One of the most popular aphorisms invoked in the ongoing discussion about the future of scholarly communication is an argument that making a global transition to open access (OA) will require no new funding – because, after all, the necessary money is “already in the system” and needs only to be directed away from subscription fees, commercial publishers’ profits, and pay-to-publish charges (which make content free but impose access tolls on publishing services), and instead applied directly to the costs of open publishing, thus enabling an effectively cost-neutral transition from a hybrid pay- and open-access landscape to a fully OA one.
Ten years ago, Martin Eve presented a trenchant critique of this line of reasoning in the UKSG journal Insights; he reprised it five years later on his blog. In the wake of growing systemwide enthusiasm for Diamond and subscribe-to-open (S2O) open-publishing solutions, I’d like to examine some similar ideas here, looking from a slightly more granular perspective at the dynamics of budget management in colleges and research universities (where much of the money and the great majority of the research outputs “in the system” come from).
The fundamental problem with the argument that “the money is already in the system” is that it’s built on a false assumption: that money currently “in the system” (i.e., the scholarly communication economy) has been directed into that system for the purpose of supporting scholarly publishing, and will stay there as long as it continues to serve that purpose.
This is by no means the case. Consider an institution like my own – a large research library with a healthy but not exceptionally rich collections budget – which puts roughly $12 million cash into the scholcomm economy every year; this money pays for journal and database subscriptions and one-time purchases of books, online archives, and other content products. This money does not belong to the library; it belongs to our sponsoring university, and is allocated to the library each year for the express purpose of buying access to content – content to which the university community would not have access if it were not paid for. Although these institutional funds do, in practice, support scholarly publishing, that support is a byproduct of their actual purpose — which is to secure access to content for the university community.
Now imagine that we switch to a system that no longer requires my institution to pay for access to content, but does some combination of the following instead:
- Requires my institution’s researchers to pay for publishing services (the article processing charge, or APC, model)
- Requests that my institution continue paying “subscription” fees to support ongoing publication of journals, the content of which has become OA (S2O)
- Invites my institution to directly underwrite the publication of open journals and/or books by means of direct subvention, often framed as “memberships” (Diamond)
As Eve and others have noted, such a system will tend to put a much greater financial burden on large institutions that produce a lot of research; a college or university that is significantly less research-productive will, even if it assumes the full cost of its researchers’ publications, end up putting far less money “into the system” than it did when it had to pay for access to content rather than access to publishing services.
But there’s a deeper problem with the idea that existing money would be likely to remain “in the system” in such a scenario. The problem is that when an institution is no longer required to pay for access to content, the money it formerly used for that purpose is now available to meet other needs – and all colleges and universities have far more needs to meet than they have resources available. Many — most — of those needs will be felt more urgently by the institution than the need to help underwrite a global shift to open scholarship.
I regularly tell my library colleagues that when a campus is under space pressure, the library can no more expect its underutilized space to remain underutilized than you can expect a dry sponge to remain dry when you stick it in a bucket of water. The same is true of library budgets. It could be true that if all of the library budgets in the world were flipped from paying conventional subscriptions to underwriting OA publishing (whether through S2O “subscriptions” or Diamond subventions), the scholarly publishing enterprise could be fully supported on an OA basis.
What has kept much of the money “in the system” is the necessity of paying for content. When you change that aspect of the system, you change the likelihood that money will stay inside it.
But the fact that the money is currently available “in the system” does not imply that if that money were no longer required for buying access, it would then remain in the system and be available to the library for underwriting open publishing. Once the money that was previously put into the system for the purpose of buying essential content is no longer needed for that purpose, it becomes a fungible resource for which any number of other university needs and priorities will compete – and at least some of those needs will be seen, by the administrators who allocate the money, as more important than supporting open publishing.
