Editor’s Note: Today’s post is by Steven Inchcoombe. Steven is Chief Publishing Officer at Springer Nature. He is a member of the Management Board and his responsibilities cover all the research publishing and editorial activities of Springer Nature, the services provided to its authors, peer-reviewers, editors and societies, the experience of its customers when using its products and publishing platforms, and its Open Access and Open Research activities. Previously Steven was Head of House and CEO at Nature Publishing Group (NPG) and Palgrave Macmillan (PM).
In January 2005, a paper was published in the Journal of Biomedical Science. “T cell responses to allogeneic human mesenchymal stem cells: immunogenicity, tolerance, and suppression” explored whether stem cells could be transplanted into other humans without the immune system (i.e., white blood cells) recognizing them as “foreign” and stimulating an immune response. An important study in its own right, having been cited over 400 times, what makes this paper even more notable for us at Springer Nature is that it marks the start of the one million immediately and fully open access (OA) primary research and review articles that we have published and are able to analyze.
These articles cover all academic disciplines and range from those which provide groundbreaking research to that which adds quietly but importantly to the academic literature and supports the development of further research. And as research shows that articles published OA have increased impact, usage, and reach, authors have also benefited.
But reaching such a milestone is one thing — more important, given the data that it makes available, is what it can tell us about how the transition to OA is going and what we need to do now to help speed up progress to a more open science future.
How is the transition going?
Overall I believe the transition is going well. From a slow start, as a sector we have now reached the point where, in 2020, a third (33%) of articles were published OA.
Looking however at our 1m, this does, though, mask quite a variation amongst different academic disciplines. From our data, medical research is overwhelmingly driving the transition, with Medicine accounting for 44% of our 1m articles. The next highest discipline is Life Sciences, which accounts for 17%. This trend of Medicine outperforming the other disciplines so strongly has barely changed from 2015.
We are starting to see some strong growth in social science and humanities. As has been documented elsewhere, these disciplines are finding transitioning to OA and meeting Plan S’s requirements more challenging than those in the medical, applied and physical sciences. This is for a variety of reasons, but mainly because of the lack of available funding. It is therefore encouraging to see that, even though this was from a low base, between 2015 and 2020 social science and humanities doubled their OA share.
Our data also brings up some interesting geographic differences. For example, Europe and Asia are strong generators of OA content, accounting for 40% and 33% respectively in 2021. Conversely, North America accounts for 18%, perhaps demonstrating the relatively low engagement with Gold OA amongst US funding bodies to date.
More broadly, in the last 5 years, these 1m articles were downloaded 2.6 billion times. That’s an average of 2,600 downloads for each article, a strong demonstration of the value being derived from OA. Translating that into something institutional librarians might appreciate, that’s a cost of less than Euro 1 per download based on all their one-time APCs.
What do we need to do now?
First, we need to move faster as a company, as an industry, and as a broader ecosystem where funding agencies and institutions are key participants. In addition to having transitioned all of our owned non-fully OA journals to be Transformative Journals, placing them all, even Nature, on the path to OA, we are also committed to significantly increasing our overall amount of OA content. Therefore, by 2024, we are targeting over 50% of Springer Nature’s article output to be OA, immediately available for all to discover, share, use and reuse from the moment it is published.
Second, we need to take a good look at the challenges and barriers certain academic disciplines are facing. This needs to be a transition which is inclusive, one which brings everyone with us. One way of doing this is via Transformative Agreements. While our early Compact agreements did not always include journals from our predominantly non-STM portfolios, we are now addressing this, working with our consortium partners to bring these into our existing and new agreements. Enabling researchers to publish OA regardless of their academic discipline is an important role TA’s play and we want to ensure they play it to the full.
Third, there are clearly areas of the world that are struggling to transition their research output to OA. There will be a variety of different reasons for this and we need to work with them to find solutions that work for them and meet their needs. We and other publishers have a long track record of providing waivers and discounts for many, but this can’t scale sufficiently and we need other solutions that the whole research community will support.
If these three issues can be addressed then I believe the possibility of a fully OA future is looking positive. Publishers absolutely have a role to play in ensuring there are sustainable OA publishing options that meet the needs of all authors. Springer Nature is ticking that box with our 600 fully OA journals, our Transformative Journal commitment enabling authors to choose from well over 2000 further journals, and the introduction of new fully-OA journals. Publishers cannot do this alone however. We need funders, institutions and consortia to come with us on this full and immediate Gold OA journey and commit to policies which clearly place Gold OA as the preferred publishing option for the research they have funded. The false promise of Green OA is one which we collectively as a sector need to move away from. As I have said before, it is simply not sustainable, relying as it does on the continuation of library subscription fees. The isolated accepted manuscript (AM) version is not what researchers want to use and leaving the fully-maintained final version of record behind a paywall slows the move to open science, or possibly jeopardizes it with AM’s lower standards, which is the opposite of what science needs.
