Editor’s Note: Today’s post is by Byron Russell. Byron is a new face on the HighWire sales team, but not a new one to publishing – bringing over 30 years’ expertise in the industry garnered at Ingenta, Pearson, Berlitz and Macmillan. Here, he reflects on some of the conversations around business models he’s been hearing in the community, and sheds some light on the ins-and-outs of the many acronyms involved.

At the recent OASPA conference in Copenhagen, there was almost universal concern that the transition to open access (OA) publication of research as posited by initiatives such as Plan S and Horizon 2020 was progressing too slowly; impeded by lack of awareness and importance by authors and researchers, and questions from publishers about long term sustainability and the appropriacy of business models such as author processing charges (APCs).

High Angle View Of People forming and arrow on the street

Problems for authors

There is still a significant lack of awareness of what OA is or means in the research community, particularly in the humanities and social sciences (HSS). Even in STEM fields, where awareness is far greater, authors are faced with a dilemma: to publish in a fully open journal, or in an established subscription or hybrid journal which may fall outside the scope of Plan S. Generally, authors want to join a club which they view as better or more impactful than the one they’re in. If they do wish to publish OA in an established hybrid publication with a high Impact Factor (IF) ranking, how to pay APCs – often accepted to be the most common business model in OA publishing — is a particular problem for those without ready access to funding sources.

In some territories, particularly in the Global South, APC agreements are regarded with suspicion and are not widely used. In South Africa, for example (according to UNESCO’s GOAP), OA is not yet integrated into everyday practice for researchers and research administrators and, “article processing charges…are definitely not the way to go for authors in South Africa”.

Problems for publishers

Publishers – especially small university presses and small society publishers — may not have marketing or administrative resources to raise the IF of their OA journal, or implement the wide-reaching changes to their systems to move to an OA model; they may get around this by publishing via larger publishing partners, losing their independence in the process. If they decide to move to OA themselves by going down the APC route, they may not have the staffing available to administer APC payment processes.

Plan S guidance states funders will not fund APCs for hybrid OA journals where the journal is not part of a transformative agreement, unless the author can deposit their article in a recognized repository at time of publication without any embargo. But this green OA approach, while fast-tracking the article’s availability, is essentially is a fudge which may impact negatively on future publisher revenues and is unlikely to be sustainable long-term.

New business models for OA publishing

In September this year Information Power published its Society Publishers Accelerating Open Access and Plan S (or, more conveniently, SPA-OPS) report, which identified and assessed a range of potential models through which learned societies could successfully transition to the requirements of Plan S. HighWire was pleased to be able to contribute logfile data on behalf of two of our society publishers (Portland Press and The European Respiratory Society) so that the usage of OA articles by various organizations across the globe could be analyzed as part of the project. SPA-OPS describes 27 potential business models, of which the APC-based models number just three.

Transformative models, which repurpose existing institutional spend with publishers in order to move them away from paywalled content to an OA model, are widely viewed as the most promising. The key benefit is that they offer a predictable, sustainable funding stream for authors and publishers. There appears to be  considerable support – at least in principle — from library consortia and their members to target existing spend in this way in order to help society publishers make the move; according to SPA-OPS, more than 75% of respondents indicated such financial support was “very likely or likely” to be forthcoming.

As libraries and library consortia provide publishers with most of their subscription income in any case, it makes sense for such funds to be used to “open up” content – and in theory it is far easier for publishers to administer consortium agreements rather than thousands of APC micro-payments. However, it may still be a challenge for smaller independents, which may well not have the staffing resources required. One way around this is for a third party to manage the various actors.

One of the pioneers of transformative agreements (TAs) was Knowledge Unlatched; under then-director Francis Pinter, KU developed a consortium-powered model to enable publishers to flip books to OA. This type of TA – a “choreographed” model, in which an independent body manages the relationships between author, publisher and funder — has several advantages; not least the reduction of the administrative burden on all key players.

There are at least seven different transformative models currently operating globally. There is insufficient space to outline them all here, but it is worth looking at three of the most interesting:

Read and Publish, Publish and Read: Read and publish (RAP) agreements involve a single fee paid by the institution to cover both subscription access and OA publishing for affiliated authors, with the balance tilted toward subscription charges. Publish and read (PAR) agreements allocate the majority of costs toward OA publishing at the article level, with read-only access and perpetual rights to subscription articles included as a benefit of the agreement.

