Goals and evaluations of transformative agreements: an exploratory study of consortial approaches

Terminology

There are several different terms used in this space: transformative agreements, transitional agreements, read‑and‑publish agreements and publish‑and‑read agreements, with some terms used interchangeably and others used with specific differentiation in mind. The intricacies of different kinds of agreements related to OA publishing have been well‑covered by others (Borrego, et al., 2021; ESAC, n.d.; Hinchliffe, 2019; Rothfritz et al., 2024; Wise & Estelle, 2020) and thus will not be addressed here. As noted by Borrego et al., the term ‘transformative agreement has become an umbrella term that encompasses a variety of deals’ (2021, p. 220). Thus, for simplicity, the term transformative agreement (TA) will be used throughout the following discussion.

Diversity in terms and conditions of TAs

Along with differing terminology and definitions of TAs, the terms and conditions of the agreements themselves vary widely. Even among agreements in the same category, ‘it is nearly impossible to find two identical contracts. Institutions negotiate different clauses with different publishers, and publishers offer different agreements to their customer portfolio’ (Borrego et al., 2021, p. 227). Elements such as costs, financial shift, coverage, license type and article types covered vary considerably (ESAC Initiative, n.d.). This variety is also highlighted by Rothfritz et al. (2024) in their study of over 1000 TAs in the Efficiencies Standards for Article Charges (ESAC) registry. In Borrego et al.’s 2021 analysis of agreements from the ESAC registry, the only commonality in their sample of 36 full‑text TAs was allowing authors to retain copyright (p. 227).

One important example of how TAs differ is whether a given agreement includes gold journals or is limited to hybrid‑only journals. This differs not only by publisher, but by individual agreement. As explained by Baquero‑Arribas et al. (2019), ‘publishers do not include gold journals in most of Read & Publish…in their view this type of journal has never been part of subscriptions portfolios, and therefore libraries have never paid any fee for them’ (p. 19), which is in alignment with Schimmer et al.’s (2015) original vision for TAs: transforming existing subscription fees into funding for OA publishing. Yet the fact that some publishers include publications in gold journals further complicates the picture. This inclusion may be tied to the fact that some agreements do not simply transform existing subscription spend but add extra costs – for example, including previous APC spending in the fee. However, it is difficult to easily map gold inclusion or exclusion to any underlying fee structure (Borrego et al., 2021, p. 227).

The fee structures associated with TAs also vary. Some agreements are essentially one overall cost that includes reading and publishing, with no specific cost associated with each element (Pinhasi et al., 2021, p. 5; Siewicz et al., 2021, p. 27), whereas others specifically outline which portion of the fee is for reading versus publishing and may involve a specific price per paper, essentially using an APC as part of the calculation.

Some agreements include caps on the number of OA articles published, and there may be additional fees for any articles over a contract’s allocation (Bieg, 2022, p. 5; Walsh et al., 2024, p. 7). In contrast, other agreements provide unlimited publications. For some, not only are there caps, but if an institution does not publish its full allotment, articles do not roll over into the next year (Muñoz‑Vélez et al., 2024, p. 89; Rawlins, 2024, p. 17). For the Springer Compact agreement in Sweden, Olsson et al. (2020) noted that authors only published 81% of the expected number of articles and the remaining prepaid articles were not reimbursed, representing a lost value of €1,713,800 (p. 13).

Regarding assessment and evaluation, it is vital that the specific TA in question is evaluated based on its unique terms and that outcomes of evaluations of other TAs for other institutions or consortia are not assumed to apply to a different contract. Caution should be taken in any attempts to evaluate TAs as a whole, as the agreements are too diverse for overarching evaluations to be meaningful when applied to any individual agreement. It is also important to question the complexity itself, as it ‘obfuscates their true cost and this model’s long‑term implications remain undetermined’ (Farley et al., 2021, p. 298). This complexity may divert attention from identifying more significant underlying issues in the model.

