REF 2029 Open access consultation and engagement summary

The consultation on the proposed open access policy for REF 2029 aimed to gather insights and perspectives from the higher education sector and other stakeholders on key issues related to open access requirements for research outputs. As part of the consultation process the funding bodies published a summary of proposals for open access in REF 2029, inviting responses from interested and affected parties.

The consultation consisted of 19 questions, including both categorical and open-ended questions, allowing for detailed feedback on various aspects of the proposed policy.

A total of 279 responses were received and analysed, including those submitted through the survey system and a small number of additional responses received via email. The analysis focused on identifying key themes and messages emerging from the narrative responses to the open questions, supported by quantitative data from the categorical responses.

The consultation questions covered a wide range of topics, including:

• changes in the open access landscape since REF 2021
• requirements for journal articles and conference proceedings
• requirements for longform outputs (monographs, book chapters, edited collections)
• implementation timelines
• exceptions to the open access policy

Sector engagement

In developing the policy proposals for the consultation, the REF team obtained input from the sector through meetings and attendance at events with relevant organisations, bodies, groups and individuals.

Further engagements included:

• Russell Group Librarians group
• Universities UK PVCR network
• small specialist institutions
• open access publishers and university presses
• the British Academy Learned Societies and Subject Associations Network
• CoAlitionS

The REF team worked with Universities UK to deliver a webinar open to all interested parties, with over 250 registrations from across the sector. This included broad representation from institutions, academic sector bodies, publishers and funders.

The REF team also received supplementary written submissions from:

• the British Academy
• the Association of Learned and Professional Society Publishers
• International Association of Scientific, Medical, Technical Publishers (STM)
• the UK Committee on Research Integrity (UKCORI)

The feedback gathered from these engagement events and communications, representing a wide range of interests across the sector, was captured and fed into reporting and decision making arising from the consultation. This feedback is incorporated into the relevant question responses, given below.

Analysis

The analysis of the feedback received highlights areas of consensus, concerns, and suggestions for further consideration. The findings have informed the development of the final REF 2029 open access policy, ensuring it is both robust and supportive of the evolving open access landscape. The feedback will also be used in developing revised proposals for the longform open access policy to be in place from January 2029.  

The consultation consisted of three questions (questions 1 to 3) capturing demographic information to enable a closer understanding of respondent makeup and views. Sixteen questions focussed on general and more specific issues related to open access and proposals for REF 2029. These covered key landscape changes in recent years, open access in the REF and proposed changes to submission requirements for journal articles and longform outputs.

Questions 1, 2 and 3 (responding as Individual or Organisation, Organisational affiliation, Country) concerned respondent characteristics used for demographic analysis. Most submissions were from institutions/organisations (74%), with 26% being individual responses.

Among both respondent types, the largest group was affiliated with universities/HEIs (70% of total). This included 86% of individual and 62% of organisational respondents. The next most significant respondent categories were learned societies (11% overall) and sector organisations (8% overall). Sector organisations encompassed a diverse group, including mission groups, umbrella bodies, disciplinary societies, collective groups, and service/infrastructure organisations.

Geographically, most submissions came from England (67%), followed by “Other” (18%), which primarily included learned societies and sector organisations with pan-UK roles, as well as publishers and those in the publishing sector with international interests. Scotland and Wales accounted for 8% and 6% of responses respectively, with Northern Ireland contributing less than 1%. Notably, nearly 45% of Welsh and 35% of English responses were from individuals, compared to 18% of Scottish responses, with none from Northern Ireland or “Other”.

Analysis of specific feedback received for each consultation question, providing a more granular understanding of the diverse perspectives and concerns expressed by respondents is set out below. As noted above, feedback from sector engagements is incorporated into the themed responses to consultation questions.

General feedback

While the questions were focussed on particular aspects of the policy proposals, respondents expressed views on the proposed open access changes more broadly, with many expressing some concerns (including where supportive of aims) or opposition, much of this relating to the incorporation of longform outputs within the proposals.

While a smaller proportion expressed support for the proposals, for some this came with reservations.

