Should open access be removed from REF requirements?

In the past week, three senior research strategy figures at the University of Oxford have called for removing the open access ‘burden’ from the rules for the next Research Excellence Framework (REF). For the past REF excercise, open access has been a requirement for all submitted journal articles and UKRI are also now consulting on plans to include books within the rules for the next exercise. The Oxford authors make a number of points that the REF open access rules are bureaucratic, not commensurate with value and work against the broader adoption of open research practices through a narrow focus on open access. This is a significant statement from a UK university and represents a more combative attitude towards open access than we have seen in recent years.

In a subsequent article, Research Information presented a series of responses from influential figures within research culture and open research, all of whom were highly critical of Oxford’s intervention. Sally Rumsey, former head of scholarly communications at Oxford, described it as ‘short-sighted’ while Dorothy Bishop emeritus professor at Oxford considered it a ‘“a retrograde step to question benefits of open access”. Stephen Curry of the Research on Research Institute said: “If Oxford was looking to burnish its reputation as a defender of the status quo with this blogpost, then it has succeeded admirably.”

I think it’s helpful to have a conversation about whether the REF OA rules are still fit for purpose. To be sure, I’m not against OA in any way whatsoever, but I have been consistently critical of the burden of OA compliance. With reference to interviews with librarians and policymakers, my PhD thesis explored how the UK open access policy landscape created a culture of compliance that libraries were meant to orchestrate, in some cases forcing them to redirect staff from library publishing towards the REF policy. This burden was in addition to the fact that mandating open access via an increasingly bureaucratic set of regulations is not a particularly good way of getting researchers on board with broader cultures of openness. We still see a huge amount of apathy or antagonism to open access because of this.

Much has changed since my research was conducted and compliance is now firmly embedded within the UK library sector. Many open access library teams are organised according to a function of getting papers into repositories, ensuring deposit regulations are followed and advising researchers on what they should do to comply. This requires significant expertise in funder policy within a service that seeks to deposit as many of an institution’s papers as possible. This service operates in parallel with a more general approach to open access that seeks to shift subscriptions over to agreements with publishers that offer access to read content and the ability to publish bundled into one cost.

You might think that, with my general dislike for compliance-based approaches, I’m in favour of the Oxford authors’ proposals to remove the OA component from the REF. While I think it’s a good conversation to be having, I would caution against removing the requirement entirely. For journal articles, the UK university sector has taken great strides towards open access policies based on rights retention. These policies set the default for research to be made open access through repositories and allow authors to opt out if they would prefer to keep it closed. The UK university sector is technically and infrastructurally well prepared to make all journal articles OA in both green and gold forms, while also seeking to nurture an ecosystem of community-led and innovative models for no-fee OA. This perhaps could mean that a more relaxed requirement is needed, one that does not require all individual papers to be checked for compliance but assumes a general position of good faith that universities will make their research openly accessible through rights retention policies. This means that I would be interested in a discussion around relaxing the OA requirement for journals rather than removing it entirely. Any money saved on compliance could then be reinvested in experimentation with open research practices and long-form scholarship, something the Oxford authors tacitly support although fail to make a strong case for.

This is a delicate argument because the UKRI’s approach to OA has simultaneously enabled more OA through compliance but also more capacity for ethical forms of OA through its block grant that can be spent on a broader range of infrastructures and staff time in service of OA in the UK. Compliance is very much a feature of UK OA, but so are initiatives like the COPIM project, Open Library of Humanities, and a host of new OA university presses. My argument stems from the fact that we should continue to support these kinds of projects while paying less attention to compliance-based approaches without removing the general need for institutions to make all their research OA. The UK is in a very good position to adopt this stance.

So I would have liked the Oxford authors to have made more of an argument around the benefits of open research and the ways to stimulate open practices, which they say are held back by the REF policy, as a way of working through compliance towards more ethical and pluralistic publishing futures. Without this argument, you end up opening the door to more conservative voices that seek to return to the halcyon days in which public access was not a concern, as is the case with Rick Anderson’s piece in the Scholarly Kitchen blog. Here, Anderson argues that we should remove the ‘ideological orthodoxy’ of open access requirements and instead allow researchers to ‘exercise their own judgment’ around open access, in an ironic appeal to an Americanised ideological orthodoxy of individual freedom. Fortunately for us over the pond, we have a chance to care about and nurture the broader collective benefits of open access.