Open access: reflections on revised Plan S

Published: 2019-07-01

I have previously discussed open access and Plan S in my blog, and in our Lancet commission on Global Governance for Health we argue that knowledge must be available to all. While I strongly support the idea and legitimacy of open access there is a need to be attentive to the demands and academic freedom of the individual researchers. Open access is well supported by the academic community, but this support may easily be undermined by a forced and uncritical implementation process such as that outlined in the original version of Plan S. This plan has now been revised and improved, and much of the critique has been taken into account.

The revised version of Plan S was published by cOAlition S last month. The fundamental principle of Plan S remains: it asks for full and immediate open access (OA) for all peer-reviewed scholarly publications. Immediate open access means that no embargoes are allowed. However, the timeline for funders to implement Plan S has been extended from 2020 to 2021. There is one more year for all parties – researchers, funders, publishers, higher education institutions – to adapt.

Revision has brought more clarity when it comes to the various routes to OA. The previous version placed emphasis on ‘pay-to-publish’ OA journals. It is now clearly stated that researchers can publish in a broad range of journals if they deposit the accepted version of the manuscript in a repository such as Europe PubMed Central, PubMed Central or KI Open Archive. This – the plan says – must be done immediately and with license for reuse.

What does this mean for our researchers?

For researchers with grants from Plan S funders this means that you have to make your publication immediately open access, by publishing in OA journals/platforms or in subscription journals that allow depositing the manuscript in a repository. As stated above, this must be done immediately on publication.

Researchers are also allowed to publish in so-called hybrid journals, i.e., subscription journals where single articles are made OA for a fee that is often quite substantial. The  precondition is that the journal is covered by an agreement that sets a path for transforming the journal to open access. This route is permitted until 2024.

Plan S funders or research institutions will cover publication fees in OA journals. This can be done by providing financial support through agreements with publishers of OA journals, as happened recently when several Swedish funders entered into an agreement with Springer Nature.

The KI library has signed several OA publishing agreements where publishing fees are either reduced or fully prepaid for KI researchers. The library will continue to assess and enter into cost-effective agreements that will make it as easy as possible for KI researchers to comply with Plan S.

According to Plan S, funders will monitor compliance and sanctions may be introduced for non-compliance by individual funders. Sanctions may include withholding of grant funds or excluding grant holders from future funding calls.

Open access in the ideal world

As I wrote in my blog on the first version of Plan S, it would be ironic if a plan intended to promote open access (a goal I warmly embrace) should end up doing the opposite. The freedom to choose publication channel is an integral and essential aspect of academic freedom. Terence Karran and Lucy Mallinson include among “The Four Pillars of Academic Freedom” the freedom to “determine the avenues and modes (conference presentations, journal articles) of disseminating research findings to one’s peers, and the wider world”.

In the ideal world, society should have free access to research while researchers should be free to publish in the journal of their choice. The art is to let these freedoms thrive side by side. The revised Plan S brings us closer towards this goal. Hurdles remain, and these must be tackled with due attention to the voices of the academic community. Without support of the academic community, any plan for OA implementation will fail.

Further information is available at the KI Library


Related posts