Key concerns and criticisms

• Longform outputs: key concerns were around mandated open access for longform publications, with concerns over lack of funding to support open access publishing
• Maturity of longform open access:  while there is some positive support for inclusion of longform in future, current levels of maturity of open access practice and publication options were considered challenging
• Practical barriers: concerns about potential misuse of research (noting AI as an area of potential concern) and need for clearer guidance for researchers and institutions
• Perverse incentives: fears that proposals as they stand could create perverse incentives, undermining the REF’s credibility and effectiveness, and misaligning with open access goals
• Negative impacts: worries about disproportionate impact on AHSS disciplines, smaller publishers, and less well-funded institutions
• Compliance-based approach: some respondents considered that a compliance-focused approach, as proposed, could lead to negative consequences, particularly for AHSS and smaller publishers. Some suggestions that focus on HEI engagement with open access might be more effective

Support for open access and alternative proposals

• General support: acknowledgement of the positive principles behind open access, even among those opposed to specific proposals
• Nuanced perspectives: some supportive respondents suggested the proposed approach might not be suitable for all authors, disciplines, or publication types, and highlighted the need to address specific concerns related to trade and particularly crossover books
• Alternative proposals: suggestions included assessing institutional engagement, aligning with UKRI policy

While there is general support for open access principles, concerns about implementation, potential negative impacts, and considerations of alternative approaches suggest a need for further refinement of the policy, particularly in respect of longform outputs.

Question 4 – Changes to the research landscape following the REF 2021 open access policy

Respondents identified several key changes in the open access landscape, including:

• Growth of open access options: the increase in reputable diamond open access publishing models, particularly for journals.
• Greater visibility and funding: increased political and funder support for open access, changes in publisher practices and improved funding for open access, particularly in STEM.
• Transformative agreements and Jisc support: transformative agreements and support from Jisc were seen as positive, however challenges were also acknowledged.
• Challenges in AHSS: concerns about the lack of adequate funding for open access in AHSS, potentially limiting scope and quality of REF submissions.
• Rights retention: increase in rights retention policies across HEIs was noted positively, but with challenges highlighted for implementing and embedding these policies, especially for smaller institutions.
• Publisher conservatism: respondents observed a reluctance among publishers, particularly in longform publishing, to adapt to the changing open access landscape.

Question 5 – Deposit requirements post-acceptance

The consultation question on whether to maintain deposit requirements post-acceptance for publications not immediately open access elicited a range of views, with a negative skew, but no clear consensus emerging (47% disagreement and 27% agreement).

Key themes and concerns

• Acceptance versus publication date: respondents were divided on whether the deposit deadline should be tied to the acceptance date or the publication date. Some preferred the publication date, arguing that it aligns better with the actual availability of the research. Others favoured retaining the acceptance date, citing clarity and ease of tracking.
• Burden of tracking and compliance: some respondents questioned the necessity of tracking deposit information, suggesting that it creates an unnecessary administrative burden. Some proposed making data collection non-audited.
• Impact on academic engagement: concerns that shifting the deposit deadline to publication could negatively impact academic engagement with the open access process. The acceptance date is seen as a trigger for action; respondents worried that urgency might be lost with the proposed change.
• Proposed one-month time limit: the suggested one-month deposit time limit post-publication was met with concern. Respondents considered this insufficient, citing potential delays in retrieving and processing outputs, staff availability, and limitations of repository systems. Alternative timeframes, ranging from three months post-acceptance to 12 months post-publication, were proposed.
• Clarity on publication date: respondents requested a clear definition of “publication date,” as different interpretations (for example, AAM added by the publisher, date first available online, issue publication date) could lead to confusion and compliance issues.
• Compatibility with conference papers: a few respondents questioned whether the proposed policy aligns with the publication cycle of conference papers, which often involves different timelines and processes compared to journal articles.

While simplification and reduced administrative burden are seen to be desirable, there are concerns about timely availability of research and maintaining academic engagement with the open access process.

Question 6 – Alignment with UKRI open access policy on licensing

The proposal to align the REF 2029 open access policy with the UKRI policy on licensing for journal publications, requiring CC-BY or CC-BY-ND licenses, elicited a mixed response (43% disagreement, 29% agreement).

