What We Heard Report: Engagements on the review of the Tri-Agency Open Access Policy on Publications (2024)

---

1. Executive summary

On July 4, 2023, Canada’s federal granting agencies—the Canadian Institutes of Health Research (CIHR), the Natural Sciences and Engineering Research Council of Canada (NSERC), and the Social Sciences and Humanities Research Council of Canada (SSHRC) ("the agencies")—announced plans to revise the 2015 Tri-Agency Open Access Policy on Publications. In that announcement, the agencies committed to engage with a wide range of stakeholders to "identify the key features of an effective, comprehensive, sustainable and equitable immediate OA Policy for peer-reviewed articles, and the incentives and supports required for the Policy's successful implementation."

Engagement activities included an online survey; information sharing and engagement sessions with agency and tri-agency advisory structures (NSERC Leaders, SSHRC Leaders, CIHR University Delegates, etc.); thematic engagement sessions with representatives from the Canadian researcher, librarian and scholarly publishing communities; conversations with members of tri-agency Indigenous advisory structures; an engagement session with large, for-profit scholarly publishers; and engagements with organizations such as the Fonds de Recherche du Québec (FRQ) and the Canadian Research Knowledge Network (CRKN). Several partners such as Innovation, Science and Economic Development Canada (ISED), the Office of the Chief Science Advisor (OCSA) and the FRQ were involved in the organization of the engagement activities.

This report summarizes the main ideas that have emerged from these activities, and is divided into three themes central to the Policy review: 1) Pathways to OA - focused on the routes available to achieve open access (OA); 2) Measurement of OA - focused on monitoring and supporting Policy compliance; and 3) Culture of OA - focused on the ecosystem of OA in relation to Open Science, at all levels from research-performing institutions to research funders, and other partners.

Key observations

• Many participants voiced support for OA models that are low- or no-cost to the individual researcher, such as diamond OA and green OA. Participants also supported the inclusion of preprints in the publishing life cycle, and in the revised OA Policy. Copyright-based approaches to support the revised Policy such as a Rights Retention Strategy and Copyright Act reform targeting research outputs were also recommended.
• Widespread concerns were voiced about OA publishing costs, in particular (increasing) Article Processing Charges (APCs); these were considered the paramount barrier to publishing OA and a significant barrier in particular for early-career researchers and equity-deserving researchers. Researchers strongly questioned the use of research funds to pay APCs as opposed to other project costs (e.g., student support) and suggested cost-containment options for APCs (e.g., a limit on the amount of agency grant funds that can be used for APCs). There was strong support across those consulted for a revised Tri-Agency OA Policy that would reduce the financial burden of OA publishing for the individual researcher.
• The agencies heard the need for more investments in OA infrastructure, including repository infrastructure and direct support for scholarly journals beyond the support currently available via SSHRC to journals in the social sciences and humanities fields, to further incentivize cost-effective OA practices.
• Participants identified early-career researchers as a key group who could potentially be negatively impacted by the Policy revision. Disciplinary differences and different research dissemination practices were also one of the main concerns raised by participants. Multilingualism in local communities, as well Indigenous knowledge and Indigenous research, were also highlighted as areas that need to be addressed by the new Policy.
• OA is a fast-moving, evolving activity, with international policies that are continually progressing. Many of those consulted stressed the need for the revised Tri-Agency OA Policy to be aligned with the international research environment, while remaining responsive to the Canadian context, local challenges and needs; and leaving room for change and adaptation.
• More support and guidance are needed to help researchers navigate the Policy. The landscape of OA and the various options to achieve compliance with the OA Policy are not fully understood by researchers. To this end, there were recommendations and support for the agencies to require OA institutional strategies to help identify and align supports and incentives within institutions, as well as between institutions and the agencies.
• It was suggested that the Policy ensure researchers have access to the necessary resources to make OA content accessible for everyone. Currently, many green OA publications do not meet basic accessibility guidelines. For instance, some older publications and publications available in repositories are not readable with a screen reader.
• Many of those consulted expressed support for the agencies to incentivize, monitor and enforce OA compliance, including the use of Persistent Identifiers (PIDs) at every level (e.g., author, paper, institution, funder, etc.) to more accurately track agency-funded research publications. Additionally, there was interest expressed in the agencies collecting more data on the amount of grant funds spent on APCs, to ensure the sustainability of the revised OA Policy.
• Members of tri-agency Indigenous advisory bodies noted that the OA Policy is presently only concerned with peer-reviewed journal articles, which does not reflect the diversity of knowledge mobilization/transmission activities undertaken by, and in support of, Indigenous communities and their knowledge needs. It was suggested that these different forms and uses of knowledge, from an Indigenous perspective, should also be promoted and valued in research assessment.
• Large publishers expressed that transformative agreements (TAs) and gold OA are the key enablers for the transition to more equitable, high-quality and immediate OA publishing. These respondents had concerns about the sustainability of funding for diamond OA journals, their reach and impact. They support the sharing of preprints and self-archiving of journal research via repositories (with varying levels of copyright retention and licensing). They view APC-supported publishing as key to maintaining the integrity of the publishing system and providing support and access to other open science outputs (e.g., preprints, data, software, code) linked to the Version of Record (VoR).