16 Thoughts on "Guest Post: What Can We Learn from One Million Open Access Articles?"
Congratulations to Springer Nature on achieving the one million article milestone. I will always be grateful to Bev Acreman for bringing me into BioMed Central and for the nine years I enjoyed learning about the world of open access.
As long as Nature is treated as a cash-cow, instead of the most vital part of your portfolio that needs to be open for humanity, decisions will be made to slow down Nature going open and speed up the rest of your portfolio. Have you honestly approached multiple big funds about opening all of Nature?
Thanks for this summary Steven and for Nature’s work to advance OA. I do think you left out a couple of key items from your list of what we’ve learned to date, though, namely: (1) The gold APC solution is utterly unaffordable to researchers in many countries, and the growth of this model is widening the gulf between the haves and have-nots, exchanging paywalls for even more insidious playwalls; (2) Many would argue that transformative agreements are calcifying our use of APCs; (3) The zero embargo component of one-size-fits-all Plan S-type solutions is a nonstarter in much of HSS and exhibits a certain tone-deafness to the diversity of needs and concerns in the research space.
Of course, finding the right way forward is all obviously a work in progress and this community is learning things as we move along. I think we need to put a higher priority on listening to the broad global community, making course corrections to account for the unintended consequences of our current policies, and also continuing to work together as a community on developing the right solutions. To this end, for example, it’s entirely premature to throw Green OA under the bus. For many, this is the way forward, especially as an easy and effective way to share data quickly and widely. We should also learn from the best practices that are out there in data and article sharing—from GenBank to PMC—and in doing so keep an open mind about what’s workable, scalable, sustainable, and so on, rather than moving forward with the mindset that we know the answers and just need to work harder to make the world accept our solutions.
The future for open is filled with potential and promise—on that point, most people in scholarly communication will agree. The best way to realize this potential and promise, though—I would argue the only way—is to truly work together.
“ The zero embargo component of one-size-fits-all Plan S-type solutions is a nonstarter in much of HSS and exhibits a certain tone-deafness to the diversity of needs and concerns in the research space.”
I haven’t come across this argument before and it doesn’t match my experience, which is that zero embargo works better for HSS than pay to publish (noting of course enormous disciplinary differences within HSS). Can you share more?
Sure. There is so much variation by field, region, institution, career stage, etc. It’s entirely plausible that your experience doesn’t align with others (which is kind of the point I’m making). But across the HSS spectrum, what we generally see is that monographs are the gold standard, not journal articles. Noting this, HSS researchers naturally have a strong desire to hang onto their research for years as they polish it into a final format—not release their research notes immediately in CC-BY format. And then there are the details: very little APC funding support in HSS compared to STM, permissions concerns with some fields (especially the visual arts), the absence of mega indexes like PubMedCentral, and more. This isn’t to say that HSS researchers aren’t intrigued by the potential of open (in creating, for instance, better access to digitized vulnerable cultural heritage artifacts), but as a group, the same specific solutions that might work well for one tight and narrow STM field in one region of the world do not translate well at all across all fields in all regions, and especially from STM to HSS.
The false promise of Green OA is one which we collectively as a sector need to move away from. […] The isolated accepted manuscript (AM) version is not what researchers want to use and leaving the fully-maintained final version of record behind a paywall slows the move to open science, or possibly jeopardizes it with AM’s lower standards, which is the opposite of what science needs.
This is exactly why no one trusts PubMed!
I’ve never heard this criticism before. Can you provide a source? Most of what’s in the green repository PubMedCentral (not PubMed, which is different) is the final published version of articles after they come off embargo, regardless of license type (most articles carry a traditional copyright but this is changing fast as most newly published articles have some variation of OA license). In cases there an AAM is included instead (as with a lot of green repositories, in order to comply with funder OA policies), a link to the final published version is included where available, but AAM’s are only a small part of the PMC collection. As far as I know, PMC is an extraordinarily successful and reliable model of green OA, supporting millions of downloads per year and enabling 90+% compliance with the US Public Access program’s open policy.
Exactly. It was a joke; people trust PubMed including the AAMs in PubMed. : )
Ah—good to know. My bad.
You have indeed made the point on many occasions that “green” OA is “not sustainable, relying as it does on the continuation of library subscription fees”.
What you have failed to do however, is show whether the self-archiving of Author Accepted Manuscripts is having any negative impact on subscription revenues.
Although logically you might expect subscription revenue to plummet if many AAMs are free to read, in the real world has this happened and if it has can, can this really be put down to some AAMs being made available?