To put it more simply: What has kept much of the money “in the system” is the necessity of paying for content. When you change that aspect of the system, you change the likelihood that money will stay inside it. Of course, if the system were to flip universally from pay-for-access to APC-funded OA, money would likely be much more effectively contained, because publishing is just as essential to an institution’s research mission as access to content is. But in light of the massive inequities this shift would produce, and as more and more people have recognized that the APC model is simply another form of toll access, it is becoming increasingly politically unpopular and the current momentum seems to be in the direction of Diamond and S2O. S2O and Diamond models have the potential to spread costs a bit more evenly — but rely on voluntary participation by institutions that are constantly trying to balance competing needs with strictly limited resources.
All of this has serious implications for the long-term sustainability of publishing programs supported by S2O and Diamond funding models. To the degree that these models become prevalent, the scholarly communication ecosystem changes in fundamental ways and will no longer attract or contain institutional money as effectively as it did when spending money was the only way to secure access to essential content. Under the old system, failing to allocate large budgets to the library meant the loss of access and the hamstringing of the university’s research and teaching missions. Under a global (or even predominantly) OA system, an institution can choose to do other things with the money that the old system forced it to allocate to the library. The purpose of the allocation thus becomes supporting OA publishers (an optional investment), rather than buying access to content (an essential one). Supporting OA publishers may be a noble purpose, but it joins a crowded field of other noble purposes that are all competing for the university’s budget.
To illustrate, let’s consider just a few of the infinite pairings of choices that a university might be faced with once it’s no longer required to pay for access to content. Which is more important to the university’s mission:
- Providing scholarships to five first-generation students, or helping to ensure that five toll-access sociology journals can flip to OA?
- Replacing outdated equipment in the physics lab, or helping to underwrite the publication of 10 open physics textbooks?
- Giving staff employees a cost-of-living increase, or supporting the open publication of a suite of economics journals?
The institution’s best, wisest, and most mission-aligned option in such scenarios will not always be the scholcomm one. And every time the scholcomm option loses out in one of these allocation choices, more money leaks out of the scholcomm system.
Supporting OA publishers may be a noble purpose, but it joins a crowded field of other noble purposes that are all competing for the university’s budget.
This is why I constantly (and generally without much success) warn publishers to stay away from S2O. I’m very concerned that many are being lured into this model by the very real and very sincere enthusiasm of librarians – who appear to be, but are not in fact, the ultimate controllers of collections budgets. S2O sounds great to us in libraries; it allows us to keep our well-established workflows (and if there’s one thing we librarians love, it’s our well-established workflows) and allows us to contribute to open publishing, which we absolutely love to do. But we will only have the ability to continue doing that if it is seen by our sponsoring institutions as the best possible use of the money they allocate to us. And in many budget-choice scenarios, it will not, in fact, represent the best possible use. S2O and Diamond models that offer premium benefits of some kind (enhanced access to archival content, say, or exclusive access to research tools) will be more likely to retain their paying customers — as long, of course, as those premium offerings provide the right kinds of benefits to those customers’ institutions, in amounts that are felt as proportional to the cost.
Accordingly, the other advice I’ve been giving lately has been to my colleagues in libraries: if you’re going to participate in an S2O program or contribute to a Diamond publishing project, make sure your institution is supportive. Have this conversation up front, before making any commitment to a publisher. Your institution may well be supportive (though how long that will remain the case will depend greatly on other factors). But if your institution finds out that it has been underwriting open publishing for some period of time when it thought it was buying access to content, you run the serious risk of losing your sponsor’s trust. And that trust is worth more than money.
Discussion
90 Thoughts on "Subscribe-to-Open Is Doomed. Here’s Why."
Hi Rick. I’m just an onlooker here (from the UK) but you seem to assume that existing funding needs to support OA AND the existing system? It doesn’t: the system we need is one whereby universities publish their own employees’ work.
How they do it – more economically via joint ventures with other universities (cf. https://universitypress.whiterose.ac.uk for an example), or perhaps online only, is a different matter. But the system we need is one whereby they are funded to do so, including of course appropriate peer review, etc. to ensure adequate oversight and academic excellence. They are already funded to produce research, after all, and what use is research without publishing?