In RAP agreements, an institutional consortium pays a pre-agreed amount for papers published by affiliated authors, and everyone in the library/consortium gets access to the subscription content for no extra cost. The German Projekt Deal agreement with Wiley (the third largest publisher of German research) is an example of RAP. Corresponding authors affiliated with eligible institutions may publish OA under the agreement in Wiley’s subscription and fully OA journals. However, there can be a tricky balancing act involved: those institutions with larger research programs may pay more than those with lighter research outputs, yet both sets of institutions benefit equally. Publishers offering a RAP model include the Royal Society of Chemistry and – somewhat more controversially – Elsevier.

A PAR agreement is similar, but here the amount of money paid to the publisher for the deal is guaranteed, and in exchange authors (or a subset of authors, depending on the arrangement) can publish OA within that publisher’s journals without paying an APC. Publishers offering PAR models include Springer Compact and Oxford University Press.

As should be immediately clear from the examples above, publishers offering PAR and RAP agreements are bigger fish – those with sizable amounts of content which creates negotiating efficiencies for libraries. So as with the Big Deal, opportunities for smaller publishers and societies to move to a PAR or RAP model are limited.

The Collaborative Organization: Several of the speakers at OASPA stressed that “going it alone” was no longer an option, and that collaboration and sharing would be key to a successful transition to OA. COPIM (Community-led Open Publication Infrastructures for Monographs) was set up in June 2019, and — among other work plans — aims to pilot diverse business models for OA book publishing, and so to reduce the dependence of publishers on author-facing book processing charges (BPCs), as well as to better facilitate new OA publishing start-ups. The initiative is attracting a lot of attention, and will be funded by a £2.2 million grant from Research England.

Such collaborative initiatives are surely to be encouraged; as mentioned earlier, the transition to OA is more likely to negatively affect small-to-medium sized publishers; particularly those involved in HSS monograph and book publishing. This problem was widely acknowledged at the recent OASPA conference, where Paul Peters (CEO, Hindawi) suggested an OA switchboard whereby a central agency would choreograph APC payments and automate repository deposition, thereby freeing both funders and smaller publishers from the burden of administration. OASPA itself is looking to kickstart this project with $10k in funding from 10-20 initial stakeholders.

While there is still life in the APC model, the overall view seems to be  that it is a transitional model which is in decline. What is certain is that if real progress is to be made towards OA without significant damage to publisher revenues (and in consequence academic publishing), greater collaboration will be necessary between major stakeholders. Transformative agreements underpinning such collaborations are the most promising business model to-date in the OA ecosystem; the SPA-OPS report’s accompanying toolkit includes an implementation model and specimen contract templates, and the hope and expectation is that societies will use this to offer transformative agreements from next year.

HighWire founding director John Sack is currently planning a workshop to be held early 2020, to identify which models are proving most popular and share knowledge around early business impacts. We welcome input from publishers and societies who would like to be involved.

Discussion

23 Thoughts on "Guest Post — The Future of Open Access Business Models:  APCs Are Not the Only Way"

I think this speaks to the enormous confusion over the structure and nomenclature of these deals. When I spoke to OUP colleagues, they noted that we call our deals “Read and Publish”. Given that each deal by each publisher with each library seems to be fairly unique and bespoke, I have little hope that much clarity is going to emerge around naming conventions.

Actually, that page does contain sufficient info. And, I would agree that all of those arrangements are “read and publish” rather than “publish and read” …. at least per the SK primer (https://scholarlykitchen.sspnet.org/2019/04/23/transformative-agreements/) piece I wrote. Conenient to invoke oneself as the authority though? 🙂 Of course there is no legal authority or universal definitions of these things.

I agree, David. One feels that open access has been (and is being) over-thought, with the net consequence being that there is considerable confusion among smaller publishers and academics as to what is going on, let alone what to do about it – and consequently progress, even in science publishing, is severely hampered. The comments below about what constitutes RAP and PAR are symptomatic of this confusion. As Lisa points out, there is no legal or even universal definition or understanding of many OA policies, and the multiplicity of licence types further complicate matters. For example, I’ve experienced a number of established societies asking if they can make certain articles open access “for a couple of months”, for promotional purposes.

I’ve recently been noticing a number of journals that offer an “open access” option where the article is made freely available but remains under full protection of traditional copyright, which the author assigns to the journal.