Increase in number of OA papers

A commonly cited marker of the success of TAs is the increase in the number of OA papers, either globally or consortium‑wide (Brayman et al., 2024; Campbell et al., 2022; de Castro et al., 2024; Rothfritz et al., 2024; Olsson et al., 2020, p. 29). Campbell et al. posit that ‘there can be no doubt that TAs increase global access to research’ (2022, p. 1), but this must be balanced by the fact that in an analysis of the metadata of over 700 agreements and nine million journal articles, Jahn (2025) found that ‘current levels of implementation of transformative agreements is insufficient to bring about a large‑scale transition to full open access’ (p. 2). This finding is echoed by Brayman et al. (2024), who – based on OA rates from 2018 to 2022 – found that it would take 70 years for the Big Five publishers to become fully open access (p. 14), and that ‘unless these TAs are extended to cover a broader range of the publishers’ output, these TAs are unlikely to convert these publishers’ portfolios to fully OA’ (p. 15).

Although there is broad agreement that TAs increase the number of OA papers, it is unclear if some of these articles are just replacing papers that would have been OA via the green route. Wenaas (2022) found that an increase in hybrid OA from 2019 to 2021 coincided with a decline in green OA, suggesting that TAs may have replaced one OA modality with another (p. 18). Bakker et al. (2024) similarly noted that when comparing the proportions of publications before and during TAs, the overall proportion of hybrid publications increased, while green publications decreased during that same period (p. 83). Catterall et al. (2025), in their analysis of the progress of OA in Australia and Aotearoa New Zealand, similarly found that promoting TAs ‘had the unintended consequence of depressing the alternative pathways to OA, the “green” route’ (p. 13). In contrast, de Castro et al. (2024, p. 19) found that accepted manuscripts continue to be added to institutional repositories, even when the versions of record are published OA via a TA.

Further complicating the situation is the fact that gold or hybrid OA status overrides green OA classification in some data sources, such as OpenAlex and Unpaywall; therefore, it is possible that some gold or hybrid publications also have a version available as green OA (Bakker et al., 2024, p. 83; de Castro et al., 2024, p. 34). It is also important to note that, even if an overall correlation is found between a TA being in place and a decrease in green OA, it remains a conjecture that the TA itself is causing individual authors not to pursue the green route, since we cannot determine with certainty what an author would have done in the absence of an agreement.

Another important qualification is that green OA is not definitively viewed as an equivalent to a hybrid or gold OA publication, as the green article is not necessarily the version of record. Writing from the publisher perspective, Inchcoombe (2022) posits that green does not offer the full benefits when compared to hybrid or gold and ‘isn’t the version researchers want’ (para. 11). Shearer (2022) counters this, arguing that publishers have a vested interest in opposing the green route, as they have ‘a huge financial stake in transitioning the system to a pay‑to‑publish model to maintain their significant profits’ (para. 8). Regardless of whether the green route provides equivalent value to gold or hybrid options, it is important to balance the assertion that TAs increase the number of OA publications with the fact that a portion of those papers could have been available, in some form, via the green route to OA.

Impact on author publishing choices

A significant concern is if, or how, TAs influence an author’s decision on which publisher to choose. There are differing conclusions in the literature about the impact on author behavior, which may be due to differing methodologies or varying definitions of a change in behavior. It is also possible that authors in different countries are differentially influenced, especially given differences between policies and funder mandates. However, findings can vary even within the same country. An analysis of publishing patterns in Germany showed that ‘eligible researchers have increased their publications in DEAL journals at the cost of other journals’ (Haucap et al., 2021, p. 2027), whereas Schmal (2024), also looking at publication patterns in Germany, found that TAs ‘hardly transform the publication behavior of academics in the sense that they shift their publications to eligible publishers’ (p. 1964).