Key themes and concerns

• Misuse and misappropriation: concerns about potential for misuse and misappropriation of research under proposed licensing, particularly in context of AI and text/data mining. Some highlighted a need for guidance and safeguards to protect intellectual property.
• Overreach and perverse incentives: some considered the proposed alignment as overreach by the REF, creating perverse incentives for researchers and publishers. They argued it could undermine credibility and effectiveness of the REF by prioritising open access compliance over other research quality indicators.
• Rights retention and institutional capacity: while acknowledging the positive trend of increased rights retention across HEIs, respondents noted that this is not universal. Smaller institutions face challenges in implementing and enforcing rights retention policies due to limited resources and legal expertise.
• Impact on creative works and commercialisation: concerns were raised about the removal of the NC (non-commercial) license, particularly for publications in the arts and humanities. This could lead to commercial exploitation of creative works without proper attribution or compensation. Limiting the use of ND (no derivatives) licenses might limit creators’ ability to reuse their own materials.
• Compatibility with specific disciplines and journals: respondents highlighted that for disciplines, such as medicine, many gold open access journals require NC licensing. The proposed changes could render outputs ineligible for publication in these venues, limiting researchers’ options and impacting the quality of REF submissions.
• Disruption, timing and publisher response: concerns potential disruption from implementing these changes midway through a REF cycle. Respondents emphasised the need to understand how publishers would respond to changed licensing requirements and whether they will adapt policies and practices accordingly.
• International collaboration: potential negative impact on international collaborations and publishing was raised, with some questioning the rationale for removing the NC safeguard and its potential implications for global research partnerships.
• AI and erosion of standards: the potential for AI-driven misuse and erosion of research standards were recurring concerns. Respondents highlighted uncertainties surrounding AI technologies, highlighting that this may not be the right time to adopt more strictly permissive licensing terms.

While a significant proportion opposed the proposed alignment, some noted general support but with reservations about the removal of NC with some also supporting SA (share-alike) licenses. They argued that these offer important safeguards, particularly in certain disciplines where commercial reuse or misattribution is a concern. Some respondents questioned the necessity of strict alignment with UKRI policy, suggesting REF can adopt a more nuanced approach recognising disciplinary differences and needs in the research community. There is a general favourability for increased openness and accessibility, there are also concerns about potential misuse, perverse incentives, and impacts on specific disciplines and publishers.

Question 7 – Recognition of alternative platforms

The proposal to recognise alternative platforms as meeting open access requirements for primary publication received a mixed response, with a significant majority in agreement (46% agreement, 18% disagreement).

Key themes and concerns

• Peer review and quality concerns: respondents expressed some reservations about the quality of research published on alternative platforms, citing lack of rigorous peer review processes on some.
• Disciplinary differences: the suitability of alternative platforms was seen to vary across disciplines. While pre-prints and similar platforms are common in STEM fields, they are less prevalent in AHSS disciplines. However, some AHSS subjects, like linguistics, do utilise such platforms.
• Increased burden: the potential for increased burden on individuals, institutions, and the REF team was highlighted. Respondents anticipated challenges in evaluating and ensuring quality and compliance of publications on diverse alternative platforms.
• Need for clear guidance: respondents stressed the importance of guidance on acceptable alternative platforms, evaluation criteria, and alignment with UKRI policies. This would need to be regularly updated to accommodate the evolving landscape of alternative publishing models.

Responses indicate a growing recognition of the role of alternative platforms in scholarly communication. While there is some support for their inclusion in the REF, concerns about quality assurance, disciplinary differences, and increased burden need to be addressed.

Question 8 – Proposed changes to embargo periods

The proposal to reduce embargo periods for journal publications elicited a mixed response, with a clear majority expressing concerns (51% disagreement, 23% agreement).