Hi, David —
What you’re proposing is commonly referred to as “Diamond” OA — a system whereby research institutions directly underwrite the open publication of their researchers’ work. This can take several different forms: S2O is one; directly underwriting third-party publishing programs outside of the subscription model is another; institutional in-sourcing (whereby the university simply takes on the role of publisher) is another. All such models involve the redirection of institutional resources away from a mission-critical need that can’t be outsourced (providing local access to scholarly content) to one that can be (publishing services). Nothing is stopping institutions from adopting the latter approach, of course, except institutional priorities and mission alignment.
I think all of us likely agree that publishing research is essential. The difficulty is figuring out the most effective, efficient, fair, and mission-aligned way to support publishing. For all the reasons I’ve outlined in this piece, I think the S2O model in particular has a dim future.
Thanks for commenting – 1) outsourcing means directing funds to third-party profit; 2) what I’m proposing is what many universities (Cambridge, Oxford, Liverpool, Stockholm, Michigan …) used to do, but in many case have now outsourced it.
I don’t think “outsourcing” necessarily implies a for-profit third-party service provider. If I outsource a project to a nonprofit service provider, that project is every bit as outsourced as it would be if I gave it to a for-profit one.
Thanks Rick. I have never personally followed the logic of the ‘there is enough money in the system already’ sentiment. The money is there, but it will only shift to a different purpose (assumed to be more virtuous and enabling OA) if University X decides to make that change. As you say, University X may have other uses in mind.
You are being harsh on the S2O proposition. Many (most?) S2O offerings do contain the ‘premium benefits’ you mention, through archival access or something else. The transition is also very simple in budgetary terms: a subscription becomes support for an S2O journal, usually with no change in pricing. So there is both a reason for continued library support (an archive) and a very simple change needed to enable S2O. That is different, for example, to Diamond OA, where a completely new funding mechanism needs to be created.
I would take your point further as well. At some stage, all agreements with be OA. We should bring interested stakeholders together now to decide how we attribute revenue to publications when that happens. If everything is OA, the ‘global rider’ problem becomes universal. It would be sensible to form compacts by which research publishing can be properly funded, on clearly understood principles, with the needs of libraries fully taken into account, before we reach that point.
Hi, Rod —
Many (most?) S2O offerings do contain the ‘premium benefits’ you mention, through archival access or something else.
And as I said in my post, to the degree that such benefits a) exist and b) provide a mission-critical institutional benefit that is c) proportionate to the ongoing cost, the institution will be more likely to be happy about continuing to pay a “subscription” fee.
The transition is also very simple in budgetary terms: a subscription becomes support for an S2O journal, usually with no change in pricing.
Simple? Absolutely. But for each paying institution, that structural/procedural simplicity masks a radical shift that matters very much: the shift from paying for a mission-critical service (providing access to scholarly content) to a service that is not mission-critical (underwriting the provision of free access to scholarly content to everyone in the world). Librarians who hide or obfuscate this shift in their dealings with the institutions that fund them run the serious risk of losing their institutions’ trust and, eventually, their funding.
At some stage, all agreements with be OA.
I think that’s a very risky assumption. Right now, 25 years after the OA movement took off, the majority of scholarly content is still published on a toll-access basis; furthermore, the market share of toll-access publishing is growing, while the market share of Gold OA appears to be leveling off and Bronze and Green are both shrinking. My suspicion — and, frankly, my hope — is that we’re going to have a healthy diversity of scholarly publishing models, including toll access, into the indefinite future. I would urge libraries, funders, institutions, and publishers to be very careful about buying into predictions of a universally OA future — such predictions are, I believe, the stuff of advocacy and lobbying rather than serious analysis.
“Simple? Absolutely. But for each paying institution, that structural/procedural simplicity masks a radical shift that matters very much: the shift from paying for a mission-critical service (providing access to scholarly content) to a service that is not mission-critical (underwriting the provision of free access to scholarly content to everyone in the world).”
S2O looks different to me. The library payment performs multiple roles with S2O: underwriting provision of free access for everyone (as you say), providing access to scholarly content (still true, as the journals will revert to subscriptions and closed access if the support disappears), archival file access and contributing to the costs of publishing research. Those factors all apply to S2O and they are essential to how we publish research today.
