Thoughts on Plan S

First see the plan itself:

cOAlition S: Making Open Access a Reality by 2020: A Declaration of Commitment by Public Research Funders

Plan S: Accelerating the transition to full and immediate Open Access to scientific publications

Preamble by Marc Schiltz: Science Without Publication Paywalls: A Preamble to: cOAlition S for the Realisation of Full and Immediate Open Access

Members of cOAlition S on the day of launch, September 4, 2018

Statement by Carlos Moedas: 'Plan S' and 'cOAlition S' – Accelerating the transition to full and immediate Open Access to scientific publications

Press release: cOAlition S: Making Open Access a Reality by 2020

Overview. The plan is admirably strong. It aims to cover all European research, in the sciences and in the humanities, at the EU level and the member-state level. It's a plan for a mandate, not just an exhortation or encouragement. It keeps copyright in the hands of authors. It requires open licenses and prefers CC-BY. It abolishes or phases out embargoes. It does not support hybrid journals except as stepping stones to full-OA journals. It's willing to pay APCs but wants to cap them, and wants funders and universities to pay them, not authors. It will monitor compliance and sanction non-compliance. It's already backed by a dozen powerful, national funding agencies and calls for other funders and other stakeholders to join the coalition.

There are two main weaknesses or aspects to watch closely. First, on one reading, the plan welcomes both gold and green OA, which is good. Its key principle requires distribution through "compliant Open Access Journals...or compliant Open Access Platforms." But a section elucidating this principle damns green OA with faint praise, endorsing OA repositories only for preservation, not for OA itself, repeating the mistake of the Finch Group in 2012. It's not at all clear how far the coalition will let green OA satisfy the upcoming policies.

Second, the plan promises support for OA infrastructure, which is good. But it never commits to open infrastructure, that is, platforms running on open-source software, under open standards, with open APIs for interoperability, preferably owned or hosted by non-profit organizations. This omission invites the fate that befell bepress and SSRN, but this time for all European research.

- - - - - - - - - -

Here's a fuller picture. I start with some strengths and then mention some weaknesses. I treat the authoritative preamble by Marc Schiltz as part of the plan itself.

1. The plan is roughly what you'd want after 15 years of funder experimentation with weaker policies. Those policies proved that OA unquestionably helps research and researchers. They proved that that the sky does not fall, and they proved that recalcitrant publishers and misguided incentives still hold back progress. After all that experience you'd want an important body to push for stronger policies on a large scale. You'd want to recognize that there's much more progress to be made, good reasons to seize the opportunities, and -- for the public interest as opposed to private interests -- no good reasons to hold back.

This isn't the first bold funder policy. But it's one of the few that I'd call strong, and it's among the strongest.

If it matters here, I'm an incrementalist. I always wanted strong OA policies, as strong as this one and even stronger. But I saw the political obstacles to adopting them 5, 10, or 15 years ago. Policies which might look weak today, but which went about as far as they could in their own time and place, helped the cause by showing that OA benefited researchers, showing that the sky did not fall, showing that removable obstacles still remained, and building the case to remove the remaining obstacles.

2. The plan recognizes the duty of public research funders to promote research. So far, so good. But it also recognizes their "fiduciary responsibility" to spend public money in the public interest. That's exactly right and too often ignored. When drafting OA policies, too many public funders compromise the public interest in order to accommodate private corporations. Sometimes the blame belongs to national legislatures for letting publisher lobbying water down strong first-draft policies.

3. The plan requires open licenses and prefers CC-BY. That's exactly right. Again, earlier weaker policies didn't go this far, and settled for gratis access (removing price barriers alone) and didn't hold out for libre access (removing both price and permission barriers). In the early days it might have been politically infeasible to go beyond a gratis mandate to a libre mandate. But it's no longer politically infeasible, and any policy-makers with the power of the agencies behind this plan should take full advantage of that.

4. Without naming Elsevier, the plan recognizes the difficult negotiations with Elsevier in Germany, France, and Sweden. It regards them as one more reason to act now, which is correct. Just as public funders have been slow to act decisively in the public interest, with minimal compromise, universities have also been slow to act decisively in the interest of research. These two families of non-profit, public-interest institutions can embolden each other, and it's good to see recent bold action by universities echoed in bold action by public funders.

5. The plan wants to "monitor compliance" and "sanction non-compliance". That's good. The funders with pioneering OA policies did little of the first and none of the second. Even though we've long since learned best practices here, far too few contemporary funder OA policies include steps to address non-compliance.

6. The plan is already backed by 11 major agencies, both at the EU level and the member-state level. It intends to apply to both EU-level policies and member-state policies (that is, to "national and European research councils and funding bodies"). In addition to the major agencies already on board, it calls for other agencies as well as "researchers, universities, libraries, and publishers" to join the coalition. Despite my criticisms (below), I'll join as soon as I'm able.

Now for some weaknesses, building up from smaller ones to larger ones.

7. The plan uses the term "science" to include the humanities. That's needlessly confusing. Instead of calling for open access to "science" and letting most people misinterpret its scope, it would have been easy to call for OA to "research" or "scholarship" in all fields.

