Skip to content
The UK's only independent think tank devoted to higher education.

Open Access: A Benefit Not a Burden That is Worth the Cost

  • 20 June 2024
  • By Stephen Curry, Dorothy Bishop and Martin Paul Eve
  • This HEPI blog was kindly authored by Stephen Curry, Professor of Structural Biology and Consul at Imperial College London and Director of Strategy at the Research on Research Institute; Dorothy Bishop, Professor Emeritus of Developmental Neuropsychology at the University of Oxford; and Martin Paul Eve, Professor of Literature, Technology and Publishing at Birkbeck, University of London.
  • The article below responds to this piece published by HEPI last week, which called for the open access requirement to be excluded from the Research Excellence Framework.

Sometimes it seems as though the debates on open access (OA) move in endless circles. In fact they are slowly spiraling towards their conclusion, just as OA policies are moving, albeit circuitously, towards full implementation. The policy trajectory has not been straight because so many technical, commercial and cultural arguments have had to be overcome to forge a way ahead.

How long it will take to complete the journey remains to be seen. There are a few more obstacles to clear out of the way first. The argument presented recently on this blog by representatives of Oxford University decrying the linkage of OA to the REF (which is now proposed to apply to long-form publications such as monographs and book chapters) touches on some valid questions about implementation, but we do not find their arguments convincing.

Patrick Grant, Oxford’s Pro-Vice-Chancellor for Research, Tanita Casci, Director of the Research Strategy & Policy Unit, and Stephen Conway, Executive Director of Research Services, claim that the proposed policy, on which Research England and the other HE funders in charge of the REF have recently consulted with the sector, is “unaffordable and excessively bureaucratic”. They assert that “The OA compliance cost for the new requirements for long-form outputs has been estimated at £20M over a REF cycle for our university alone” and “This estimate does not include the cost of researchers and professional services teams engaging with a convoluted, technical policy.”

Grant and colleagues go on to suggest that the expanded OA demands of the REF will prevent institutions from selecting their very best outputs for submission to the 2029 assessment exercise, but remain silent on the question of why Oxford scholars would not want their very best work to reach the widest audience to give it the largest possible impact within and beyond the academy. They round out their argument with the claim that a policy that clearly signals the importance to the UK of open research “detracts from opportunities to promote open research practices” by fostering a compliance culture. While this may be a risk, it is disappointing that senior figures at one of our leading universities are not more alive to the responsibilities and opportunities that attend the realisation of a more inclusive OA policy.

Perhaps most discouraging is the question raised by Grant et al. of why the REF should have an OA policy at all. This echoes objections raised a decade ago when OA for the REF was first mooted but ignores the fact that the REF awards public money. Denying any section of the public access to the outputs of that work is asking for funding without accountability.

Their question also downplays the significant public good that the mandate has achieved, including the transformative effect of the REF OA requirements implemented in 2016, which produced a huge boost in the accessibility of UK research. Indeed, open access has many benefits for many parties. It allows research to reach the widest audience, particularly among those who lack university journal subscriptions or library access. It facilitates the detection of fraudulent research practices. It ensures global participation in the research endeavour. And it helps people with disabilities who cannot access physical academic libraries with any ease. We create a fairer, more equal, society by removing barriers to access for everyone. 

The proposed enlargement of the REF mandate to include long-form humanities work (books) will extend these benefits to these outputs. Yet the outcry over this requirement remains, despite suggestions for how it could be equitably funded. We would only note that it is a strange pose for humanities scholars continually to decry the lack of public funding for their disciplines and yet spend so much effort resisting moves to make their research accessible to the public that provides that funding.  This is particularly pertinent at a time when humanities departments are under threat. If outputs from subjects such as English, history and classics are invisible to the wider world, it makes it far easier to dismiss them as irrelevant. The REF OA policy helps make the case for these disciplines, embodying Gibbons’ 1999 vision of a “transparent and participative” social contract between academics and society.

Researchers are, of course, right to resist the imposition of unnecessary bureaucracy and Research England should listen to reasoned suggestions for streamlining the process. The numerous exemptions and caveats to the policy, though, are evidence that they have done and continue to do precisely that.

Yet we would ask on our turn what Oxford and other institutions have done proactively to prepare the way for a policy that was trailed back in 2016, or during the years since as the rising tide of OA has only surged higher? Oxford’s estimate of £20m costs should be considered in the context of the estimated £1,200m that Oxford is likely to receive over a seven-year REF cycle*. Is it so outrageous, when awarding over a billion pounds of public funding, to insist that a mere 1.7% of it is spent to ensure the public can see the results? Further, if truly seeking to contain costs, they might implement robust and responsible research assessment exercises that would free scholars from the demands of ‘prestige publishing’ and the price hikes of prestige publishers in the first place. They could also consider investing QR funds to support scholar-led publishing operations, such as the Open Library of the Humanities, University College London’s open access press, or Cambridge University’s Diamond OA journals platform. We must all of course continue to engage in the technical, commercial and cultural arguments thrown up by the REF and OA, but let us do so with a clearer view of all the issues at stake.

*This calculation is based on the published QR allocation of £166m to Oxford University for the year 2023-24. It makes no allowance for inflation or changes in government funding.

Get our updates via email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.


  1. Joshua says:

    Not having a REF OA Policy makes complete sense, what this article fails to acknowledge is that the previous article and other arguments against REF having an OA policy are not against OA at all – indeed they argue for Open Research to be assessed in the People, Culture & Environment REF element. If you apply the costings not only to what you finally submit, but to all monographs published during a REF period the costs are far greater – and not all institutes receive the same QR funding as Oxford does – but publish the same number of monographs – or do you only fund those by staff whose work you believe will be at least 3 star, but not those at 2 star or below – creating a class system… Those advocating to not have a REF OA policy, are not against OA at all. But whether research is open or indeed is open, but took more than a month to be deposited in a repository (e.g., staff was on FTE contract of 0.4 and heavily involved in marking when article was accepted for publication, so deposited within 4 months, but publisher policies only allow open after 6 months – so made no difference to when others could access – but not eligible for REF… that has to be wrong). Let HEI’s determine their own open access policies, publication policies, workflows for openness and assess that as part of the environment – looking not only at compliance, but culture, open software, open educational resources etc. Without the need for another convoluted policy that wastes a great deal of admin time and money and that results in some excellent research not being assessed.

  2. Li Chend says:

    It’s telling that OA advocates are using the very argument that publishers used – that publishing costs are a small fraction of the overall research cost to try to counter the points of the Oxford scholars.

    However, with OA there are now multiple extra layers of complexity and bureaucracy to pay for – that people are finally pushing back on.

  3. John says:

    The authors are right to acknowledge that it is the public that provides the bulk of the funding to cover OA costs, but do they have any evidence that this is how the taxpayer wishes its money spent? – writing big cheques to publishers over funding the NHS, schools or providing a national defence?

Leave a Reply

Your email address will not be published. Required fields are marked *