Any given library may focus its attention on only one or a few of those dimensions of S2O, but as Luke Vilelle states below, different libraries have different goals. This difference does not invalidate S2O. On the contrary, it makes clear that S2O is a step forward from subscriptions, providing the benefits they give and significantly more besides for the same expenditure.
S2O looks different to me. The library payment performs multiple roles with S2O: underwriting provision of free access for everyone (as you say), providing access to scholarly content (still true, as the journals will revert to subscriptions and closed access if the support disappears) […]
But, importantly, the risk any individual institution faces of losing access if the journal reverts to a subscription model is functionally nonexistent. Either the gamble pays off (other institutions for whom supporting S2O is actually mission-critical continue their support at a sufficient level) and the content remains free, or the content reverts to subscription-only and the institution simply goes back to the status quo ante, having lost nothing. This fact is one of many that challenge the sustainability of S2O.
[…] archival file access and contributing to the costs of publishing research.
As I said previously: if archival file access is available only to S2O “subscribers,” and if that access is worth the cost of continuing to pay the full former price of the subscription, then libraries are more likely to continue paying and should have no problem justifying it to their sponsoring institutions. Of course, in that case, the journal is continuing to charge for access — just not to current and future content. (Is that really OA?) Over time, of course, the value proposition of access to the archive is likely to decline (at a rate that depends, among other things, on the nature of the content).
… different libraries have different goals. This difference does not invalidate S2O.
I don’t believe anyone is arguing that S2O is “invalid.” The question is whether it’s sustainable. And the answer to that question depends much less on libraries’ goals than on the goals and missions of the institutions that provide them with collection funding.
On the contrary, it makes clear that S2O is a step forward from subscriptions, providing the benefits they give and significantly more besides for the same expenditure.
But this is my central point: S2O does not, in fact provide the same benefits as a real subscription, and depending on the specifics of the model, the benefits it does provide may or may not accrue to the institution based on whether it pays. No amount of rhetorical reframing changes this reality, and it’s the reality that will ultimately shape the flow of funding.
Most librarians I know aren’t looking to manage their collections like a stock portfolio, where they divest Publisher X because they have moved to an S2O model and are now free to access, then buy them back the next year if S2O ends. There is a lot of trust built into how libraries manage acquisitions. Librarians trust that a journal that published good content this year will do the same again next year, and subscribe in advance. Publishers assume that they can continue publication next year because most libraries will continue to support them. This time shifting in the decision-making process has always been present. S2O simply moves the time shift from the year before to the year after.
Of course, extreme financial pressure can break these bonds of trust. But in normal operations, the key question is not whether a librarian can shave a few dollars off an acquisition budget by cancelling a handful of journals. It is whether the publisher is providing value for money on publications that aid the faculty of the host institution. If S2O journals demonstrate that they are good investment on a range of library acquisition criteria, they will do just fine.
Most librarians I know aren’t looking to manage their collections like a stock portfolio, where they divest Publisher X because they have moved to an S2O model and are now free to access, then buy them back the next year if S2O ends.
Calling it “manag(ing) their collections like a stock portfolio” seems like a radically inaccurate and unfair way of characterizing the work of librarians who are, invariably, trying extremely hard to balance the needs of a highly diverse constellation of constituencies in a context of extremely limited resources. The reality is that this is what they do because they have no choice; they’re not managing investments for the purpose of maximizing shareholder return; they’re trying to buy products and services that will best support their institutional mission. Even within the context of an allocated collections budget, the decisions we have to make are very difficult, and I can promise you that — especially at institutions with more limited budgets — librarians are asking themselves very tough questions about what to spend money on.
Librarians trust that a journal that published good content this year will do the same again next year, and subscribe in advance.
Unless the publisher has signaled that next year’s content is expected to be available at no charge, in which case the librarian will now have to make a potentially very difficult decision: continue paying for content that no longer costs money to access, or decline to pay, and use the money for other urgent purposes. “Value for money” is only a consideration when you have to pay the money to get the value. Of course, by declining to pay, an individual library is (slightly) increasing the likelihood that the content won’t actually become free. If that happens, they can always start paying again. Some publishers (like Annual Reviews, I believe) build a penalty into such scenarios to increase libraries’ incentive to keep paying. That’s the “stick” version of the “carrot” approach of having premium offerings for those who keep paying.