This isn't just a quarrel about words. The coalition is built mostly of "science" agencies whose remit does not cover the humanities.

Moreover, the plan may assume too quickly that the OA principles that apply in the sciences apply without qualification in the humanities. The one exception it acknowledges is that it may take longer to implement an OA book mandate than an OA article mandate.

8. The plan says in bold type that "access to research publications...cannot be monetised in any way." Three paragraphs later it says that publishers "may charge fair value for [their] services in a transparent way." This is an apparent inconsistency. But the two principles are compatible and an easy revision could make that clear. The first principle applies only to research access. Access must be free of charge, or cannot be monetized. On the second: the entire history of OA shows that free access is compatible with publisher revenue, even profit.

9. The plan says that the "so-called 'hybrid' variants [on subscription journals]...should be terminated." Later it also says that "the 'hybrid' model of publishing is not compliant with [these] principles." But in between these two unqualified statements it adds the qualification that hybrid journals are acceptable as stepping stones to full OA. "Therefore, it is acceptable that, during a transition period that should be as short as possible, individual funders may continue to tolerate publications in 'hybrid' journals." In context, this exception leaves room for "offset" agreements, which only apply to hybrid journals. I accept the fully nuanced position here -- refusing to pay APCs at hybrid journals except those that can demonstrate that they are converting to full (non-hybrid) OA. But the plan contains an apparent inconsistency that a simple revision could fix.

10. The plan wants an exceptionless OA mandate: "_all_ European scientific publications should be freely accessible by 2020" (emphasis added). But it recognizes that author freedom to submit new work to the venues of their choice must take priority, in part because some European jurisdictions assure this freedom by law. The problem here is not deference to author freedom. The problem is the apparent inconsistency between the calls for an exceptionless mandate and this important carve-out.

Note the plan's two strategies to minimize the exceptions that this carve-out could allow.

Knowing that the growth of OA depends largely on author choices, it aims to influence their choices. It exhorts researchers to "realise that they are doing a gross disservice to the institution of science if they continue to report their outcomes in publications that will be locked behind paywalls." This is a true observation about the consequences of certain author choices twisted into an exhortation. You don't change researchers' behavior by berating them for doing a gross disservice to science.

Knowing that author choices are largely influenced by existing incentives, it aims to change those incentives. It doesn't acknowledge all these incentives, but it does identify the use of bad metrics like JIF. It recommends the DORA principles (from the San Francisco Declaration on Research Assessment). Again, I support the recommendation. The main point here is that it's a mere recommendation. The plan adds, "We therefore commit to fundamentally revise the incentives and reward structure of science." OK. But it says nothing about how it will carry out this commitment, apart from recommending DORA. Those of us who have worked on this front know that the needed changes must be made by individual hiring, promotion, and tenure committees in individual departments of individual universities (and by the individual counterpart review committees at individual funding agencies). This is a huge, messy problem that cannot be waved away by declaring a good commitment and endorsing good principles.

On the plus side, the plan recognizes the limits of the coalition's policy-making power. On the minus side, it offers nothing more than recommendations to minimize this large carve-out, and seems to underestimate the difficulty of implementing its recommendations.

11. It wants participating funders to pay fees at fee-based OA journals. But it seems largely unaware that there should be strategic restrictions on how to spend this money in order to create good incentives for the authors and publishers receiving the money. The only restriction it acknowledges is that the fees should be "standardised and capped". There are many more it ought to consider.

12. The plan criticizes predatory journals by saying that they "misuse the Open Access model of publishing". The problem here is not the diagnosis of predatory journals, but the assumption that there is just one business model for OA journals, namely, the APC model. This is an elementary mistake that shouldn't appear in a doc from informed, high-level policy-makers. It would have been very easy to say that predatory journals "misuse the APC model of OA publishing" and drop the implication that there are no other models.

This is not just a quarrel about wording. The plan doesn't acknowledge the existence of no-fee OA journals, let alone their dominance. Most peer-reviewed OA journals today are no-fee, and always have been. Nor does the plan acknowledge the importance of no-fee OA journals in the larger ecosystem of OA options. No-fee OA journals are critical for unfunded and under-funded researchers. The no-fee business models resist predation. (There's a reason why most predatory journals charge APCs.) If the plan wants to sustain fee-based OA journals by paying APCs, as it does, then it should also want to sustain no-fee OA journals. It might do that if it acknowledged their existence, their preponderance, and their value.

13. The plan's bold-type "key principle" says that publicly-funded research "must be published in compliant Open Access Journals or on compliant Open Access Platforms." This principle opens the door to both gold and green methods of compliance. That's good. But it almost completely shuts the door on green OA just a few paragraphs later when elaborating the principle. "The importance of open archives and acknowledged because of their long-term archiving function and their potential for editorial innovation." There's no acknowledgement of their importance for OA itself! This is the same mistake made by the Finch Group in 2012, which was inexcusable even at the time, and should never be repeated by informed, high-level policy-makers. Going forward, we'll have to watch with care to see whether the new OA mandates can be satisfied by green OA, or whether they will be gold-only or gold-favoring policies like those that have so notably failed in the UK.

