Introduction

Research libraries have a long tradition of funding the market for academic publishing. Library budgets pay for the acquisition of monographs, standing orders for series and for subscriptions to journals and databases. Whereas purchases by individuals might admittedly be an additional stream of income (especially in the context of society publishing), it is fair to state that library budgets were, and still are, one of two essential elements in the traditional business model of scholarly publishing behind a paywall – the other one being the (for the most part) free labour performed by scholars in the production and reviewing of manuscripts. It comes as no surprise then that these same libraries are called upon to act as funders for publishing in open access (OA). After all, does it really matter whether libraries pay for the acquisition of the finished product published behind a paywall or for the production costs of the same publications in OA? If a large part of libraries’ budgets is used already to finance the market for academic publishing, should they not simply adapt when that market changes? And if it is true that there is enough money in the system to simply switch to OA publishing entirely,1 then what is stopping us from investing the whole of the acquisition budget of research libraries, as is, into OA publishing?

Unfortunately, things are not that easy. Firstly, there is the fear of freeloaders. The return on investment in the old model is obvious: if you spend money, you add to your collection and only those who have access to your collection profit. But what if a library pays for the production costs for OA so that everyone has access? We would hope that not too many librarians consider that to be a problem, as it is the essence of our profession to make information accessible to as many as possible. But colleagues in the financial administration of academic institutions might think otherwise: if access to content could be maintained with somebody else picking up the bill, then it is very tempting to reconsider spending money on it. Secondly, the conversion to OA does not happen overnight, so we are faced with a budget need for publications behind a paywall as well as to invest in OA publishing. Thirdly, there are many indications that a switch from buying the finished product to paying for its production will actually result in a further rise of costs if we try to achieve this through so-called transformative agreements with the currently predominant publishers.2 It is true that, ‘the idea that open access could be part of the solution of the affordability issue is more or less abandoned’.3 Even the LIBER principles for negotiations with publishers in the context of offsetting deals seem to admit this, as the statement ‘No Open Access, No Price Increase’ implies that a price increase is acceptable if the contract contains an OA part.4 Lastly, unless we make the change from financing product to financing production for all publishers, all research outputs and all academic disciplines simultaneously, we will end up enforcing the oligopoly which already exists in academic publishing, as a lot of ‘smaller players’ will not be able to make the transition as quickly and successfully. (^{Note 1}) We could come to a (very costly) new deal for mainstream scholarly publications with a handful of big technology companies, achieving maximum OA for these, but we will drive smaller presses out of the market and kill bibliodiversity in the process. (^{Note 2})

We should consider why we strive for OA. If it is only to arrange broader access to the published record, then the easiest approach undoubtedly is to continue working with the traditional partners in scholarly communication and simply accept the rise in costs. But what about the other hopes and expectations for OA? What about the goal to open up participation in scholarship (^{Note 3}), thus overcoming the exclusivity enjoyed by researchers affiliated with the richest universities in the world – which implies that you cannot trade a barrier to reading for a barrier to publishing (as happens in approaches to OA which make use of author-facing charges)? What about the desire to regain control, as an academic community, of the ecosystem of scholarly communication (^{Note 4}) – which will not happen if authors continue to give up copyright in exchange for access to this ecosystem? And what about the wish to control the cost of scholarly communication (^{Note 5}) – which we abandon if we are willing to pay whatever it takes to be able to publish in OA?