To me, we are having a glass half full/half empty discussion here.
Your view is that it is rational for librarians to cancel free resources, with some caveats. And then possibly restore them the following year if they cease to be OA. In my interactions with librarians over 25 years working in STM publishing, this isn’t a behaviour pattern I have often come across.
My view is that the glass is half full. S2O has most of the characteristics of a traditional subscription and the additional virtue that it makes content available on a permanent OA basis. For me, that is a big step forward – and it works very well for libraries who have an explicit mission to support open resources. It isn’t the end of the OA journey (a thought behind part of your critique, I think), but it represents significant progress. There is now a challenge to embed it more deeply and make sure it is sustainable, and all S2O publishers recognise that.
I tend to see the optimistic elements of the S2O model first and foremost – and my mind then runs on to the practical question of how do we address some of the vulnerabilities you identify and make them less problematic.
Your view is that it is rational for librarians to cancel free resources, with some caveats. And then possibly restore them the following year if they cease to be OA. In my interactions with librarians over 25 years working in STM publishing, this isn’t a behaviour pattern I have often come across.
I do, in fact, believe that it would be less than fully rational for a library to say “We’d better keep paying for content that is available at no charge because if we don’t keep paying for it we’ll run the risk of having to start paying for it again.” But that’s not the main point I’m making with this post. My primary point is that academic libraries are not independent organizations with their own money that we’re free to use in any way we wish. We are agents of universities that vouchsafe university funds to the library for specific purposes — in this case, we’re talking about collections budgets that are allocated to the library for the purpose of securing access to scholarly content. When those funds are no longer needed to secure access to that content, there’s a very good chance that they’ll be redirected to other needs, many of which are at least as worthy as (and may even advance the institutional mission more effectively than) the cause of underwriting open publishing.
My view is that the glass is half full. S2O has most of the characteristics of a traditional subscription and the additional virtue that it makes content available on a permanent OA basis.
But the big problem with S2O is that it abandons the most fundamental characteristic of a traditional subscription, and the one most likely to retain paying customers: the fact that paying a subscription fee is the mechanism by which the customer retains access to the subscribed content. The other characteristics (workflow mechanisms, invoice management, regular recurring payment, etc.) are superficial and far less dispositive. Libraries don’t maintain subscriptions because maintaining subscriptions is easy and convenient; we maintain subscriptions because the institutions we serve need access to the content. When paying a subscription fee is no longer required for access, the entire calculus changes. And this is why the past behavior of librarians is a poor index for what’s likely to happen in the future: S2O is a new business model and creates a new decision-making calculus both for us and for the institutions whose money we’re spending.
As I say below in response to another comment: I believe S2O is doomed because I see no reason to believe that underwriting open access publishing will rise to the top of the fiscal priority list for a critical mass of academic institutions, over time.
Hi Rick
You present an interesting case . However, ‘or the content reverts to subscription-only and the institution simply goes back to the status quo ante, having lost nothing. ‘- is part of the argument that I am unconvinced by. There will be the danger that of the S2O model works one year, and an institution stops subscribing and just uses the ‘free’ content, and then it’s back to a standard subscription model the following year (because not enough institutions sign up), that the institution will have lost access to the money the last time it paid for the resource 2 years previously. That may even happen in the year the institution is not paying for the subscription e.g. if the institution thinks it won’t need to pay for access then suddenly discovers during that year that it has to.
Hi, Elizabeth —
Thanks for your comment. But I’m not sure I follow, and I may be misunderstanding your point. If a library pays for access in Year 1, and then declines to pay when payment is no longer required for access in Year 2, and then has to begin paying again in Year 3 (because the journal doesn’t meet the contribution threshold), then in what way has that library lost the money it paid in Year 1?