 

top of page

---

2. Background

On July 4th 2023, the Presidents of NSERC, SSHRC and CIHR announced a review of the Tri-Agency Open Access Policy on Publications, originally published in 2015. The revised Policy, which will be released by the end of 2025, will require peer-reviewed journal publications arising from agency-supported research be made freely available, without subscription or fee, at the time of publication (i.e., immediate OA).

A public, online survey was also launched on July 4, 2023, concurrently with the tri-agency announcement. The survey captured information and research community views on immediate OA as a Policy requirement, including any perceived benefits and barriers, preferences for OA models, and supports the agencies could provide. Over 1,400 responses were received from members of the research community and partners. The survey results were made available in early 2024.

From summer 2023 to spring 2024, the agencies engaged with various members of the research community and partners from the scholarly publishing ecosystem, in Canada and internationally. The objectives of these engagements were:

1. to engage the full breadth of the Canadian research community in discussions on the revision of the Tri-Agency OA Policy; and
2. to explore the various perspectives of the research community, publishers, and other key partners towards the revision of the Policy in the context of other relevant trends facing the research ecosystem.

 

top of page

---

3. What we heard

3.1 Methods of engagement

In addition to the survey, the agencies participated in and conducted in-person and online meetings with various members of the research community and partners. Participants included experts (researchers, librarians, etc.), national and international granting agencies, small and large publishers, university presses, institutions, research administrators, early-career researchers, members of tri-agency Indigenous advisory structures, consortiums and non-profit organizations, editorial board members and volunteers from the survey, amongst others. The agencies also received multiple written submissions from various groups of researchers, publishers, and librarians. Other sources of input included feedback from CIHR’s University Delegates, as well as NSERC and SSHRC Leaders.

Nine invitation-only community-wide engagement sessions (6 in English, 3 in French), covering three broad themes (Pathways to OA, Building a Culture of OA, and Measuring OA) that emerged from the survey results, were held in February and March 2024. A cross-section of invitees was selected iteratively based on their expertise, experience, availabilities, interests and on how the Policy revision might affect them. A total of 111 participants (30 in French, 81 in English) contributed to the sessions. Each session lasted 90 minutes and included between two and three break out groups composed of four to eight people who freely discussed questions related to the themes, followed by a group plenary.

3.2 Key findings

3.2.1 Theme 1: Pathways to Open Access

In this theme, the agencies witnessed substantial interest and discussion about the different models and tools for OA publishing. Embedded in these discussions were concerns about equity and financial sustainability.

Article processing charges and transformative agreements:

• Many participants expressed concern about Article Processing Charges (APCs), primarily for reasons of costs. Opinions included: APCs are too expensive; their costs keep rising; they encourage predatory publishing; and studies have shown that they do not accurately reflect the cost of publishing.
• Many participants stressed the importance of ensuring that changes to the OA Policy not lead to more and/or higher publishing costs. Participants noted that researchers may already be faced with difficult choices between publishing OA and funding their research, and that increased costs for OA publishing would be particularly harmful for early-career researchers, researchers in equity-deserving groups, and others without stable research funding.
• Discussion of APCs raised issues of cost containment. Proposals included limiting or capping the amount of agency grant funding that can be used by a researcher to pay APCs, based on specific criteria such as the number of papers published or as a percentage of total grant budget. Other participants argued against this, pointing out that APCs are a current feature of the system that would be difficult to transition away from entirely given the incentive structures that drive researchers’ selection of publication venues.
• International funders who participated in the engagements noted that their grant funds can be used to pay APCs, but hybrid OA journals are often not eligible, unless they are a part of a transformative agreement (TA). At the Canadian level, participants mentioned the existence of several deals between Canadian Research Knowledge Network (CRKN) member institutions and large publishers (e.g., Elsevier, Wiley, Sage, etc.).
• Participants noted that some libraries have special funding available for researchers to pay APCs; however, the funds available are usually not enough to cover all the fees requested.
• Large publishers expressed that TAs and gold OA are the key enablers for the transition to more equitable, high-quality and immediate OA publishing. They view APC-supported publishing as key to maintaining the integrity of the publishing system and providing support and access to other open science outputs.
• While some participants mentioned that TAs allow libraries to save some money on subscriptions, the sustainability of these agreements was criticized by others:

Repositories and preprints (green OA):

• Numerous participants recommended the creation of a large-scale tri-agency repository where funded authors would be mandated to deposit all peer-reviewed journal publications arising from agency-supported research, citing the success of the National Institute of Health’s PubMed Central as a successful example of this model. The previous PubMed Central Canada repository was also mentioned as a reference point for a potential future venture.
• Some participants mentioned the Canadian Association of Research Libraries’ (CARL) efforts that are underway to create a federated model of institutional repositories on open infrastructure,^{Footnote 1} as well as efforts to support smaller institutions by creating joint repositories. It was suggested that the agencies coordinate with CARL to potentially leverage this initiative for the revised Policy.
• Some participants noted that many repositories (databases and search engines) are struggling with a lack of visibility, discoverability and accessibility.
• Engagement with members of tri-agency Indigenous advisory structures indicated support for repositories including a potential database for Indigenous knowledge, controlled by Indigenous rights-holders.
• Many participants and large publishers expressed support for the sharing of preprints (with varying requirements for where the Version of Record (VoR) is saved and/or linked) and self-archiving of journal research via repositories (with varying levels of copyright retention and material licensing for accepted manuscripts) as a way to comply to the Policy.

Free for authors and readers model (diamond OA):

• The agencies heard that diamond OA was often the preferred model by the research community and viewed as an aspirational goal. Engagement with international funders revealed that some are currently supporting diamond OA initiatives. Engagements also reflected opinions that a transition to a fully diamond OA landscape will need time and support, in terms of funding, human resources and adoption by researchers.
• SSHRC, through its Aid to Scholarly Journals (ASJ) grants, directly supports diamond OA initiatives. Some participants suggested that the ASJ grants be expanded to include journals across all scholarly disciplines, or that NSERC and CIHR create their own mechanisms to support diamond OA journals in STEM and Health Research disciplines.

Copyright:

Two different copyright approaches were discussed by participants:

• The first approach proposed was a complete reform of copyright law, establishing a new framework with set prices between publishers, the research community, and a copyright board that would be re-negotiated after an agreed period. This approach would be similar to the statutory licensing of the music and the video gaming industries. It was noted that the role of the agencies would not necessarily be to lead this reform, but to use their unique position to convene all the relevant parties to the table for negotiations.
• The second approach proposed was requiring a rights retention strategy or the implementation of an amendment to the Copyright Act, enshrining secondary publishing rights. This would allow researchers to disseminate their peer-reviewed journal publications through alternative routes, such as repositories.

Other observations: Enabling policy implementation

• Participants suggested that a "top-down" approach, such as Plan S and the The [U.S.] White House Office of Science and Technology Policy (OSTP) Nelson Memo, would lead to more meaningful changes; participants stressed the role of the agencies in setting direction and standards across the research community. At the same time, participants stressed the importance of a flexible Policy that would leave room for change and adaptation.
• Some participants highlighted how the adoption of institutional strategies (similar to those mandated by the Tri-Agency Research Data Management Policy) would facilitate implementation of the revised OA Policy. The adoption of institutional strategies would also help mobilize institutions, beyond the libraries where engagement on OA is currently concentrated. At the same time, participants also indicated that institutional strategies can be resource-consuming and burdensome, particularly for small institutions.
• Participants highlighted that the existing Érudit-Open Journal System infrastructure (Coalition Publica) should be supported and leveraged to advance immediate OA. The expertise of librarians and libraries more broadly should also be used for training researchers and managing infrastructure.
• Participants highlighted that the broader context of Open Science (e.g., the UNESCO Recommendation on Open Science, the Office of the Chief Science Advisor’s , etc.) should be considered, and the new Policy should be aligned with these values.

3.2.2 Theme 2: Building a culture of Open Access

In this theme, participants discussed the best ways the agencies can help foster a culture of openness among the research community. Various related topics such as awareness, training, incentives, and research assessment were discussed by the participants.