Key themes and concerns

• Publisher adaptation and international context: a major concern was that proposed changes do not reflect publisher practices and there is uncertainty about whether journals, especially international ones, will adapt embargo periods to align with the REF. Some considered that REF’s influence may be limited in the global publishing landscape.
• Challenges with rights retention: while acknowledging the importance of rights retention in facilitating open access, respondents expressed concerns about the time and resources needed for wider adoption, particularly for smaller institutions. This questioned the feasibility of implementing these changes within the proposed timeframe.
• Impact on smaller publishers: respondents highlighted potential negative impacts of reduced embargo periods on the viability of smaller publishers, particularly those operating with tight margins.
• Disciplinary variations: concerns about the differential impacts of reduced embargo periods across disciplines. AHSS disciplines, particularly arts and humanities, are at greatest risk, with negative consequences for quality and diversity of REF submissions.
• Increased burden on HEIs: respondents anticipated a significant increase in the administrative burden to monitor and ensure compliance with changed embargo periods.
• Concerns over mid-cycle implementation: concerns about implementation of changes part-way through the REF cycle, with potential for disruption and confusion.
• Need for research on publisher viability: calls for research to be commissioned on the impact of reduced embargo periods on publisher viability, particularly for learned societies that rely heavily on publication income.
• Need for publisher engagement and further evidence: some respondents emphasised the importance of engaging with publishers to understand their perspectives and potential responses to the proposed changes. They also called for further research to assess the potential impacts of reduced embargo periods on different stakeholders.

Positive impacts and alternative views

• Increased access and impact: some respondents highlighted positive impacts of shorter embargo periods, emphasising increased accessibility and potential for greater impact.
• No embargo periods: a small number of respondents advocated for eliminating embargo periods altogether, arguing for immediate open access upon publication.
• Past experience with open access mandates: a few respondents pointed out that previous open access mandates, including REF 2021, have not led to significant disruptions in the publishing industry, and that concerns about the impact of reduced embargo periods might be overstated.

While there is support for increased openness and reduced barriers to access, there were concerns about the potential impact on publishers, disciplinary differences, and administrative burden on HEIs.

Question 9 – Implementation date for journal-based publications

Overview of responses

The proposed implementation date of 1 January 2025, for the revised open access policy for journal-based publications had considerable opposition. The majority of respondents (62%) disagreed with the proposed implementation date, while only 14% were in favour.

Key themes and concerns

• Burden of multiple policies: respondents highlighted the challenges of managing three concurrent open access policies (REF 2021, REF 2029 for journals, and REF 2029 for longform outputs) with differing requirements and timelines.
• Insufficient time for implementation: many respondents felt that 2025 was too soon to implement the changes, emphasising the need for a longer lead-in period to adapt. Concerns were raised about inadvertent non-compliance due to publication timelines, and insufficient time for HEIs to adapt internal processes, and CRIS providers to update systems.
• Challenges for smaller HEIs: the limited time for smaller HEIs to develop and implement rights retention policies was a particular concern, given the potential advantages of rights retention in meeting open access requirements.
• Impact on green open access and journal licensing: respondents noted that journals might need time to update licensing agreements and institutions relying on green open access might be disadvantaged by a rapid implementation.

Suggestions for a revised timeframe

• Alignment with longform outputs: a common suggestion was to align the implementation date with that for longform outputs (January 2026) to streamline policy management and communication efforts.
• Minimum twelve-month delay: many respondents advocated for a minimum twelve-month delay from the announcement of the final policy to its implementation, allowing time to address concerns and prepare for the changes.
• Delay to next REF cycle: some suggested delaying implementation until the next REF cycle to avoid the complexities of managing multiple policies within one cycle.

Additional requests and suggestions

• Clarity on the application of REF 2021 criteria until the new policy’s implementation
• Definitions of key terms like “submitted,” “accepted,” and “published” in the context of the new policy
• A single, consolidated REF open access policy
• Reversion to a three-month deposit window post-acceptance, as the proposed one-month window post-publication was deemed too short
• Development of a tool to clearly identify journals that comply with REF requirements

The majority of respondents opposed the proposed January 2025 implementation, citing concerns about the feasibility, potential for non-compliance, and need for more time to adapt. Feedback suggests a strong preference for delaying implementation, aligning it with the longform output policy, or even deferring it to the next REF cycle. Additionally, respondents called for guidance and tools to support compliance with the new policy when it is implemented.

Question 10 – Tolerance limit for articles and conference proceedings

The proposal to retain the 5% tolerance limit for non-compliant articles and conference proceedings received a mixed response, with the largest group of respondents in favour (43% agreement, 29% disagreement).