From a small society publisher's perspective, Anderson et al. reached a similar conclusion, finding that there was no large increase in submissions after the implementation of a TA; rather, the TA simply supported an institution’s typical output to become fully OA (2022, p. 8). There are also institutional examples, such as McLain and McKelvey (2024), who found that in the seven TAs in place at their library, the number of articles published matched previous publishing numbers, suggesting limited changes to author behavior (p. 4). A case study in Sweden provides a counter‑argument, albeit anecdotal, of a change in author behavior, noting that there has been “a shift in researchers’ questions from ‘which journal is best for my article to get published in’ to ‘where could I publish my article without having to pay an APC’?” (Parmhed & Säll, 2023, p. 6). In an analysis of TAs in Austria, Pinhasi et al. (2021) indicated that there was an increase in the number of papers published in their Springer and Wiley TAs (p. 9), but it was not clear if this was at the expense of publishing in other venues or merely an increase in publishing in general.

Given the complexity of the environment and the challenge of controlling additional variables beyond the presence of a TA, it is challenging to draw definitive conclusions regarding the impact of a TA on author publishing choices. However, it remains important to monitor and evaluate how these agreements influence author behavior to ensure that publishing decisions continue to reflect academic priorities instead of being shaped by financial factors or choices institutions have made to enter into agreements with specific publishers.

Journal prestige and the promotion and tenure system

Another theme that emerged in the literature is the role of journal prestige, including its consequences for promotion and tenure and how this limits the ability to move away from journals provided by large commercial publishers. Even though researchers are generally in favor of OA, it is not typically one of the main reasons they choose a particular journal; rather, topical match and journal reputation play a larger role (Inchcoombe et al., 2022, p. 11; Olsson et al., 2020, p. 21; Wenaas, 2022, p. 2). That OA is not a main factor in journal choice is echoed by Harris et al. (2024) in their study of publishing patterns of researchers funded by Fonds de recherche du Québec, which found that ‘dominant commercial publishers are still the primary places where researchers are publishing’ (p. 43). Similarly, in an evaluation of TAs in the UK, a Jisc report concluded that expanding TAs to include a broader range of publishers beyond the major commercial ones ‘does not appear to have changed author behaviours and UK authors continue to choose traditional publishers to disseminate their research.’ (Brayman et al., 2024, p. 15).

To reduce the reliance on certain prestigious journals, a change is needed in the way researchers are assessed, evaluated and recognized. This change in research assessment would need to come first, to enable libraries to move away from the large commercial publishers. This move also requires collective action, as explained by Haucap et al. (2021), stating, ‘while all scholars and scientists may jointly be better off if the best research would be published OA in low‑priced journals, no individual scholar has strong incentives to be the first to move’ (p. 2029). This collective‑action problem, along with the slow pace of change in the culture of research assessment, means that libraries are somewhat constrained in how they support the transition to a fully OA publishing system. Although libraries can advocate for, educate and work with stakeholders at their institution, changing research incentives is not within their purview.

As explained by de Castro et al. (2024), one of the challenges in the development of OA ‘has been the severe decoupling between the end‑users of the research publishing ecosystem – the researchers – and those advocating for its reform’ (p. 67). Researchers are often isolated from the financial and long‑term impacts of the scholarly publishing ecosystem. This means their publishing choices are more influenced by research assessment criteria and culture than by elements such as cost, the publisher’s business model and long‑term impacts on the sustainability of the scholarly publishing system. This separation of the end‑user, whether readers or authors, from the cost also applies to the subscription model; this is simply a continuation of the problematic nature of that separation.