We believe that research libraries should continue to act as funders of the ecosystem of scholarly communication, playing their part as ‘public agents of learned patronage’5 by transitioning from spending on product to spending on production cost. However, we are also convinced that we should not miss this opportunity to cure the schizophrenia showcased by librarians, who understand perfectly well that ‘the aims of commercial … publishers are not the same as theirs’.6

KU Leuven Fund for Fair Open Access

At many institutions, this reasoning was the impetus behind establishing mission-driven university presses, frequently organized as library publishing programmes (especially if these are recently founded). At KU Leuven, this step was taken a long time ago, with the establishment of Leuven University Press (LUP) as a separate entity in 1971.7 In line with its mission, LUP embraced OA early on and is recognized, for example, by the European Research Council (ERC), as a trusted OA publisher. Together with the encouragement of green OA through the institutional repository Lirias8 (supported by a policy that KU Leuven authors should deposit a copy of their work and by Belgian copyright law, which enables the dissemination of pre-final versions9), LUP has thus, for some years already, played a prominent role in the promotion of OA. However, KU Leuven Libraries wanted to do more, which led to the foundation of the KU Leuven Fund for Fair OA in 2018.10 This distinct fund, established as a budget line in the financial structure of the library separate from the acquisition budget for academic publications behind a paywall, focuses exclusively on non-profit and community-led (^{Note 6}) approaches to OA publishing. The current fund is set up in two parts, one consisting of subsidies for OA books published by LUP (€100,000 in 2021) and the other one (€60,000 in 2021) to support a wide array of open scholarship initiatives such as financial contributions to diamond OA (^{Note 7}) programmes. As such, it is a centralized fund to finance publishing in OA as found at many other universities. However, its strict focus sets it apart from budgets that are (also) used for profit-driven approaches to OA and explains why the fund’s size is relatively small (representing little more than 1% of the entire operating budget of KU Leuven Libraries). The fund is overseen by a committee consisting of representatives from various academic disciplines, LUP and KU Leuven Libraries, and chaired by KU Leuven’s vice-rector for research policy. The day-to-day management is in the hands of two members of the library’s staff (namely the authors of this article), who take up this responsibility alongside their other tasks. Together, they spend approximately the equivalent of 0.3 FTE on the care and promotion of the fund and can count on administrative support (such as processing payments or updating the website) from other library staff members (about 0.1 FTE).

There are several advantages to creating a distinct fund of this sort. Firstly, it provides extra visibility, even with a relatively small budget, both for the institution (which can highlight its support for non-profit and community-led initiatives) and for the publishers and providers supported by the fund (who get promoted on the fund’s website11 and were highlighted in a social media campaign during the summer of 2021 via the KU Leuven Open Science Twitter account). Secondly, creating a separate fund is a way of securing part of the library budget for alternatives to for-profit OA, which is a necessity since this approach to OA creates (to use Jeff Pooley’s phrase) a ‘read-and-publish funding lockdown’,12 thus threatening to eat up all the available funds and driving non-profit alternatives out of the market.13 A distinct fund, in essence, constitutes the ‘dedicated budget line for open infrastructure’ that Peter Kraker argued for recently and is a way to set aside a small percentage of the library budget for investment in open infrastructure as suggested by David W. Lewis.14 Lastly, while the initial set-up of a detached budget line might be a bit cumbersome, it actually accommodates decision-making and administrative processes further down the line. Many libraries struggle to join library membership or sponsorship programmes in the field of open scholarship – such as those run by Open Book Publishers, punctum books or the Open Library of Humanities, or those co-ordinated by SPARC (SCOSS), LYRASIS (OACIP) or COPIM (Opening the Future) – sometimes not even because the funds are not available to do so (as the required investment is typically quite limited), but because the decision-making for novel types of acquisitions is unclear and processing payments for such programmes can be complicated. Having a discrete fund dedicated to support exactly such (and only such) endeavours implies that the decision to sign up for any of these programmes, and the related administration, becomes a lot easier.