Key themes and concerns

• Disproportionate impact on smaller submissions: the most common concern, raised by both supporters and opponents of the 5% limit, was its potential disproportionate impact on smaller submissions.
• Stricter OA rules: several HEIs, sector bodies, and learned societies highlighted the proposed OA rules for REF 2029 are stricter than in REF 2021 due to shorter embargo periods and licensing requirements. Several called for an increase in the tolerance limit.
• Clarity and guidance: there were requests for clearer guidance on the open access policy, including audit requirements, application across output types, and handling of pre/post-January 2025 publications.
• Administrative burden on HEIs: several respondents, particularly from smaller institutions, expressed concerns about the administrative burden associated with managing open access compliance with the granular requirements of the proposed policy. Increasing the tolerance limit was seen as a way to reduce this burden by minimising the need for exceptions.

While the proposal to retain the 5% tolerance limit received the most support, there was no consensus. Concerns about its impact on smaller submissions, stricter OA rules in REF 2029, and administrative burden on HEIs were prominent. There were calls for clearer guidance and potential reconsideration of the tolerance limit, especially for units with fewer submissions.

Question 11 – Proposed exceptions for journal publications

The proposed exceptions for journal publications were generally well-received, with a clear majority expressing agreement (57% agreement, 15% disagreement).

Key themes and concerns

• Removal of embargo exception: a notable point of contention was the exception allowing for embargo periods exceeding the stated maximum. Several respondents, including some HEIs, suggested removing this exception, arguing that it could undermine efforts to encourage shorter embargo periods and incentivise authors to seek open access options.
• Need for clarity and guidance: respondents sought further clarification on several aspects, including the criteria for determining the “most appropriate publication venue,” the definition of a “substantive connection” between an output and an HEI, and the handling of conference proceedings published with an ISBN. They also requested more information on audit triggers and thresholds, and recording and evidencing compliance.
• Additional exceptions: a range of additional exceptions were proposed, reflecting the diverse needs and circumstances of the research community. These included exceptions for cases where the publication venue cannot offer compliant licensing, where original content is generated under more restrictive licenses, and where administrative errors lead to non-compliance. Respondents also suggested reinstating some exceptions from REF 2021, such as those related to delays in securing the final peer-reviewed text and portability of outputs between institutions.

While the proposed exceptions were generally viewed favourably, feedback highlighted areas where further clarification and refinement are needed.

Longform outputs

Question 12 – Deposit requirement for longform publications

The proposal for longform publications, mandating immediate availability as open access upon publication (or within 24 months if subject to an embargo) faced significant opposition (49% disagreement, 22% agreement).

Key themes and concerns

• Disruption and cost implications: respondents expressed concerns about the feasibility of implementing this change within the current REF cycle, highlighting potential disruption and financial burden on HEIs and academics. The short timeframe and need to manage multiple open access policies simultaneously were cited as major challenges.
• Applicability to longform outputs: several respondents questioned the appropriateness of applying open access mandates to longform publications, particularly in the arts and humanities. They argued that open access is more relevant to journal publications and STEM disciplines, and that mandating it for longform outputs could negatively impact scholarship in certain fields.
• Potential negative impacts on research and scholars: concerns were raised about potential unintended consequences, such as limiting quality and volume of longform publications submitted to the REF, creating a skewed picture of the research landscape, disproportionately affecting early career researchers, those in less well-resourced fields, and those at less well-funded institutions.
• Challenges in the longform publishing market: respondents highlighted the differences between the journal and longform publishing markets, emphasising that the latter is not yet as mature in terms of open access options and sustainable funding models. They argued that the proposed changes could lead to increased costs for authors and institutions and might not be feasible given the current state of the market.
• Concerns about the Author Accepted Manuscript (AAM): respondents raised concerns about the status of the AAM for longform publications, highlighting that it is often a more provisional and less polished version compared to journal articles. They argued that mandating sharing of AAMs could negatively impact the perception of research quality and discourage authors from participating in open access.

Support and nuances

Despite the opposition, some respondents expressed support for the proposal, citing its potential to decolonise knowledge and challenge the dominance of larger publishers. Others suggested that parity with journal publication requirements would be beneficial. However, even those in favour acknowledged the need for wider adoption of rights retention and a potentially later implementation date to ensure the policy’s workability.

