has issued guidelines for the dissemination of research data and scientific articles, which lays the foundation for the activities at UiB.’

When the national guidelines were drafted, all institutions were invited to comment. They expressed strong overall support for the guidelines. The National guidelines proposal further asked institutions to revise their policies in accordance with the guidelines, of which six complied. The remaining four wanted to introduce stronger measures than what the guidelines proposed.

As expected in a country with one research council, its influence is seen in university policies: nine institutions refer both to the Research Council of Norway and the international Plan S. NMBU and UiA explicitly write that their policies are based on Plan S, while NTNU and UiB state that the council’s policy (which includes Plan S) is the foundation of their own policy.

University open access policies share many similarities, and we observe a strong coherence with the national level and the national guidelines. All current institutional policies default to green open access with opt-out possibilities for final dissemination, and all mandate depositing. Although the universities accept hybrid open access, this is only tied to financial support through national agreements and not in dedicated institutional open access funds. This is also in accordance with national guidelines.

Eight universities further state a preference for publishing in gold open access journals when researchers are faced with the choice of otherwise equal options. The last two do not state an explicit preference, but still base their policies on compliance with the national guidelines. Preferences for gold open access publishing are usually formulated cautiously and encouragingly.

As mentioned, UiO, UiB, UiT and USN connect open access to different types of local incentives or as a basis for the national results-based funding indicator. These institutions have therefore implemented mechanisms that are planned but not yet implemented at the national level.

Our observation is that national documents lend legitimacy to and anchor institutional policies, not least seen in revisions performed after the national guidelines appeared. Thus, all Norwegian Universities show a high degree of commitment to open access, and the motivation for open access as expressed in the national guidelines is unanimously shared.

The academic heartland and open access

We have found coherence among university policies and strong links to national policies. This is not merely a process of adaptation and compliance–universities have been formally asked for advice and have been able to influence the national policies. As such, the idea of open access has probably also become institutionalised in the policy system, including the Ministry and research funders.

Still, we observe tensions surrounding open access, where Plan S is accused of interference with academic freedom [13]. Open access has wide support, but not when perceived as in conflict with core values in academia, captured in the expression ‘the academic heartland’ [59]. Clark noted tensions and conflicts with the introduction of third mission activities (ibid.), where stimulating collegiality, autonomy, academic freedom and achievement was of critical importance for institutionalisation. We argued in the introduction that such tensions are less likely in the open access case, since open access is more in alignment with scientific norms and traditional missions. Our main empirical findings offer support: national open access guidelines were largely applauded by the universities.

However, this is not the case with Plan S. The most relevant difference is that the Norwegian national open access guidelines do not view any (type of) journal as ineligible, while Plan S excludes the choice of journals that do not meet the requirements of the policy. In addition, the legal aspects of the exclusion of journals is considered within the law in the case of Plan S, as it ultimately is a voluntary arrangement to sign Plans S contract agreements [70]. This is not the case when scientists perform research as regular employees according to Norwegian legislation, and this is the domain of institutional policies. In the foundational texts of current legislations, it is stated explicitly that the choice of journals belongs to the researcher (ibid.).

In a broader perspective, this consideration helps explain the profile of local and national policies. Open access is not simply an issue of academic ideals and Mertonian norms, but also of academic freedom and norms protected by (Norwegian) law, and many universities have pushed for an even stronger formal and legislation-based protection of academic freedom. There are thus limits to how strict and compulsory rules of open access can be designed. No institution ultimately mandates open access. There are opt-out-options available even though the researcher must comply for the benefits of open access to become realised.

This reasoning is also found in our material on the discussions around academic freedom and open access and echoes that researchers’ choice of publication channel is regarded a part of their individual choice as university staff. The centrality and legal status of academic freedom naturally leads to opt-out clauses and explains why gold open access publishing at best is stated as a preference.

Legal matters are also the reason policies contain opt-out clauses for final dissemination in the case of green open access. This is to avoid situations where researchers run out of options if co-authors or publishers refuse to grant permissions for green open access. Instead, the first initial step of depositing is made mandatory as depositing poses conflict neither with legal considerations nor with academic freedom. These considerations are not only found at the institutional level, they are also part of the national guidelines [68].

We interpret this as a confirmation of the co-evolutionary aspect of the institutionalisation process. National guidelines and governmental decrees have respected the ‘academic heartland’ in all policy iterations and sought advice from the scientific community when implementing them. We therefore find fewer tensions in the introduction of open access as expressed in the national guidelines, compared to the introduction of the third mission [6–8, 14, 15]. The influence of Plan S on some of the polices was unexpected mainly because of the controversies surrounding the policy. However, it seems Plan S has served as an inspiration for institutional policies rather than as a script for detailed implementation. The academic home remains protected, and we interpret the influence of Plan S as a sign of commitment.