Research assessment

• Discussions of what is OA culture and how to encourage OA frequently prompted responses about current norms in research assessment and the dominance of the prestige system of research (e.g., the use of indicators such as journal impact factor)
• Participants underlined the importance of alignment at the national and international levels (e.g., through initiatives such as DORA, COARA and the Barcelona Declaration on Open Research Information) to not penalize researchers prioritizing openness in their career choices.
• The importance of aligning national with institutional incentive structures was also noted by many, although participants also indicated that the agencies could play a leadership role in this area and work to be as transparent as possible about potential changes in assessment.
• Many participants supported including OA criteria within research assessment, for example by only allowing OA publications to be considered in peer review of grant applications, or by only allowing researchers who had complied with the OA Policy to reapply for agency funding.
• Many ideas to help the agencies adapt their excellence criteria to better align incentive structures with OA publishing were also suggested, such as limit eligible outputs to five, only consider or give more value to open outputs, remove journal names on CVs, give more value to local platforms and journals, give more value to knowledge mobilization activities beyond journal publications (e.g., editorial work, popularization of science, etc.), join COARA, implement a narrative-style CV or other more qualitative assessment methods that would allow for research assessment that is increasingly focused on quality and contribute to the valorization of OA.
• Participants also believe that the agencies' program officers, peer review committee Chairs, and peer reviewers need to be more empowered with clearer merit review criteria in grant application evaluation to ensure that some of the disciplines that are more resistant to change are encouraged to act on it, leading to a broader culture change and better consideration for DORA principles.
• Discussion of assessment also raised concerns about unfair impacts. Participants identified early-career researchers as the first group who could potentially be negatively impacted by the Policy revision. It was noted that they usually have a lot of pressure to publish in prestigious journals and transitioning to immediate OA could lead to them having to pay more APCs while having limited funding.
• Disciplinary differences and different research dissemination practices were also of concern to participants. The scope of the current Tri-Agency OA Policy only includes final, peer-reviewed manuscripts and/or publications, however many disciplines in the social sciences and humanities have a very strong monograph culture. There is a need for more awareness about the value of these outputs.
• Similarly, members of tri-agency Indigenous advisory structures also noted that the current OA Policy only pertains to journal publications, which does not reflect the diversity of knowledge mobilization products and activities undertaken by researchers.
• It was suggested that different ways of knowing and forms of knowledge mobilization (KM) should also be promoted and valued in research assessment. Respondents indicated a desire to create space within the Policy for additional Indigenous knowledge activities. Because of the history of Western extractive research practices, it was emphasized that research, including outputs such as data and publications, must directly benefit the community and must be handled very carefully so as not to be used against the community.
• Along the same thinking about inclusion and recognition of different knowledge products, some participants believe that other types of outputs should also be monitored by the agencies. These outputs include data, codes, peer reviews, software, etc. Mentions were also made to accessibility issues and formats to disseminate research results, such as American Sign Language videos.
• The importance of multilingualism in local communities was pointed out as a major stake for OA and research assessment going forward. The potential of artificial intelligence (AI) for (mis-)translating publications was discussed, although many participants underlined that the translation of scientific concepts and theories is a considerable issue.
• Looking at the international level, participants mentioned that current OA models are not inclusive in terms of participation, and that they are not working for all international partners and the Global South, in general because they are too expensive.

Other challenges to culture of Open Access

• In addition to discussions about research assessment, the engagements surfaced other observations that could be considered as barriers to building a culture of OA. These included: predatory publishing practices, unclear or ambiguous policies (e.g., how embargos work, the place of preprints, etc.), and how cultural shifts implicate many players and take a long of time and effort.
• Institutions and librarians emphasized that there is a certain misunderstanding of the Policy and of OA in general within the research community (e.g., many view APCs as the only way to publish in OA). They also raised the strong link between culture and compliance monitoring, and how there is a need to ensure that funded researchers have a better grasp of the Policy and how they can comply. Participants also suggested that being more explicit in the funding notice of decision about grant holders’ responsibilities regarding requirements for funder acknowledgements and OA publishing could make a difference.
• As for potential solutions to these barriers, participants suggested more funding directed toward OA publishing, more training and support to help researchers navigate the system (e.g., for-profit, not-for-profit, predatory publishing, authors’ rights etc.), more TAs through library consortiums such as CRKN, institutional OA strategies, and OA mandates.
• It was suggested that the OA Policy ensure researchers have access to the necessary resources to make OA content accessible for everyone. Currently, many OA publications do not meet basic accessibility guidelines. For instance, some older publications and publications available in repositories are not readable with a screen reader.
• Looking at the international level, participants mentioned that current OA models are not inclusive in terms of participation, and that they are not working for all international partners and the Global South, in general because they are too expensive.