I appreciate that the Lancet is not published by SN, but I wonder whether subscriptions to this journal have fallen off the proverbial cliff, as so much of the content has been made free to read by the publisher?
Indeed, according to Dimensions, the Lancet has published 1544 articles in 2021, of which 866 (56%) are fee to read. And of this 56%, some 85% has been made OA by Elsevier (what Dimensions calls “bronze OA”), presumably reflecting Lancet’s important role in publishing research related to COVID and Elsevier’s commitment to ensure that this critical research does not remain hidden behind a paywall.
So, though I agree with your assertion that “green OA” relies on the continuation of library subscription fees, this source of funding seems to be highly resilient.
At the end of the day, I believe we both have the same ambition – to ensure that all research is OA. Where we differ however, is that I believe that, at least for now, there are multiple routes to deliver this, including, “green OA”, the costs of which will continue to be met by library subscribers.
Robert, I can’t comment on what is or is not having a commercial impact on other publishers but I would note that the numbers you reference are from 2021 articles – so any impact on subscriptions would be in the coming years and impossible to prove now. Just because subscriptions are still being paid now doesn’t mean that that will continue, especially with all the new tools available to libraries as highlighted by David Crotty in his comment below.
But to be clear my/our concern is not to ring-fence and protect subscription income in isolation but to transition this spend to full, complete and immediate (Gold) OA given its proven benefits and thereby create a sustainable future that enables a more efficient research communication ecosystem.
And what of the unitended consquences? Gold OA will no doubt work well for parts of wealthy Europe, but it’s hardly a stable and sustainable solution with proven benefits for much of the rest of the world—quite the contrary in fact (at least as global Gold is currently envisioned and resourced). Why not have our cake and eat it too? We can work together to reap the benefits of open, and in doing so make sure our solutions work for everyone everywhere.
To back up Robert’s comments, I’m not aware of any library cancelling subscriptions because of a publisher allowing Green OA. In fact, as someone who is actually invovled in subscription decisions concerning read and publish deals at a university library, I can assure you that we actively want to see publishers allowing immediate Green Open Access when we look through the offer proposals they give us, i.e. we are more likely to be well disposed to subscribing when publishers have liberal Green OA polices, not less.
I’ll also add it’s a bit cheeky for a publisher to imply, as Steven’s last paragraph does, that it’s publishers heroically leading the way on OA and that funders and institutions have to join them on this mission – it’s the funders and institutions who are actually pushing for OA. That last paragraph kind of reads as simply “give us more money”.
Unsub (https://unsub.org/), a tool designed to help libraries cancel subscriptions when there are free alternative routes to accessing subscription articles, states that it is in use by over 500 academic libraries. The site is filled with testimonials and press clippings about its use to cancel subscriptions. SPARC has an ongoing tracking page of canceled Big Deal subscriptions (https://bigdeal.sparcopen.org/cancellations).
Rather than claiming that Green OA has no effect whatsoever on subscription decisions (we know that at least in some cases it does, as here’s at least one librarian publicly stating that it is a factor https://scholarlykitchen.sspnet.org/2017/02/21/forbidden-forecast-thinking-open-access-library-subscriptions/ and here are several librarians publicly stating that they won’t pay for things they can get for free https://scholarlykitchen.sspnet.org/2019/10/09/roadblocks-to-better-open-access-models/), I think a better argument is that up until 2019, Green OA had little effect on subscriptions due to the information asymmetry in the market. Publishers often knew how much of their content was available for free, but libraries had no systematic way of compiling that information. Enter Unsub and the playing field has been leveled, and now at least some libraries seem to be actively reducing their holdings where free substitutes can be found.
There’s a lot of confounding factors involved — as you note, a lot of different things go into library spending decisions (for example, economic impacts of the pandemic), but I don’t think it’s unreasonable to state that libraries are looking at free alternatives (including Green OA) when they make those choices. This would make for an interesting study, interviews with the users of Unsub and how it impacts their decisions, because if 500 libraries are using it, they must be doing something with it.
“We and other publishers have a long track record of providing waivers and discounts for many, but this can’t scale sufficiently and we need other solutions that the whole research community will support.” Rather than “waivers and discounts,” how about truly market-sensitive pricing (for transformative agreements). The world’s poorest regions and supporting institutions deserve no less than the richest to share equally in knowledge generation and knowledge consumption. This must proceed independent of, and in defiance of, market logic and business logic.
A laudable milestone, but I find it a little odd, and ahistorical even, that 2005 is selected as the start of SpringerNature’s OA journey, when a pivotal moment was Springer’s acquisition of BioMed Central, which had already been publishing exclusively OA research for several years by 2005.