Can you clarify what you mean by “traditional copyright”? Perhaps you mean “traditional license / reproduction rights terms”? Open Access has always been about perpetual free-access license terms granted by the copyright holder who continues to actually hold the copyright under law, not an abandonment of copyright ownership itself. The only alternative to “copyright” is “public domain” which no one has been talking about in the OA context.

What I mean is that I’ve recently been coming across journals that have an “open access” option whereby the author pays an APC and the paper is made freely available for all to read, but the author must transfer copyright to the journal and the article is published and marked as “copyright the journal/society/publisher”, without any of the reuse rights one would expect from an open access paper. Yes, under the Berne Convention all papers are covered under copyright, but the BOAI and other common definitions of OA require that the paper is then licensed for reuse via a Creative Commons license. Here, no CC license is applied and reuse is limited to what falls under Fair Use.

Wow, that sounds suspiciously close to fraud. These journals need to be “named and shamed”!

Thank you Byron for this overview. Perhaps another way forward is to fund OA through platform fees instead of article payments? There are a few journals and societies that seem to work well on the diamond OA model.

The OA switchboard: I actually tried to set up a similar national initiative a few years ago. I had a decent number of participating (small) publishers and university presses but both the national funder and the university consortium were not interested: I was told the number of annual publications within the switchboard network was not enough to justify their efforts in getting any sort of deal going. It seems that a switchboard need to represents a significant market share in order to be taken seriously, which certainly is a challenge when dealing with small publishers.

The American Thoracic Society would like to attend John Sack’s workshop in 2020 and/or receive more information.

Thanks for this overview, Byron. You mention that APC models are not widely used in the Global South, but there are also challenges with transformative agreements if the institution (or consortium) negotiating receives many of their subscriptions through deeply discounted arrangements, or even completely free-of-charge through a service like Research4Life. Their negotiating power will be minimal and yet their researchers have just as great a need to publish (perhaps even greater) as their peers in the North.

Helpfully, Information Power is also considering one or more pilots with consortia in Africa to explore how they too can benefit from a shift to Open Access; after all, we mustn’t end up with a research communication system that is actually less equitable than the one we have today!

A couple of comments:

1) “Plan S guidance states funders will not fund APCs for hybrid OA journals where the journal is not part of a transformative agreement, unless the author can deposit their article in a recognized repository at time of publication without any embargo.”

This is not correct – depositing articles published in hybrid journals not part of transformative arrangements makes them (or the author) compliant with Plan S (provided conditions for embargo, license and copyright retention are met), but funders will not fund APCs for such articles.

2) While Byron Russell characterizes green OA as ‘a fudge’ in this piece, the HighWire white paper on Plan S options has this to say on the expectation of declining subscriptions when AAMs are shared immediately:

“Most publishers have not considered this because of the assumption that free versions of many articles would cause significant subscription loss. However, four HighWire society-based publishers now publish author manuscripts upon acceptance, and make them freely accessible. Three of these publishers (with eight journals among them) make these open versions persistently available even after the final version is published in an issue. Two have been doing this for nearly twenty years and do not believe this practice has contributed to subscription decline. And these publishers have been doing this for all original research, not just for certain funders.”
(https://www.highwirepress.com/app/uploads/2019/02/HighWire-PlanS-Whitepaper.pdf#page=12)

3) “Publishers – especially small university presses and small society publishers — may not have marketing or administrative resources to raise the IF of their OA journal” – fortunately, there is a growing commitment from insitutions and funders, including Plan S funders, to move away from journal metrics as IF, as evaluation criterion for research output.

Hi Bianca,

On point 1, yes, this is inelegantly phrased. Worth also pointing out (at least in my understanding) that authors can publish OA in hybrid journals and remain compliant with Plan S, but Plan S will not pay the APC costs for that publication. Assuming the author has an alternative set of funds, or collaborators who are funded elsewhere that can cover costs, this is another route to compliance allowing authors to publish in hybrid journals.