3.2.3 Theme 3: Measuring Open Access

In this theme, participants discussed whether the agencies should more closely monitor adherence to the OA policy. Additionally, more technical aspects such as the most realistic ways to monitor OA publications resulting from agency-funded research (e.g., data sources, methods, resources, time, etc.) were discussed.

Monitoring OA

• Overall, participants believe that more and better data on funded researchers' outputs are needed at every level. For example, the agencies do not currently track how much money is spent on APCs or the compliance rate of the OA Policy.
• Many participants also emphasized that monitoring would increase Policy compliance and indicated that the results of monitoring activities should be public and always up to date. For instance, they noted how it could be built into, allowing the implementation of an automated monitoring system (e.g., through PIDs).
• Many also discussed the need to streamline compliance. For example, manually self-reporting through the Canadian Common CV (CCV) was considered wasteful and counterproductive as this data is hard to enter, remains private and is not reusable.
• It was also noted that one of the limits of monitoring is the timeline of grants compared to research outputs. Some research outputs that are the product of the financed research are published beyond the life cycle of the grant, and after researchers have submitted their final reports to the agencies.
• Some participants believed that only open resources should be used for monitoring compliance given that for-profit organizations have been using their data as leverage in negotiations with universities and governments. However, it was also noted that although open data sources are more inclusive and will keep improving, the metadata quality of these resources is not yet at the same level as proprietary ones. The principles of Open Scholar Infrastructure were mentioned.

Mechanisms to increase compliance:

• Some participants mentioned research that shows the best way to have higher compliance rates is to implement meaningful consequences of non-compliance (e.g., withholding funds, less points in evaluations, etc.), otherwise researchers will never take the Policy seriously.
• It was also noted that the agencies need to make compliance with the Policy easier through various means such as tools, education and capacity building, and not just adding more administrative work for researchers.
• Many participants also stated that giving more positive incentives to comply with the Policy (e.g., more points in evaluations for open research outputs, additional funds for researchers with high OA Policy compliance rates, more funding for APCs, prizes, etc.) would lead to a more successful Policy overall.
• Participants noted that any implementation of consequences for non-compliance should be phased-in to give the research community an opportunity to adapt and, for those who are non-compliant to become compliant.

Persistent identifiers (PIDs) and measurement of OA:

• Many participants pointed out the use of PIDs at every level, ORCID for researchers, DOIs for publications and grants, ROR for institutions, etc. is key for interoperability. Open Science dashboards such as The ANR Open Science Monitor and McGill University’s Tanenbaum Open Science Institute (TOSI) Open Science Dashboard were given as examples of how the agencies should report their monitoring.
• ORCID was suggested as a potential solution to replace the CCV. There was no consensus on if it should be mandated or not. Some security concerns were also pointed out, given that ORCID information is stored on servers outside of Canada; however, others pointed out that none of this information is private. Some concerns were also expressed about the lower adoption of ORCID in certain disciplines, notably in the social sciences and humanities.
• While the majority of participants were in favour of the implementation of PIDs, some reminded the agencies of the importance of supporting the infrastructure needed in their implementation (e.g., journals, platforms, etc.). The implementation of PIDs would also require a lot of coordination at every level (e.g., Canadian institutions, Canadian PIDs Working Group, other funders, open platforms, etc.) to ensure proper adoption.
• Participants mentioned that PIDs could help overcome many problems and even reduce biases related to disambiguation, especially for Black, Indigenous, and People of Colour (BIPOC) researchers, such as name duplicates and variations.

 

top of page

---

4. Next steps

The agencies will draft a revised OA Policy, based on the consultation feedback received to date and future, and pursue engagements with partners as needed. The new Policy will be published by the end of 2025.

Contact us:

• CIHR: openaccess-libreacces@cihr-irsc.gc.ca
• NSERC: openaccess@nserc-crsng.gc.ca
• SSHRC: openaccess@sshrc-crsh.gc.ca

 

For information on the Tri-Agency Open Access Policy on Publications (2015), please visit: https://science.gc.ca/site/science/en/interagency-research-funding/policies-and-guidelines/open-access.

 

top of page