Failure to disrupt the scholarly publishing oligopoly

Although TAs have been criticized for failing to disrupt the scholarly publishing oligopoly and continuing to consolidate the scholarly publishing market (Farley et al., 2021, p. 300; Shu & Larivière, 2024), these agreements are not actually aiming to completely change the system. In a paper that lays out some of the foundational principles that led to the development of TAs, Schimmer et al. (2015) stated that ‘the goal is to preserve the established service levels provided by publishers that are still requested by researchers, while redefining and reorganizing the necessary payment streams’, with the ultimate goal of all publications being OA (p. 1). Researchers want, or need, to publish in certain journals, many of which are hybrid commercial journals. Therefore, if the main goal is to increase the number of OA publications, then TAs offer an avenue to do so without requiring authors to change where they publish.

The ‘transformation’ in TAs, as defined by ESAC, is to transform ‘the business model underlying scholarly journal publishing… from one based on toll access (subscription) to one in which publishers are remunerated a fair price for their open access publishing services’ (ESAC, n.d.), not to transform scholarly publishing from a commercial to a non‑commercial system. This is not to suggest that the issue of whether TAs contribute to the current oligopolistic situation in scholarly publishing should not be explored, but this concern does not inherently reflect a flaw in the agreements themselves, particularly when the likely alternative involves continuing to pay substantial fees for subscription‑only packages from the same providers. While it is fair to point out that maintaining the current oligopoly is a potential and significant downside to these agreements, it is important to contextualize this criticism in light of the comparable drawbacks of a traditional subscription‑based model, which are also highly dependent on commercial publishers.

Changing role of academic librarians and the academic library

In addition to providing access to content and looking at the cost per use of resources, librarians working in collection development now must take into consideration factors such as author experience, publishing patterns, rights retention and other publishing‑related elements as part of their work (Baquero‑Arribas et al., 2019, p. 23; McLain and McKelvey, 2024, p. 3; Walsh et al., 2024, p. 18). This shift towards agreements that include publishing also changes evaluation criteria and approaches, as ‘such complex transformative agreements cannot be evaluated with the conventional tools used by libraries to benchmark subscription‑only conditions’ (Schimmer et al., 2021, p. 25).

In their paper examining how TAs are transforming libraries, Campbell et al. (2022) posit that ‘the core business is no longer the provision of access, but open access publishing’ (p. 2). This represents a significant change to the role of the library and could have implications for workflows, job descriptions, the organizational structure of libraries, budgets and more. This change also raises the question of how this transformation serves a library’s population of readers. While readers benefit from being able to freely access OA papers, they are not necessarily directly served by libraries focusing their efforts on matters relating to publishing.

Theme: number of OA papers

One of the most‑mentioned elements, both in goals and evaluation, was the increase in OA papers resulting from these agreements. This is perhaps among the most straightforward elements of TAs to evaluate and is easily quantified. This assessment is usually discussed either as an increase in the number of OA papers within a particular agreement or as the existence of multiple TAs contributing toward an overall proportion of the institution’s authors’ output being made available OA. Although analysing the results of the evaluations undertaken by the consortia was not in the scope of this study, it should be noted that almost every consortium that commented on the number of OA papers indicated a substantial increase and named this as a clear benefit of the agreement. This marker of success is also frequently noted in the general discourse on TAs, as outlined in the preceding literature review.

Theme: global impact

While nearly all consortia reported an increase in OA publications, it is important to distinguish these localized increases from trends at the global level. This can be seen in Bibsam’s evaluation, which states that TAs have contributed to Sweden’s high proportion of OA papers, and notes that, globally, the overall change is very slow (Beyond Transformative Agreements Working Group, 2023, p. 7). There is a difference between institutional impacts, country impacts and the global impact of TAs. Gains in the proportion of OA papers for an individual institution or country, due to a given TA, should not be conflated with an overall change to the system, as can be seen in the analysis provided by Brayman et al. on behalf of Jisc, where they observed that ‘the increase in global TAs does not appear to have made a material impact on levels of OA, either. We observed low rates of journals being “flipped” to fully OA between 2018 and 2022’ (2024, p. 14). Four of the consortia included the global impact of TAs in their evaluations, with two specifically noting that despite the rapid increase in TAs, globally, the publishing system is changing very slowly.