I think the idea is that you pay a subscription to an S2O journal, then you decide not to, and you either spend that money on something else essential or your institution takes that money out of your budget. Then the journal flips back and you no longer have the money to resubscribe.
David, you’re right — that is indeed a risk. Of course, the upside of that scenario is that you’ve now subscribed to something else that you didn’t have access to before. The library is then left needing to decide whether to stick with the new thing or go back to the old thing — or, in other words, needing to decide between two desirable things, not being able to afford both.
In other words, effectively the status quo ante.
The risk of having institutional budget redirected doesn’t arise when the library cancels one subscription and starts another; it arises when the institution realizes that the library is paying for free content.
For the journal though, this makes S2O feel something more like a ratchet mechanism. We know that library budgets are limited, which is why it is very difficult to sell a new subscription journal into the market (and hence, most new launches are OA journals). For an S2O journal, if you have to flip back to subscriptions, and the library has diverted that money elsewhere, you’re now in the position of selling them essentially a new subscription. Which means they’re going to have to find new money (unlikely) or cancel something else they’re doing to buy back your journal.
Yes, I think that’s right. It’s yet another reason why journal publishers should do very serious due-diligence work — rather than just listening to cheerleaders and advocates — before deciding to take the plunge into S2O.
Incidentally, one of the many long-term challenges that will be faced by publishers who embrace S2O will be the challenge of starting a new journal.
Rick, to some extent I agree with you about being up front with our institutions, but I’ve found it very hard to engage with my administration on how our library funds are supporting OA. A memo that I wrote to our Provost’s office last year detailing our OA investments went unanswered. Maybe the problem is that the OA landscape is relatively complicated and not easily explained to administrators; in fact the whole universe of resource acquisition has become so complicated that for me to be completely up front about how we spend our budget would require time that our high-level administrators usually don’t have (not that I don’t try). I can see that a time may come when they would intervene against spending money on OA, but I feel like we’re a long way away from that right now. So I guess I’m not ready to call S2O doomed– there’s still time to figure out how we might make it work in ways that benefit institutions directly. I also remind myself that libraries are very good at forming cooperative alliances; not that we always do it perfectly, but we’re used to putting cooperative structures in place, and where cooperation is essential we have been able to retain funding for work that isn’t completely self-interested but contributes to the good of the whole. So overall I guess I’m more hopeful than you are about the future evolution of sustainable OA models.
Hi, Anne —
I’ve found it very hard to engage with my administration on how our library funds are supporting OA. A memo that I wrote to our Provost’s office last year detailing our OA investments went unanswered.
Do you report to the provost? If so, your meetings with her or him would be a great opportunity to follow up on that memo and take her/his temperature on the library’s current OA support strategies. But it’s also possible that the lack of response signals approval of what you’re doing — the provost’s posture may be “Thanks for the heads-up; we’re good.” But it would be good to clarify that if you can, especially if you plan to increase the library’s redirection of content-access budget towards OA underwriting.
Rick, I report to the provost, and have 30 minutes a month with them. Because I’ve had four provosts in my three years here, much of that 30 minutes is spent just getting them up to speed on all the things that the library does– it’s not nearly enough time for sustained engagement on this topic, or really any topic. It seems that that level of turnover isn’t unusual (many deans & directors across our PALCI consortium report similar levels of turnover in their administrations). I guess my point is just that I don’t think we’re close to having the plugged pulled on our OA investments. The chaos in higher ed isn’t leaving much time for that level of scrutiny.
Fair enough! Every library leader needs to evaluate that situation in their own institution and take the course that makes the most sense to them.
Anne – if I may ask, roughly what percentage of your Provost’s annual budget do you spend? If you have delegated authority to spend 5% of the university’s funding, I think you ought to be getting more than 20 minutes!
And do you think it’s representative of similar libraries?
5% is off by at least an order of magnitude. https://scholarlykitchen.sspnet.org/2024/04/09/guest-post-speak-finance-gain-university-support-for-open-scholarship/
Well, I guessed it might be!