See my critique of this Finch Group mistake at the time, in my newsletter for September 2, 2012.

I made another version of the same point in this interview with Richard Poynder, July 23, 2013.

I support the plan's call for unembargoed OA under open licenses. Many people, perhaps including the authors of this plan, believe that green OA must be embargoed. That's false. Many also believe that green cannot stand under open licenses (that is, provide libre access). That's also false. Some green OA is embargoed and some isn't. Some green is libre and some isn't. I'd be fine if the policy only allowed green OA when it was both unembargoed and libre. But because some green OA can be both, the policy should drop the language that virtually closes the door on green OA after emphatically opening it. It should cut this apparent inconsistency, and be explicit that unembargoed libre green OA is just as satisfactory as unembargoed libre gold OA. It should acknowledge the value of repositories for OA itself, not just for preservation and editorial innovation.

I've often called for green OA even when it was embargoed and without the benefit of open licenses. I still support that, especially when unembargoed or libre OA are out of reach. What's promising about this plan is that it is determined to put unembargoed libre OA within reach. Under these circumstances, it should at least support green and gold paths to this goal, and not rest on the mistaken assumption that one of them cannot lead to the same goal.

Here's another perspective on this: The easiest and most effective way to provide unembargoed libre green OA is with systematic rights-retention policies. If authors retain the rights need to deposit works in OA repositories, then the repository copies can be unembargoed and libre. Funders like the NIH and Wellcome Trust have rights-retention policies, though too few other funders have followed their lead. Nearly 100 universities worldwide have rights-retention policies for faculty. Several European countries (the Netherlands, France, Germany) have provided rights retention by statute, with different conditions and exceptions. The plan insists that "authors retain copyright...with no restrictions." That's more than enough to make unembargoed libre green OA possible. Hence, allowing unembargoed libre green OA to satisfy the future OA mandates is not likely to be a small concession taken up by few. It could be taken up by any and all. It would give authors a serious choice between green and gold OA, for every new work. It would solve the problem of insufficient funds to pay APCs at fee-based OA journals. It would solve the problem facing early career researchers who must comply with their funder policies while still satisfying their promotion and tenure committees (committees that didn't get the memo recommending that they change their incentives).

14. There are a few important areas of vagueness to be clarified later. These are not objections as much as spots to watch.
* The OA mandate for articles must be in place by January 1, 2020, but the OA mandate for books can take longer, provided that the extra time is "as short as possible". How short is that, and who decides?
* Similarly, offset agreements and paying fees at hybrid journals are acceptable for a transition period provided that the period is "as short as possible". Again, how short is that, and who decides?
* If authors will retain copyrights, then journal publishers will get some kind of license to publish. But what kind of license, and who decides? Some licenses give publishers no more right of first publication, and some are nearly equivalent to transferring permanent exclusive rights.
* "Funders will ask universities, research organisations, and libraries to align their policies and strategies" with this plan. Alignment would be good, and it's good that these funders will ask for it. But how persuasive will they be? What leverage will they have? Do they appreciate that this kind of diplomatic mission is very far from the simplicity of policy-making?
* The plan limits itself to principles and deliberately omits implementation details. "Science Europe, funders, the European Research Council and European Commission will work together to clarify and publish implementation details." That's entirely reasonable. But we should expect many devils in those details. Moreover, European institutions could bypass some of the implementation debates if the guidelines in the plan were more clearly written and more sensitive to details of the OA landscape.

15. The plan promises "support...for Open Access infrastructures where necessary." So far, so good. But the plan is silent on the importance of open infrastructure, that is, platforms running on open-source software, under open standards, with open APIs for interoperability, and preferably owned or hosted by non-profit organizations.

The plan is signed by Marc Schiltz (President of Science Europe), and says that it was initiated by the Open Access Envoy of the European Commission, Robert-Jan Smits (formerly the EC Director-General of Research and Innovation). While it credits other organizations, its chief co-authors seem to be Schiltz and Smits. Earlier this year both Science Europe under Schiltz and the EC under Smits called for new infrastructure for OA research. Both calls were good ideas, regrettably flawed. Both conspicuously failed to call for open infrastructure. It's as if Science Europe and the EC were unaware of the risks of hosting Europe's OA research on a proprietary platform, or unaware of what happened to SSRN or bepress.

See the Science Europe call for infrastructure (March 21, 2018)...

...and my note that it didn't call for open infrastructure.

See the EC call call for infrastructure (March 31, 2018)...

...and my note that it didn't call for open infrastructure.

- - - - - - - - - -

I gave comments on Plan S to Holly Else for Nature News, and Martin Enserink for Science Magazine, which appeared at the same time as the plan itself.

I thank them for quoting me, and completely understand why they couldn't include more details from my take on plan's strengths and weaknesses. As you see here, a fuller account is longer than the plan itself.

Shared publiclyView activity