Theme: cost

Cost‑containment, or cost‑neutrality, was the most‑mentioned goal or principle and was also the most frequently used element of analysis among the consortia sampled. Although cost‑neutrality should be quite straightforward, there is an important distinction between what constitutes cost‑neutral for libraries, for institutions, and for academia as a whole. While many consortia indicate that TAs should be cost‑neutral or contain costs, it is important to note which elements are being compared to make this determination. The importance of defining what cost‑neutrality is being measured against is highlighted by Borrego et al. (2021), ‘if the expression “cost neutral” excludes APCs, there is little evidence that TAs are constraining costs’ (p. 227). Some of the consortia that rate TAs favorably in relation to cost generally do so in relation to total previous spending with the publisher, including APCs paid directly by researchers outside of the agreement. As outlined by Pinahsi et al. in their discussion of the finances of OA, ‘what exactly “cost neutrality” entails is often debatable. Depending on each individual publisher, it may be based on a combination of the University’s historic spend, previous publishing output and, in some cases, APCs paid to date’ (2020, p. 6). Jisc’s review of its TAs clearly outlines that it considered the APCs previously paid when determining that ‘TAs have been a successful mechanism to constrain costs when reviewing at the sector level.’ (Brayman et al., 2024, p. 12), but when analysed in a more general sense, Jisc does indicate concerns about long‑term financial sustainability. If TAs are meant to redirect subscription funds towards OA publishing, as outlined in the original white paper by Schimmer et al. (2015), one could argue that it is inaccurate to include previous APC spend in the calculation of cost‑neutrality.

Setting aside complications regarding determining cost‑neutrality, several consortia have indicated that they are indeed paying more for TAs. Bibsam, the Consortium of Swiss Academic Libraries and analysis from Finland all mention these agreements being costlier than subscriptions or that they have increased costs overall (Ala‑Kyyny, 2024; Olsson et al., 2020, p. 13; Swissuniversities, 2023, p. 10).

It is also important to question whether cost‑neutrality is in fact a positive outcome. Although not paying more is worthwhile to highlight, positioning cost‑neutrality as a beneficial outcome reinforces the idea that status quo pricing, both for subscriptions and APCs, is acceptable. Libraries have been objecting to high subscription prices for many years, especially within the context of ‘big deal’ agreements; therefore, simply maintaining the costs of those large packages is not necessarily a positive outcome in the long term.

Theme: paying for publishing rather than subscriptions

A desire to pay for OA publishing rather than subscriptions was explicitly expressed by 40% of the consortia. On its own, this goal does not represent a meaningful change to the publishing system, but it was generally discussed in the context of paying for publishing as a service, with emphasis on the service element as creating a change in the overall structure. As explained by Schimmer et al. in their paper on DEAL’s cost‑modeling tool, this is ‘a redirection of current financial flows away from the outdated structure and logic of subscriptions and towards models, complete with new processes and workflows, based on open access publishing services’ (2021, p. 6), which is more broad‑reaching than merely changing which element of the process is being paid for. Although paying for publishing was cited as a desired outcome, it was rarely mentioned as an element that was being specifically assessed to see whether it was being achieved.

Theme: transparency

A desire for more transparency was noted by more than half of the consortia in their stated goals and by three when discussing their evaluation of the agreements. Mentions of transparency were related to cost, contract details and the workflows or processes themselves, with cost being the most frequently highlighted aspect. Many referred to transparency generally, rather than in relation to specific elements. It is somewhat difficult to objectively evaluate whether a particular agreement has increased transparency unless there is a concrete marker, such as the elimination of a non‑disclosure clause or the sharing of previously undisclosed information. This may explain why it appeared less frequently in the evaluations compared with the goals.

