Introduction

In spite of the indisputable economic, social and ethical case for granting open access (OA) to our research findings1,2,3,4,5,6,7, the journey towards OA feels painfully slow8,9. Indeed, the practical challenges of moving towards OA are great enough to make the route so hard-going as often to seem impassable. For this reason advocacy remains essential, but we believe a different kind of discussion is needed: one that can engage with all sides of the debate, can understand and draw together all perspectives, and can move things forward in ways that we can all agree are the most sensible and practical.

The authors of this article are most concerned, for obvious reasons, with the implications of OA for funders like The Higher Education Funding Council for England (HEFCE) and the other three UK HE funding bodies, with whom HEFCE has developed its forthcoming open access policy. It is from this perspective of research funder that we write this article. Funders cannot act alone, though, and we must understand the issues from all sides before we implement any policy that could have profound effects on the research community. This is why HEFCE and the other funders undertook a double consultation on open access in relation to the research assessment exercise that follows the current REF10.

HEFCE provides £1.6 billion in funds to English higher education institutions for conducting research11. HEFCE funds research as a public good: the creation and dissemination of knowledge and insights has the power to transform our world for the better. For this reason, it is our long-standing position that the research we fund should reach the widest possible audience to maximize its chances of having a positive economic, social and cultural impact12,13.

Challenges

It should be clear, then, that the benefits of open access publication of research as enabled by the new technologies of the internet are of considerable interest to HEFCE and other research funders. However, for OA publication to truly take hold, a number of important practical questions will need to be answered.

“…we must understand the issues from all sides before we implement any policy …”

Who should pay?

The most obvious of these questions is the financial one. Though some would believe that the internet will ultimately enable the free propagation of all information, as well as its free use14, we are clearly not at that point yet. Indeed, that idealistic aim may actually hinder the more rapid adoption of open access options15. The process of scholarly publication undoubtedly costs money, and for information to be free to the reader, publishers (whether commercial operators, learned societies or other not-for-profit organizations) must receive income or labour from another source. Arguments have gone on for years over whether, on the one hand, publishers' subscription income should be protected by delaying free access to the information or, on the other hand, their subscriptions should be replaced with new income, for example, from article processing charges (APCs) payable by the author or author's institution or funder16. From our perspective, there are benefits and drawbacks to both of these. To delay free access to vital publicly funded medical research by substantial lengths of time seems unethical today, given the potential of the internet to facilitate free and immediate dissemination of knowledge to those in need of it. To expect all researchers, including independent scholars of modest means, to find the funds to publish their work seems equally problematic.

“… the costs of publication … are an inseparable part of the public cost of research.”

This is, of course, an oversimplification of a more complex debate, but these perspectives are frequently enough encountered to be highly illustrative of the spiny nature of the financial question. We are grateful to all those seeking to raise the standard of debate around open access, not least the British Academy whose important research into the economics of journal publishing in the humanities and social sciences formed important inputs into HEFCE's policy17.

Though we respect the fact that complex financial questions around the various mechanisms for delivering OA are not yet fully resolved, we believe research funders are entitled to see these questions as secondary to a bigger issue. We are of the view that the costs of publication, whether these are borne by libraries or by authors, are an inseparable part of the public cost of research. Institutions in receipt of HEFCE funding, by extension, are entitled to use that funding for subscriptions, for APCs, or for both. The mechanism by which the publication of research is paid for is therefore less relevant to us than the extent to which access to the research is maximized. As such, public funding for research publication must take place for the wider benefit of the economy and society.

Retaining author choice

A separate issue for moving towards OA is around its effects on the choices of individual academics. If funders place conditions on their funding related to publication, there will be instances where these conditions come into conflict with individuals' publication choices. Accepting research grants or public research funding brings with it a responsibility for researchers to follow reasonable funder requirements but we accept the argument that individual researchers are best placed to judge the most appropriate outlet for their research. For this reason, we have protected individual author choice as much as possible in our open access policy18.