Sorry, I missed David’s question yesterday. Regardless of how much of the budget I oversee, I agree that I should have more time with the Provost– 30 minutes a month is absolutely not enough for all of the things that I need to cover with them, which include, at the moment, a major library architectural planning project. Our Provost has a huge portfolio of direct reports and there’s been a lot of competing priorities for the past few years. Again the point I was making is that I hear from other deans and directors who are in the same boat that I am in terms of administrative turnover and stress, and so it can be hard to get these conversations started.
I will have a lot more confidence in S2O as a model when I see a publisher “flip back” a journal that fails to meet its subscription threshold. So far I see a lot of organizations putting out press releases saying that they’re not bringing in as much as they think they need to be sustainable, but that they’re going to go ahead and keep the journals OA anyway. Right now it appears to be something of a ratchet mechanism, with no going back once the journal flips.
David, I’m curious — in your experience, when publishers say both that a) it’s become clear that their S2O program is fiscally unsustainable and b) they’re going to keep the journals going on an OA basis anyway, do they say how they’re going to support the ongoing OA provision? Are they just going with APCs?
I’ve seen a mix of responses. In some cases, the journals program has moved from an S2O model to a “pure publish” model, where subscribing institutions see no publication charges and non-subscribers pay an APC (IWA did this as described here: https://www.ce-strategy.com/the-brief/third-time-is-a-charm/#2). EDP suggests that they’re trying to reduce costs to make the lower earnings from S2O viable (https://www.edpsciences.org/en/news-highlights/2994-open-access-and-transparency-edp-sciences-releases-2024-transparency-report-for-mathematics-journals) as well as deciding not to move journals planned to go S2O to the model (https://www.edpsciences.org/en/news-highlights/2884-edp-sciences-decides-against-open-access-transition-under-s2o-for-radioprotection-in-2023). Others just seem to be a shrug of the shoulders and a strategy of hoping more money will turn up somehow.
Radioprotection was under the S2O model. The S2O model “worked” in that the threshhold wasn’t reached and so it wasn’t opened. That doesn’t make the journal not operating under S2O that year. That makes the outcome not open. Closed is also an S2O outcome.
Thanks for the correction. Are you aware of any journals that have made the flip back from S2O open to S2O closed?
I am not. So far, to my knowledge, any journal that made S2O-open has continued to meet the threshold for open in subsequent years. (BTW, I would not call the IWA model “pure publish” but I’ll leave that for another day.)
Thank you for this interesting post that echoes skepticism I’ve heard from others about open access models. I’d like to interrogate one of the underlying premises of this post, that library collection funds are “allocated to the library each year for the express purpose of buying access to content.” I’m not sure this is entirely true, and I’d suggest we take a step back for a broader perspective. Your library might be different, but nowhere in our library’s mission or fundamental commitments do we say anything about buying access to content. We “connect” our community with resources, by committing to “collect, support, organize, and enrich a valuable and diverse set of resources…” Collection funds are allocated to the library to ensure that our community has access to the resources they need to support their scholarship and creative work, giving the library discretion to identify how best to make those resources available.
How we ensure our community has access to those resources has changed significantly over the years. Here’s one example particularly pertinent to this post. Somebody 30 years ago might have written that collection funds are “allocated to the library each year for the express purpose of buying content.” That has changed radically in the online environment over the past three decades, and now we talk about “buying access to content” instead of buying the content itself. But the underlying goal remains the same: we want to ensure our community has access to the resources they need to support their scholarship and creative work. Open access models, including subscribe to open, enable us to explore how we might provide new routes for access that address long-standing challenges and inequities we face in the scholarly communication ecosystem.
Hi, Luke —
I’d like to interrogate one of the underlying premises of this post, that library collection funds are “allocated to the library each year for the express purpose of buying access to content.” I’m not sure this is entirely true, and I’d suggest we take a step back for a broader perspective.