While there is support for the underlying principle, responses highlighted the need to further consider the challenges of the longform publishing market, disciplinary differences, and potential impact on researchers and institutions. Feedback suggests that further consideration is needed to ensure that any policy is both feasible and supportive of a diverse and thriving research landscape.

Question 13 – Twenty-four-month embargo period for longform publications

The proposal for a maximum 24-month embargo period for longform publications faced significant opposition, with a majority expressing concerns (52% disagreement, 18% agreement).

Key themes and concerns

• Inadequate embargo period: for many a 24-month embargo was too short for longform publications, particularly considering the need for publishers to recoup investments. Respondents proposed extending the embargo to 36 or 60 months, with some suggesting no embargo at all.
• Curtailment of academic freedom and publisher disincentives: concerns were raised that the proposed embargo, coupled with other open access requirements, could limit authors’ freedom in choosing publishers and disincentivise publishers.
• Negative impacts on scholarship and publishing: respondents feared potential negative implications for high-quality publications and disciplines reliant on longform outputs, particularly in the UK context. They also expressed concerns about the viability of smaller, specialist, and scholarly presses if implemented.
• Need for publisher engagement and consideration of international context: the importance of engaging with publishers to understand the implications of the proposed embargo period was emphasised. Respondents also highlighted the international dimension of publishing and questioned the REF’s influence on global practices, especially when mandates are unfunded.
• Impact on authors and research: concerns were raised about potential negative effects on author royalties and the disincentivisation of longform publishing, particularly in the arts and humanities where publication costs may be borne by the authors. The proposal was seen as an unfair demand to surrender intellectual property without compensation, especially given the limited funding available in certain disciplines.

Support and nuances

A small number supported a 24-month or shorter embargo, aligning with UKRI policy. Some saw this as a positive step towards decolonising knowledge and challenging the dominance of larger publishers. However, even those in favour highlighted the need for wider rights retention and a potentially later implementation date.

The proposal faced significant opposition, with concerns raised about its impact on publishers, academic freedom, research quality, and the financial burden on authors and institutions. Feedback suggests a need for further consideration of the embargo period, greater engagement with publishers, and a nuanced approach that recognises the unique characteristics of the longform publishing landscape and diverse needs of different disciplines.

Question 14 – Licensing for third-party materials

The proposal to grant an exemption from open access requirements when licensing for third-party materials cannot be obtained received overwhelming support (83% agreement, 8% disagreement).

Key themes and concerns

• Lack of control and funding: respondents emphasised that authors and HEIs often lack control over third-party licensing decisions and may face challenges in securing permissions, especially due to limited funding or copyright holders’ concerns beyond financial compensation.
• Disproportionate impact on specific disciplines: the requirement to obtain third-party licenses was seen as disproportionately affecting certain disciplines, such as Art History, which heavily rely on such materials.
• Potential for negative behaviours: respondents expressed concerns that the absence of an exemption could lead to increased licensing fees or even legal aggression from third parties in cases of inadvertent copyright infringement.

Suggestions

• Exemption for unrealistic costs: respondents suggested extending the exemption to cases where licenses are obtainable but at a prohibitively high cost.
• Need for clear guidance: the need for clear and timely communication of audit and evidentiary guidance for data collectors was emphasised.

Respondents highlighted the lack of author and institutional control, funding limitations, and potential negative impacts on specific disciplines. The feedback suggests that the proposed exemption would be crucial to ensure fairness and avoid unintended consequences in open access policy relating to longform outputs.

Question 15 – Output sharing without third-party licensed materials

The proposal to share redacted versions of outputs when licensing for third-party materials cannot be obtained received mixed feedback (29% agreement, 41% disagreement).

Key themes and concerns

• Impact on output meaning: many respondents believed that removing third-party materials could render outputs meaningless, particularly in disciplines heavily reliant on such materials.
• Negative perception of green deposit: the majority of comments expressed concerns about depositing redacted outputs in institutional repositories (“green deposit”).
• Loss of author control: concerns were raised about the potential erosion of author control over their work, as redacted versions could be widely disseminated and perceived as the definitive version.
• Burden of redaction: the process of identifying and redacting third-party materials, especially in longform outputs, was seen as a significant burden.