However, there is also a downside to creating an autonomous fund, which is small in comparison to the entire library budget, namely that supporting the non-profit and community-led approach to OA can be perceived to be a marginal affair. In that sense, it could become an excuse not to engage with it on a broader scale. In Leuven, we have tried to avoid this pitfall by promoting all of our investments in community-led, non-profit open scholarship under the same umbrella (thus securing the first advantage listed above), without funding all of these solely through the Fair OA Fund. As a result, support for, for example, The Programming Historian, OpenEdition Journals and the Open Commons of Phenomenology is listed on the central webpage about open scholarship15 but actually funded by the collection budget of discipline-specific units of KU Leuven Libraries, such as Artes and the Institute of Philosophy. This way, we wanted to create awareness that there is a need to repurpose the acquisition budget to fund OA publishing (on the condition that we focus on the non-profit approach) and wanted to achieve a situation where library staff who traditionally focus on purchases start considering alternatives, in an effort to avoid that investing in OA only happens when additional budgets are created, which are handled by a separate team of OA specialists.

Proving return on investment: readership data for OA books (^{Note 8})

The largest share of the available budget is spent on providing funding for OA books published by Leuven University Press. (^{Note 9}) Depending on the affiliation and additional means (such as project funding) of authors, the fund subsidizes a third, two thirds or the whole of the publication costs. Since the launch in 2018, the fund has so far (writing in July 2021) helped to finance 23 OA books, with another 30 subsidy applications already approved. An additional three book projects were initially cleared for subsidy but ultimately not pursued as originally intended. Applications for financial support from the Fair OA Fund are open to all authors wishing to publish with LUP. The opportunity to get an at least partial intervention in the author fees is clearly being noticed: whereas 13 proposals were approved in 2018 and again in 2019, this number rose to 20 in 2020, and 10 applications were approved in the first half of 2021. The applications also highlight LUP’s international reputation: 24 out of the 56 applications submitted so far came from authors not affiliated with KU Leuven.

The reach of the books subsidized by the fund is demonstrated by the readership data. Fully aware that view or download statistics are not a sure measure of actual readership and might lead to unwanted use (such as comparing the impact of individual titles), we decided to share them anyway to underline the added value of publishing in OA venues. What is more, we wished to share readership data as proof of return on investment, affirming that the fund is offering a real contribution to the impact of authors (both from within and outside KU Leuven) who publish in OA with LUP, as well as highlighting the success of LUP as an international OA publisher.

We also thought carefully about how we wished to present this data, conscious of the fact that this is not just a simple number of total online readers. In the end, we decided to use an approach inspired by one of the presses which we support through the Fair OA Fund, namely Open Book Publishers.16 This approach takes into account one of the main challenges, namely that the same OA book is typically made available in a multitude of ways. Books subsidized by the KU Leuven Fund for Fair OA are distributed via OAPEN, JSTOR and Project Muse (and can, of course, also be shared on any other platforms the author or readers may wish to use). This is obviously desirable for the optimization of global impact, dissemination and digital preservation but complicates the matter of readership data as there is a lack of standardization in how to gather and present usage metrics. (^{Note 10}) Our methodology gathers usage statistics from OAPEN, JSTOR and Project Muse and breaks them up per platform instead of providing only the number of total readers. We also provide an interactive map of downloads and user engagement for each book, intended to reflect the geographical reach of the publication. Both the detailed figures and the interactive map are updated quarterly and made available on the website.

Numerous studies have pointed out that publishing in OA increases impact17 and our numbers only provide further confirmation of this phenomenon. If we look at the data on a title level, we see that the impact of a new title is almost instant (whereas this would take considerably longer with a book that appears only in print), with an average of more than 300 chapter readers and almost 100 book readers per title every month during the first six months after publication. The very first book published with the fund’s support, namely Brokers of Modernity: East Central Europe and the Rise of Modernist Architects, 1910-1950 by Martin Kohlrausch (published in March 2019), even reached 451 readers (both as individual chapters and the book as a whole) on average per month during the first 12 months after publication, with 751 consultations during the first month of publication. These numbers exceeded the expectations of the author:

‘There were mainly two reasons why I decided to go for Open Access. First, I expected that with Open Access my book could reach a much broader audience than a print edition, as its theme, “modernity”, is global, also a global audience. Second, I was hoping for a quick “absorption” of the book. Judging from the very high download numbers my expectations have been outmatched. Moreover, these numbers provide me as the author with quite telling insights into where the book is downloaded (and hopefully read).’ (^{Note 11})

The readership data not only provides us with insight into how often these OA books are consulted but also where. An oft-cited advantage of OA is that the content can be accessed all over the world, by everyone who would be interested in reading it and has access to the internet. We are happy to see that titles published with the support of the fund have been accessed by readers from all inhabited continents – which is particularly impressive when you realize that some of these titles were published in Dutch. With one download from Vatican City, we even reached the smallest sovereign state in the world. What is more, data for books like Oil Wealth and Development in Uganda and Beyond: Prospects, Opportunities, and Challenges, edited by Arnim Langer, Ukoha Ukiwo and Pamela Mbabazi (published in 2020), also indicate that OA actually helps to better match the geographical focus of the book with interested readers than was possible within the traditional publishing model, where readers need to have the financial ability to buy a physical copy of the book. This collection of essays studies the management of oil resources in Uganda, with chapters drawing on the experiences of Nigeria, Ghana and Kenya. Despite being published in Belgium, it has been accessed more in Africa than anywhere else.

No more APCs

The second part of the fund, currently devoted completely to library memberships in the field of open scholarship and diamond OA programmes, was initially also set up, in part, to finance article processing charges (APCs) for articles published in journals that adhered to the Fair OA principles as defined by the Fair Open Access Alliance. (^{Note 12}) The rationale behind this was that we also wanted to support (mostly) scholar-led OA journals depending on an APC-based business model. This limitation to Fair APCs made our fund stand out from other institutional APC funds, as is illustrated by a survey from 2019 conducted by Springer Nature in which OA administrators from 16 institutions indicated that they often exclude hybrid OA but do not differentiate between for-profit and non-profit OA. (^{Note 13})

The decision not to establish an APC fund that would cover any and all author fees of KU Leuven researchers was well-considered, in the first place, simply because this is financially not feasible without a massive budget increase for a library at a research-intensive university. (^{Note 14}) More importantly, however, this would have implied that the university actively promoted this business model and encouraged its researchers to publish in APC-funded OA journals. But this is not the policy we wished to accept since we (together with many OA advocates and non-profit publishers) consider any model with author-facing publication costs counterproductive to a transition to affordable, sustainable and equitable OA.18

The application process we had originally set up for financing non-profit APCs was very simple: authors could send in their submission via an online form and would receive feedback after no more than three working days. From the outset, every journal was evaluated on a case-by-case basis by the two members of the library’s staff responsible for the day-to-day management of the fund. While this was time-consuming, we consciously did not want to provide a white list of journals that qualified for support, mainly because we do not believe that the landscape of scholarly publishing with all its complexities and nuances can be translated into a set of clear-cut criteria for approval or disapproval. Moreover, no list can keep up with the fast-evolving academic publishing market, with new journals appearing on an almost daily basis. Inevitably, the conclusion is that a white list will always be subjective (which is also why we decided against working with one of the commercial lists currently available) and has to be permanently maintained, making it even more time-consuming than our case-by-case analysis.19

Our main goal for establishing this very limited APC fund was to encourage non-profit forms of scholarly communication and to increase awareness amongst researchers about how different OA publishing options affect the scholarly communication landscape. We found it crucial to foster this awareness, as many researchers are not informed about the financial impact of their publishing choices – for which we cannot blame them, as the institutional subscription model shielded them from this in the past. During its three years of operation, the APC fund has achieved this goal as it provided a new incentive for conversations about for-profit vs. non-profit academic publishing at the university. However, we are not naïve. While our approach indeed instigated discussions and created some awareness, we are fully aware of the fact that the unavailability of a library subsidy for for-profit APCs will not prevent authors from publishing in such journals, just as the existence of a small fund exclusively focused on non-profit APCs will not change the business model of for-profit companies.