Theme: research accessibility or visibility

Increasing the accessibility or visibility of research was included as a desired outcome of TAs by seven of the consortia, with five also noting it in relation to the evaluations they were undertaking. Many of the references were quite general, simply making statements regarding research being more openly available. Several of the consortia cited increases in visibility for their researchers specifically, such as MPDL stating a goal of amplifying ‘the research produced by the Max Planck Society and make it accessible to as many people as possible’ (MPDL, n.d., para. 5) and MALMAD indicating that OA publishing benefits researchers by increasing exposure to articles, higher citation rates and improved international rankings, and that it is promoting this increase via its TAs (Inter‑University Computation Center, 2024). Similarly, some consortia indicated increased visibility specifically for their researchers as a benefit they have seen from their TAs – for example, EISZ mentions that its TAs significantly increased the ‘visibility of Hungarian science’ (EISZ National Programme, n.d.), and Projekt DEAL indicates that its TAs have the ‘well‑known benefits of increased visibility and impact’ (DEAL Konsortium, 2020).

Theme: equity

Although TAs have been criticized for reinforcing inequitable structures within the scholarly publishing system (Farley et al., 2021), approximately 40% cited improvements in equity as a desired outcome. The word ‘equity’ was treated quite broadly, with references to universal access, equal access, global increases in access, providing maximum benefit to all or fairness coded as equity‑related elements. Most commonly, equity was mentioned in a general sense, such as ‘equitable scholarly publishing,’ without specifying which aspect of the agreement aimed to ensure equity. SANLiC (South Africa) was the only consortium that specifically included equity as it relates to social justice, stating that ‘inclusivity and social justice must be core’ (Truran, 2021). From an evaluation perspective, equity did not emerge as one of the themes, which may be due to the inherent difficulty of quantitively or systematically evaluating concepts such as equity or social justice.

Theme: workflows and administrative work associated with TAs

Discussions of workflows and the administrative work associated with TAs were found both in the goals and in the evaluations undertaken, though references appeared more frequently in the latter. This was a varied theme, with many comments focusing on the potential reductions in administrative burden on researchers or the inefficiencies of paying individual APCs, rather than the potential additional administrative work associated with the TAs themselves – such as managing approvals for papers, workflows for author identification or answering questions from researchers about the details of the agreement. There were also several mentions of improving the workflows for paper submissions and related processing procedures. As recommended by Bakker et al. (2024), ‘an institution determining the potential benefit of an agreement must consider the indirect costs of the agreement’ (p. 86), although references to indirect costs, such as staff time, did not appear frequently in the sample.

Theme: publishing patterns

Analysis of publishing patterns was raised by almost half of the consortia in their discussions evaluating the agreements. Changes to publishing patterns do not typically appear in the stated goals, as there is not a specific desired outcome. Rather, this could be monitored to determine whether changes are occurring and to identify any undesired consequences of these changes. The findings were also mixed regarding whether publishing patterns are in fact changing, as can be seen in the discussion on author publishing choices in the literature review above. It is also possible that shifts in publishing patterns emerge more slowly than other changes and thus may not be apparent in the initial years of an agreement.

Theme: author choice

The concept of author choice was included in both goals and evaluation by roughly a quarter of the consortia. Statements coded as author choice encompass those concerning allowing authors to publish where they want and providing authors with options, especially those that are seen as easy or low‑friction. As an example, the ‘background and objectives page’ of the Projekt DEAL website outlines that TAs ensure that the ‘journals wanted by authors are preserved’ (DEAL Konsortium, 2020, para. 4). Similarly, SANLiC indicates that one of the benefits of TAs is that ‘researchers do not have to find a new open access venue to comply with open science policy. These agreements enable researchers from participating institutions to publish in the hybrid journal of their choice’ (SANLiC, n.d., para. 11). Olsson et al. (2020), in their assessment of Bibsam’s TAs, indicated that these agreements allow ‘researchers to continue publishing without changing their behaviour. They publish in the journals they are familiar with, whose publishing process they trust and whose reputation they consider beneficial to their career’ (p. 29). This element of choice is similar to the approach taken by the 2024 Swiss National OA strategy, which includes author satisfaction with publishing options as a measurable indicator in its evaluation process (Swissuniversities & Swiss National Science Foundation, 2024, p. 21). The need to provide authors with publishing choices, especially those that require little change to behavior, is inherently tied to the challenges of journal prestige and the promotion and tenure system, as outlined in the literature review above. As discussed in the earlier section of this paper addressing the failure of TAs to disrupt the scholarly publishing oligopoly, the preservation of author choice is a key factor cited by consortia and libraries in explaining continued participation in these agreements with large commercial publishers.