Potential negative consequences for research integrity, author control, and the practical challenges of redaction were noted as serious concerns.

Question 16 – Principle of tolerance for non-compliance
and question 17 – Tolerance level for non-compliant longform outputs

While the majority of respondents (66% agreement, 14% disagreement) supported the concept of a tolerance level for non-compliant longform outputs, most disagreed with the proposed tolerance level of 10% (57% disagreement, 12% agreement). A significant portion advocated for 100% tolerance or a non-penalty pilot exercise.

Key themes and concerns

• Level of tolerance: a recurring theme was that any implemented tolerance level should be generous, potentially higher than the proposed 10%. This stems from concerns about the feasibility of achieving 100% compliance at this stage and potential negative consequences of a strict approach. Concerns were raised about the potential double weighting of non-compliant outputs and the lack of resources (funding and staff capacity) to achieve 100% open access.
• Resource constraints: respondents highlighted lack of funding and staffing capacity as significant barriers to achieving compliance, particularly for longform outputs.
• Impact on research and equity: concerns were noted that a tolerance level, even a generous one, might discourage the submission of high-quality research due to the fear of non-compliance. Additionally, concerns were raised about the potential impact on equity, diversity, and inclusion (EDI) and the parity of access to resources.
• Discipline-specific concerns: respondents highlighted that the negative impacts of a tolerance level would disproportionately affect certain disciplines, particularly those within the arts and humanities.
• Mandatory open access: a notable minority questioned the necessity or appropriateness of mandating open access for longform outputs in any form.

While the principle of a tolerance level for non-compliant longform outputs enjoys majority support, there is a call for it to be more generous than that proposed. Concerns about potential negative impacts on research submission, EDI, and specific disciplines suggest a need for careful consideration and mitigation strategies.

Question 18 – Proposed implementation date for open access policy

The majority of respondents disagreed (55% disagreement, 21% agreement) with the proposed implementation date of 1 January 2026. A significant portion considered that longform outputs should not be included in any open access mandate. There was a strong preference for delaying implementation or adopting a fundamentally different approach.

Key themes and concerns

• Disruption and timeframe: implementing the policy midway through the current REF cycle was deemed too disruptive. Respondents suggested delaying the implementation until the next REF cycle to allow for thorough consultation, development of a more supportable approach, and preparation time for all stakeholders.
• Alternative approaches: respondents proposed abandoning the open access requirement for longform outputs altogether or shifting to a non-compliance-based policy that assesses institutional commitment and progress toward open access. This could be assessed through the People, Culture, and Environment submission and include appropriate metrics.
• Cost and burden: concerns were raised about the high costs associated with the proposed implementation and the increased burden it would place on researchers, particularly early career researchers and those unable to afford publication fees.
• Publisher agreements and flexibility: respondents emphasised the need for clear policies and agreements with publishers. Concerns were also raised about the potential restriction of publisher choices and the impact on research flexibility.
• Impact on research and researchers: there are concerns that the proposed implementation could discourage researchers from pursuing longform publications and negatively impact specific groups, such as early career researchers.

There is a preference for either delaying the implementation, adopting alternative approaches, or abandoning the open access requirement for longform outputs altogether. Concerns about disruption, cost, burden, and potential negative impacts on research and researchers suggest the need for further consideration and consultation for any implementation of open access policies for longform outputs.

Question 19 – Proposed exceptions for longform publications

While most respondents agreed with the proposed exceptions for longform publications (46% agreement, 23% disagreement) there were concerns and suggestions for improvement. Respondents sought clarity and guidance on the proposed exceptions and their implementation.