“… individual researchers are best placed to judge the most appropriate outlet for their research.”

The role of the publisher

There is no doubt that moves towards OA present significant challenges to scholarly publishers of all kinds. We believe that publishers can and will rise to these challenges and that the widespread adoption of OA requires innovative forms of dissemination to be offered by both existing and new publishers.

Some see the publisher's role in the dissemination of information as little more than a vendor19. This viewpoint transmutes the many positive contributions that publishers make to the information dissemination process20 into negatives: careful quality control becomes cherry-picking of saleable material, maintenance of a publication's trusted reputation for excellence becomes profiteering from an imprimatur built by others21, and copy-editing, laying out, marketing and (for some22) peer review become unnecessary gold-plating that ought to be dispensed with.23

We believe that the existing publishers have the potential to offer much that will support open access to scholarly information, and that negative viewpoints like the one above hinder rather than aid widespread adoption of OA. We do, though, look to publishers for responses which go beyond their current position. Despite a lack of evidence that the subscription model is undermined by shorter embargoes24,25,26, there has been considerable publisher resistance to short embargo periods27,28,29,30,31. Publishers have also been ready to accept additional fees to ‘unlock’ articles and make them freely available32,33 but there is a significant risk that this will increase costs to universities – a risk that they are well aware of34,35,36,37,38. We realize that there are complex issues for publishers in managing the unlocking of articles39,40 but we believe this issue is the single biggest hurdle to the adoption of a sustainable open-access model for publishers, but at least one publisher has already overcome this hurdle41. We need a more cost-effective way of unlocking material if publishers, universities and funders are to work together towards sustainable open access.

“… existing publishers have the potential to offer much that will support OA …”

Tackling this problem leads into a more rounded view of the value added by quality control and aggregation. For publishers of all kinds to thrive in a more open world, they face a challenge in re-articulating their purpose and benefits to funders, researchers and institutions in a positive and engaged way. We believe that publishers can continue to play an important role in (as Michael Bhaksar has stated) filtering, framing and amplifying information and knowledge42. But these roles will continue to be challenged by technological advances that seek to automate43, and our desire is that publishers engage with funders as partners in the objective of achieving open access, rather than be seen as deferring or even obstructing widespread adoption.

The HEFCE policy for open access

HEFCE and the other three UK HE funding bodies want to stimulate the academic community simply to do what is already possible: to make their work available, online, for free access and use by anyone with an internet connection. We believe we can achieve this through a simple and straightforward set of rules for the research assessment exercise that follows the current REF: if an output is not available in an open access form, then it is ineligible for submission to REF 202044. With some exceptions45, which we will respect, we have discovered as a result of extensive consultation that such a policy can be readily implemented, with publisher support, in a way which supports academic choice of publication destination and without any significant increase in the cost of scholarly dissemination.

Key to achieving this aim is indeed to respect the scholarly publishing landscape as it exists today. We have chosen to begin by focusing only on those types of output for which OA has reached a certain maturity; namely, journal articles and conference proceedings. It is already standard practice for many journals and conferences to allow authors to self-archive a copy of their paper on their institutional repository or website; indeed, we note that the final peer-reviewed drafts of over 80% of published articles could have been archived under existing journal policies, had the authors undertaken to do so46, and this figure rises to 96% when looking at a sample of articles submitted to the 2014 REF47. Our policy requires that the author upload a copy of their peer-reviewed paper to an institutional or subject repository at the point of acceptance for publication. We want this to become a matter of routine for authors to ensure their eligibility for future research assessments. We have included appropriate protections for those exceptional cases where deposit was not possible. We will support repositories to get ready for this requirement when it comes into force in 2016 – key to this will be for repositories to actively respect any embargo periods specified by a publisher. These protections are a key part of our strategy to work within current constraints, and we will be seeking to keep an eye on whether these protections are adequate (or indeed necessary) as the global landscape evolves.