Theme: speed of the transition to OA

One of the benefits of TAs is the speed at which they can increase the number of OA papers. This rapid transition is possible partially because TAs do not require authors to change their behavior – they can continue to publish in traditional journals and can do so without paying an APC, which some may be unable or unwilling to do. Whereas broader efforts to advance OA through a change in research assessment practices will take time, the impacts of TAs can be seen more immediately. In the sample consortia, almost half mentioned increasing the speed of the transition to OA as one of the goals of entering into TAs, often using qualifiers such as ‘immediate’ when discussing the move to OA. For example, FinElib explicitly states that it negotiates TAs to ‘move rapidly towards open publishing’ (FinElib, n.d., para. 9).

Future considerations

Although some evidence indicates it may take 70 years or more for all publications to become OA via TAs (Brayman et al., 2024, p. 14), it is still important to consider what the future might look like if these agreements eventually achieve what they purportedly set out to do: pay for publishing rather than reading. What does a publishing‑only agreement mean for the role of the library, and for the work of the librarians responsible for these agreements? Previously, publishing problems, such as difficulties with the paper submission process, were not the business of the library. However, if publishing is the main element of an agreement, then is the library not also responsible for helping researchers with this, in the same way it would help users with access problems? There could be a future where libraries are essentially negotiating for publishing services only – this is quite different from current contracts, which are still mainly for access to content. As the role of the library continues to change, the Association of European Research Libraries (LIBER) suggests that by managing OA publishing agreements and engaging stakeholders in the process, libraries will position themselves as ‘a key and trusted partner for the future and secure the library’s role as the central point for managing research communication services within their institutions’ (Open Access Working Group, n.d.). In an environment where publishing is the main element of agreements, libraries may also have an increased role to play in providing guidance on research integrity, publishing ethics and best practices in scholarly communication. While libraries certainly possess knowledge and competencies that would be useful for this work, such as experience in license negotiations and expertise in scholarly communications, it is worth examining whether the library is truly the most suitable part of the institution to assume this responsibility.

In addition, changes to price distribution between libraries may also be required, as previous spending on subscriptions no longer makes sense if contracts are largely paying for publishing. The Bibsam Consortium has grappled with this challenge, noting in its report on the topic that ‘if the core provision of a contract is publishing and the price mechanism depends on number of publications, the cost distribution model ideally does as well’ (Vooren, 2019, p. 5). Campbell et al. (2022) also raised this issue, stating that ‘library consortia have an important role in designing new cost distribution models to help smooth the transition to open access for their library members, making gradual adjustments to their investments over time and make future agreements sustainable for all’ (p. 10). This potential redistribution of pricing would change the evaluations of these agreements and further call into question the concept of cost‑neutrality. It would also require a more concrete differentiation between cost‑neutrality at the library or institutional level and cost‑neutrality at the consortial level. As explained by Earney (2017), ‘cost reallocation would not just have to happen between institutions in one country, but across many countries’ (p. 16). The prospect of rebalancing costs globally presents considerable complexities. Paying for publishing – especially in contracts where payment is based on publication intensity and where adjustments may need to be made between countries – raises questions about the objectives of such agreements and whether the library paying for publishing is truly the most desirable long‑term outcome.