On point 2, I can anecdotally tell you that while many publishers have been able to get by under the radar for many years with very liberal Green and Bronze OA policies, those days are increasingly over. This gets back to my recent post about free riders, and how no library has such a high level of funding that they can afford to pay for things they can receive for free (https://scholarlykitchen.sspnet.org/2019/10/09/roadblocks-to-better-open-access-models/). New tools like the forthcoming Unpaywall Journals are explicitly designed to help libraries optimize their spending and cancel subscriptions to journals where a significant amount of the material is freely available. I know that in my experience, I’m increasingly seeing more and more pushback from libraries because of my employer’s mission-driven approach to trying to make as much material freely available as possible. This is an unfortunate consequence of libraries being underfunded — essentially, the publishers who are trying to do the right thing are being punished by the market, and I suspect that many will start rolling back all of the free access they’ve tried so hard to grant.

Hi David,

1) Yes, that’s indeed what I tried to say.

2) I guess time will tell. So far, I am not aware of any proven decline in subscriptions due to green OA policies, even liberal ones – if there are, I’d be interested to know. Whether that projects forward will need to be proven as well. While I agree personally that relying on green OA that is dependent on other parties still paying for subscriptions for that content is not the way forward (unlike e.g. contributing to shared infrastructure for diamond OA), I think there are at least two reasons journals would not need to see a demise of subscriptions with green OA for accepted manuscripts: one is when the added value of the published version is perceived as high enough, the other is when the value of curation and contextualization (through non-research article content) is perceived as high enough.

It’s not only cancellations; it’s also downward pressure on pricing. As for “the value of curation and contextualization,” if one has to invest in that, why invest in the creation of the content in the first place, if you can get it for free everywhere?

What proof would you consider adequate? I would think that the increasing number of Big Deal cancellations would be evidence (https://sparcopen.org/our-work/big-deal-cancellation-tracking/ and https://scholarlykitchen.sspnet.org/2017/05/01/wolf-finally-arrives-big-deal-cancelations-north-american-libraries/). I would think that the preponderance of universities cancelling their Elsevier agreements, a concept until recently unthinkable, would be evidence enough, particularly because we have heard repeatedly from those libraries that their researchers will be fine because they can get all they need through other channels. Does that not indicate that those other channels are harming subscription sales? I can tell you anecdotally that I regularly hear from journal sales reps that there is increasing pushback on the publishers that make a lot of content freely available — why should I pay for something that is largely free? As more sophisticated tools meant to explicitly cause these cancellations become available, I expect this to increase dramatically.

I do understand the libraries’ behavior here — they are in incredibly difficult positions and have to make very hard decisions about how to spend their limited budgets, so yes, it makes total sense to stop spending on things you can get for free. But the net effect is punishing the publishers that are acting as the best members of the community and offering liberal access policies. This is going to result in a reversal of those policies and a locking down of content that was previously voluntarily made freely available.

The view may be very different from different kinds of institutions, eg small comprehensive public ones versus very large ARL-type major-research ones. But I keep asking my colleagues internally and elsewhere to consider the question, “what percent of a hybrid journal (or entire Big Deal package) has to be OA before we consider cancelling it because of that? 20%? 50%? 80%?” I keep getting “crickets” to that question. No one on the library side that I talk to is ready to face the actual quant question yet, and since subscriptions are binary (you either subscribe or not), that means no cancellations are happening because of OA. I’ve seen lots of bigger libraries cancelling Big Deals because of budget problems, but not because of their OA content. However, with COUNTER CoP5 we are finally starting to get data on what portion of a Big Deal’s usage is on the OA content within that deal, so maybe after we have a year’s worth of that data (it started Jan 2019 for most publishers) we will start to see librarians more seriously engage the percentage-to-cancel question.

You’re right — librarians are cancelling subscriptions (and Big Deals) largely for budgetary reasons, but OA provides them a mechanism to find journals that can be cut with the least amount of pain to their users. I can certify that an increasing number of librarians are bringing up data from Unpaywall in their negotiations with publishers (both large and small), and asking why they should continue to pay for material that a publisher is voluntarily making freely available to the world. It absolutely makes sense, given the tough decisions that librarians have to make. However, the end result is that the publishers that are most liberal with making material free (something most librarians would support) are the ones who are being/going to be punished, and those policies are likely to be rolled back, and more and more material will go back behind paywalls. The new COUNTER statistics (and new analysis tools being released) are going to exacerbate this problem.

Joseph – I’m fresh from a workshop this week on research assessment reform put on by DORA and HHMI. Can you elaborate on why you believe movement away form the JIF is unfortunate?

Leave a Comment