Multiple factors have led us to discontinue the APC part of the Fair OA Fund, as of 2021. First of all, while we put a lot of effort into providing clear communication that this was not a library-sponsored APC fund that would pay for all (for- and non-) profit OA costs, it did create the misconception in the minds of some researchers that the university would now pay all APCs on their behalf. They also fixated on the maximum price of the APC we had set in order to determine if the APC, in fact, represented the true cost of publishing (€1,000) (^{Note 15}) and requested reductions from publishers, hoping we would then approve their application. Obviously, exceptionally reducing the cost of an APC to publish in a certain journal does not all of a sudden make it a Fair OA journal.

Clarity was the second reason why we suspended the APC part. Whereas in theory it was straightforward to state that we would only sponsor APCs of articles published in Fair OA journals, in practice we were quickly operating in a grey zone between for-profit and non-profit publishers. If we strictly applied the first Fair OA principle (‘controlled by the scholarly community’), we would hardly be able to approve any applications. We therefore decided, in the beginning, to be more lenient. However, after evaluation we concluded that to stay true to the principles of the fund it was essential that we avoided operating in this grey zone. This way, we could also equivocally state that we do not support the APC model, as we, as said before, actually do not believe that a business model for OA centered on author-facing costs is the way forward.

The third reason for deciding to cancel the APC part of the fund was that we soon learned that managing even a small-scale APC fund comes at a large administrative cost. (^{Note 16}) It, for instance, became clear that processing payments would require more than the very limited staff investment of (in total) 0.4 FTE that we had originally envisioned. Publishers have their own workflows in place to manage APC payments which do not always encourage the library to act as an intermediary in the publisher-author relationship, resulting in publication delays and a time-consuming bureaucratic nightmare for both author and institution. We were, for example, faced with the same payment issues as reported in 2014 by the University of Glasgow Library, having to request new invoices with the institution’s rather than the author’s name. Furthermore, we were likewise confronted with constant payment reminders as our financial system is simply not set up for immediate payments, leading to a seemingly endless loop of e-mail conversations with both author and publisher assuring both parties that payments had been made.20

All of this led us to conclude that instead of investing energy and money in handling APCs, we wanted to prioritize the development of an alternative OA ecosystem which holds more promise for equity and sustainability. Since 2021, we therefore focus the second part of the fund exclusively on creating partnerships with diamond OA platforms and other types of open scholarship infrastructure by contributing to their economic viability. A thorough analysis of what to support is still time-consuming but less so than judging individual journals. Additionally, administrative complications concerning payment are minimal, not only when dealing with established partners such as SPARC or LYRASIS, but also with other non-profit publishers with experience with library membership programmes such as Open Book Publishers and the Open Library of Humanities.

Conclusion

The current push for OA risks putting so much strain on library budgets that there is no money left to foster alternative approaches to scholarly communication beyond the for-profit solutions offered by a handful of big technology companies. This should not deter research libraries from investing in OA, but rather incentivize them to prioritize non-profit, community-led approaches which are better aligned with scholarly values. One possible way to do this is to create a discrete fund specifically focused on financing these alternatives, which has a number of advantages concerning visibility, speed of decision-taking, administration and budget control. The Fund for Fair OA managed by KU Leuven Libraries is an example of such a fund, solely spent on OA books published by a university press and other non-profit initiatives in the field of open scholarship (particularly when funded through library membership programmes). Our experiences with managing a Fair APC Fund confirmed our opinion that sponsoring alternative business models will result in a more equitable and scalable implementation of OA than what can be achieved through the prevailing model with author-facing charges on the level of individual publications. KU Leuven invests in OA in other ways as well, but the Fund for Fair OA puts our preferred approach to OA publishing in the spotlight and makes sure that we safekeep part of the library budget to maintain diversity in the market of scholarly publishing.