Key themes and concerns

• Clarification and rewording: respondents requested clarification on several points, particularly regarding Exception 1, suggesting rewording “the only” to “the most” appropriate publisher to mirror requirements for journals and clarify how this will be determined and evidenced.  
• Cost implications: concerns were raised about the potential cost implications of Exception 1, as it could imply an obligation to choose a compliant option even if it is expensive.
• Audit and evidence: respondents requested further details on audit triggers, thresholds, and the information HEIs should collect for audits. They also sought clarification on the evidence required to demonstrate ‘liaison and consideration’ for Exception 1.
• Additional exceptions: respondents suggested several additional exceptions, including an exception for publishers not permitting multiple chapters from the same edited collection to be deposited as Green open access, an equivalent to the ‘Other’ exception from REF 2021 for unforeseen circumstances, a formal exception for trade books, and an alignment with the UKRI open access policy for items not in scope.

While the proposed exceptions for longform publications received general support, respondents called for further clarification and guidance. Cost implications, clarifying audit requirements, and considering additional exceptions to ensure a fair and effective open access policy for longform outputs were issues raised.

Impact of consultation responses on the REF 2029 open access policy

The funding bodies made clear at the outset of the consultation their intention that the consultation should be meaningful, and that comment was being sought to contribute to and affect the final policy to be applied for REF 2029.

With feedback given on the range of issues covered by the consultation, we have sought to balance the feedback and extent and nature of any supporting information with other priorities and policy imperatives. In doing so we aim to have achieved an approach for REF 2029 that reflects a realistic and achievable policy for the sector and the funding bodies.

Some of the key changes to the REF 2029 policy, as compared to the proposals consulted on, are directly informed by the consultation responses and information from sector engagement events. This reflects the strength and weight of particular arguments, allowing us to better understand both feasibility and acceptability of particular measures at this stage. The arguments against some measures were supported by clear information providing a nuanced understanding of complexities and factors in play across the sector which supported early decision making on some aspects, communicated shortly after the end of the consultation itself.

There has been a further deeper consideration of the feedback, and further decisions taken on the basis of feedback and other factors, leading to changes, amendments and additions to the final policy.

These are outlined below.

Longform outputs

There will be no mandate for longform outputs in REF 2029. Longform outputs are not considered to be in-scope of the policy.

Journal articles and conference contributions

• Implementation: implementation date for requirement changes to the REF open access policy will be 1 January 2026 rather than 2025.
  • This will give a clear 12-month lead-in for implementing changes and adapting systems.
  • REF 2021 policy requirements will continue to apply for all in-scope outputs up to this date.
• Deposit: the policy requirement remains for deposit post publication (where not published fully open access).
  • The funding bodies are clear that earlier deposit is both accepted and supported. Where institutions have practices and systems in place for mandating deposit on acceptance there is no requirement to change this.
  • The post publication deposit requirement period is three months rather than one month as proposed.
• Licensing: publications should be licensed as openly as possible.
  • The funding bodies preference is for CC-BY or other licenses conferring the same level of openness.
  • However, licenses meeting the standards of CC-BY-NC and/or CC-BY-ND (including CC-BY-NC-ND) are acceptable.
• Embargo periods: the funding bodies are aware of and appreciate the concerns set out in the consultation responses; however, a decision has been taken to continue with the proposed reductions.
  • This is in the context of, and to encourage, moves across the sector towards rights retention and exploration of open access options for publication with short/no embargo periods, to support quicker access to research.
• Exceptions: some further exceptions have been applied and changes made to some proposed exceptions responding to information from the consultation.
  • Some proposed exceptions relating to concerns over licensing are addressed by the changes to the initial proposals.
  • Exceptions are set out in the policy, grouped by exception type.
• Audit: while audit guidance for REF 2029 is still being developed, responses to the consultation make clear that there was a need to ensure some early guidance was given on general approach and triggers for audit.
  • The audit approach for open access in REF 2029 will be broadly in line with that used for REF 2021.
  • An outline of the principles for audit are set out in the revised policy.

While many of the proposed measures have been agreed and incorporated into policy, some have not been taken forward for REF 2029. These include incorporation of longform outputs, moves towards more open licensing and aligning embargo periods to the UKRI open access policy (with the aim of improving openness across the research landscape, and support moves towards rights retention across institutions). However, these areas remain as aims for the funding bodies to progress going forward into future assessment exercises.

As previously communicated, work will be taken forward to develop an open access policy for longform outputs for implementation on the 1 January 2029, at the end of the REF 2029 publication period. This will be published well in advance of these changes to give researchers and institutions time to adapt.