“… our desire is that publishers engage with funders as partners …”

It is not clear to HEFCE that the monograph publishing world is as ready as the journal world to move to open access publishing. There are oft-cited issues around monograph publishing – rising prices, dwindling print runs, squeezed library budgets48,49 – to which OA seems like an obvious and tempting solution50. However, we do not believe it is right to impose OA rules on monographs now. We do not fully understand the issues that the scholarly book publishing world is facing, nor do we fully understand the extent to which the humanities and social science communities that rely so heavily on book publishing are ready (or willing) to embrace electronic and OA forms of book publication, especially given the pre-eminence of the print monograph within these disciplines51,52. These issues and others need much further exploration. The work that HEFCE is doing, in partnership with the Arts & Humanities Research Council (AHRC) and Economic and Social Research Council (ESRC), will be key to helping us move forward our understanding in this area, and we are very grateful to Professor Geoffrey Crossick for agreeing to lead on this work53.

“…, we do not believe it is right to impose OA rules on monographs now.”

How OA publishing will work for non-text outputs, including for creative and practice-based research, is even less clear to us at the moment. Respondents to the funding bodies' consultation on open access have shown us that there is significant desire within the arts world to move towards greater openness of its research works, but have revealed concern that this desire is not being met by current publication frameworks54. There are also similar issues around the data that underpin much scientific research – respondents tell us they want to make their data available openly, but we are not yet fully sure of the extent to which this is practical right now. We have therefore decided not to subject all of these to our OA rules, but will be working hard to increase our understanding of the issues over the coming months.

The last part of the funding bodies' policy we want to mention here relates to the undue complexity that readers face in trying to find research outputs relevant to their work. The increasing volume of available scholarly material means that traditional searching and indexing methods may no longer continue to be sufficient, and one of the clearest arguments for OA is that it allows automated searching and indexing by third parties to take place55,56,57,58. These types of sorting and retrieval technologies will help researchers and others to cut through ‘information overload’ and find the information that is most relevant and useful to them.

We fully recognize both the current benefits and the tremendous potential for new and emerging technologies to enhance the way we interact with research. We want to reflect the fact that we recognize this, and have done so in our policy by giving credit to institutions that can demonstrate they have taken steps to enabling the so-called text mining of their research outputs by automated tools as a key part of creating the healthy research environment of the future.

“… events are fast-moving, and new developments come thick and fast.”

In summary

While we believe our HEFCE policy goes a long way towards bringing open access to UK research within our grasp, we will need to be watchful as events unfold. There is no doubt that in recent years, OA has attained new prominence in UK academic life, not least since the review of OA by Dame Janet Finch. But events are fast-moving, and new developments come thick and fast. For HEFCE, this means we must ensure our policy remains attuned to the changing environment, while setting clear and lasting rules for researchers and institutions and remaining mindful of the needs of all stakeholders.

With this in mind, the final perspective we want to draw into this discussion is that of the reader. When we consider readership, it is not obvious that electronic publication, let alone OA publication, will necessarily entirely replace other forms of publication59,60. Though universal low-cost internet access is looking like an increasingly achievable goal, there are still parts of the world for which routine electronic access to journals or books is simply not yet practical. There are still others, closer to home, who judge electronic access to be inferior to print61, and we need only look outside the academy to see how traditional print sales continue to dominate non-scholarly work for a signal of how resistant readers are to transitional forces62. These issues must be considered alongside all others if open access is to become a reality.

Many of the perspectives we have shared above are, without a doubt, concerned with the practical nature of the journey towards open access. HEFCE will continue to play its part in solving the practical challenges of increasing access to the research we fund by whatever means we have at our disposal. Overall, we want to make sure that whatever the shape of things to come in the world of scholarly publishing, HEFCE takes a strong, evidence-based and realistic view and acts accordingly.