Green open access and REF compliance

Yesterday’s blog post suggested that our reliance on Hybrid Gold Open Access to meet research funder’s open access requirements was too expensive and libraries should be supporting alternative pathways. In terms of pure numbers the main way the University of Edinburgh is making access to it’s research open is via Green Open Access (OA) – where the Author’s Accepted Manuscript (AAM) version of the research output is deposited in a repository and made available as soon as publishers copyright policies allow.

Since we started compiling monthly reports in 1st April 2016 we have made 6266 journal articles and conference proceedings open access which equates to around 92% of the University’s total output.

The University of Edinburgh has always had a preference for Green OA because there are no upfront costs for authors so everyone can participate without requiring access to large research budgets. But since the REF2021 open access requirements were published, Green OA has become even more important. The HEFCE policy states that, to be eligible for submission to the next REF, authors’ final peer-reviewed manuscripts must be deposited in a repository. The University of Edinburgh has adopted this policy and provided the infrastructure and support for all it’s researchers to make their research Green OA. The programme which was adopted to facilitate open access in the University of Edinburgh’s College of Medicine and Veterinary Medicine is described in more detail in the following paper:

Large scale implementation of open access: A case study at the University of Edinburgh’s College of Medicine & Veterinary Medicine

The results of the open access implementation plan have been extremely successful and we can report a 92% adoption rate for Green OA. The Scholarly Communications Team prepares monthly reports using data from our repository to identify the current level of compliance with the REF open access policy. These monthly reports indicate the numbers of in-scope research outputs (journal articles and conference proceedings with ISSNs) and whether they meet the requirements for the next REF, as per our repository’s compliance-checker functionality.

The figure and table below show the growth in the percentage of research outputs made open access since April 2016 and also a more detailed breakdown by subject area. The reporting period is a rolling window three months behind the current date.

1 April 2016 – 31 May 2017 Number of papers Indicative compliance
Arts, Humanities & Social Sciences
Business School 123 out of 126 97.62%
Divinity, School of 22 out of 22 100%
Economics, School of 21 out of 21 100%
Edinburgh College of Art 111 out of 118 94.07%
Health in Social Science, School of 118 out of 125 94.40%
History, Classics and Archaeology, School of 86 out of 88 97.73%
Law, School 102 out of 104 98.08%
Literatures, Languages and Cultures, School of 59 out of 61 96.72%
Moray House School of Education 155 out of 160 96.88%
Philosophy, Psychology and Language Science, School of 402 out of 405 99.26%
Social and Political Science, School of 181 out of 191 94.76%
Medicine & Veterinary Medicine
Clinical Sciences, Deanery of 1183 out of 1387 85.29%
Biomedical Sciences, Deanery of 214 out of 239 89.54%
Molecular, Genetic and Population Health Sciences, Deanery of 794 out of 910 87.25%
Veterinary Studies, Royal (Dick) School of 550 out of 573 95.99%
Edinburgh Medical School 672 out of 732 91.80%
Centre for Medical Education 7 out of 8 87.50%
Science & Engineering
Biological Sciences, School of 407 out of 455 89.45%
Chemistry, School of 203 out of 231 87.88%
Engineering, School of 373 out of 385 96.88%
GeoSciences, School of 354 out of 373 94.91%
Informatics, School of 366 out of 376 97.34%
Mathematics, School of 135 out of 135 100%
Physics and Astronomy, School of 340 out of 380 89.47%
Edinburgh Parallel Computing Centre 6 out of 7 85.71%


Despite a hugely successful Green OA implementation programme at the University of Edinburgh there are a number of significant outstanding problems. To my mind the main issues are that Green OA is:

Not immediate. Most publishers (but not all) require an embargo period of between 12-36 months before the AAM can be made open access. Not only are long embargo periods are hugely detrimental for scholarly communication, but I believe they are unnecessary. Many academic publishers insist on long embargo periods to protect journal subscription revenue from cancellations. However, a number of academic publishers (including the Royal Society, Cambridge University Press, Emerald and SAGE) have zero month embargoes for selected titles are they are not unduly affected by cancellations.

Not compliant with all research funders policies. Unfortunately, these long embargo periods are not compliant with all research funders policies (e.g. RCUK) which means that many researchers are forced to pay Hybrid OA fees for their publications. As we discussed yesterday Hybrid OA accounts for 70% of our open access expenditure. In the UK researchers are faced by a myriad of different funders policies (e.g. HEFCE, RCUK, Horizon 2020) which currently require different routes to open access.

Not cost-free. If you factor in hidden costs – such a repository platform fees and staff costs for mediated deposits/copyright checking – Green OA is more expensive than you would initially imagine. Whilst gathering data for a recent HEFCE open access questionnaire we estimated that we have around eleven full time equivalent staff distributed across the entire University working on open access implementation. If you aggregate all these costs then the cost per paper to deliver Green OA is somewhere in the region of £45 which I think is too expensive (but put in context this is still pretty good value for money as our average Gold OA costs are £1,676 per paper).

In summary, Green OA is a step in the right direction but still does not answer all of the problems that we currently face. What we can do to alleviate the issues of long embargo periods, harmonising research funders policies and simplifying processes and thereby lowering costs is the topic of tomorrows blog post – the UK Scholarly Communications Licence.

Leave a Reply

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