I’m not sure what your position is in your library, so forgive me if I’m telling you something you already know. But in every university of which I’m aware (including the several for which I’ve worked over the course of my career), libraries are given specific budget allocations each year for either two or three defined functions: operations, collections, and (in most cases) personnel. Furthermore, library directors make arguments every year for collections-budget increases based significantly on the increasing cost of access to content in different subject areas. Now, there may well be institutions — particularly outside of the US — where this is not the case, but I’m confident in saying that it’s typical here. Even where the structure of the budget allocation may be different, I’m even more confident in saying that the institutions that sponsor libraries allocate large amounts of money to them with the express understanding that it will be spent to purchase content (or access to it; see below), and that the library’s unilateral redirection of that money away from purchasing content and towards underwriting OA publishing would constitute a diversion from the sponsoring institution’s intended purposes. That’s why it’s wise to counsel with the institution before undertaking that redirection.
Somebody 30 years ago might have written that collection funds are “allocated to the library each year for the express purpose of buying content.” That has changed radically in the online environment over the past three decades, and now we talk about “buying access to content” instead of buying the content itself.
Agreed. When I say “buying content” in the post, that should be taken as shorthand for “purchasing copies of physical and digital documents and purchasing either ongoing or temporary access to online content.”
Sorry for not identifying myself and my perspective more clearly. I’ve been a library director at a liberal arts college (Hollins University) for the past 13 years. In my experience, it’s too great a leap to say that the university has an “express understanding” that the library’s collection budgets will be spent to purchase content or access to content. That’s not what we tell the university in our library’s mission statement. That’s not how I talk about our collection funds with my provost. Our primary goal is to use those dollars to advance the university’s priorities. Hollins aims for academic excellence, so that’s why our library is going to collect, support, organize, and enrich a valuable and diverse set of resources. That often means traditional purchases or subscriptions, but other times means that we provide support to a critical piece of OA publishing or infrastructure. Hollins believes in an “academically rigorous undergraduate liberal arts education,” so that’s a reason why we’re a member of OA book publisher Lever Press, because it embodies and reflects the values of the liberal arts. Hollins desires to create a “just future,” so that’s a reason why we support subscribe to open models, so that all can have access to these valuable resources, which can make a difference in peoples’ lives.
And I made reference to the switch from buying content to buying access to content to point out that there have been radical changes in how the library connects our community with resources, and there will continue to be radical changes into the future. I agree that not all OA models will necessarily survive, and perhaps S2O will be one of those, but I don’t want us to make the mistake of thinking that purchasing stuff is the best or only way to connect our communities with the resources they need.
In my experience, it’s too great a leap to say that the university has an “express understanding” that the library’s collection budgets will be spent to purchase content or access to content. That’s not what we tell the university in our library’s mission statement. That’s not how I talk about our collection funds with my provost.
It sounds like in your case, Luke, it would be very easy to get a sense of what your institution really expects you to do with the allocated collections budget: just ask your provost a question like “Historically, we’ve always used our collections budget to buy access to content in various ways. There are now opportunities for us to use some of it to help make content freely available on an OA basis. Would that seem to you like an appropriate use of the budget you’re allocating to us for collections?”. Whatever the response, it seems like it would be very useful to you.
Oh, sure, Rick, I’ve had conversations about open access with my provost. She appreciates the work of the library and publishing communities to open up scholarly communication resources, and our library’s individual efforts to support this change. We don’t keep our OA support a secret — it’s part of our collection development policy, we talk about it with faculty, and we have a web page dedicated to why and how we support OA, https://library.hollins.edu/open-access/. But I’m not sure your framing of the question is the best way to have that conversation with a provost — we’re trying to make lasting change to a scholarly communication system that has kept significant chunks of content closed off from our students and faculty (except through ILL). We can help both our campus community and the larger world have easier access to this content — isn’t it incumbent on us to explore those possibilities?
But I’m not sure your framing of the question is the best way to have that conversation with a provost — we’re trying to make lasting change to a scholarly communication system that has kept significant chunks of content closed off from our students and faculty (except through ILL). We can help both our campus community and the larger world have easier access to this content — isn’t it incumbent on us to explore those possibilities?
Explore the possibilities? Absolutely. There are tons of ways that the library can explore supporting open publishing, some of which would be clearly in line with the intended use of allocated resources. (For example, if you have funding for a scholarly communications librarian, and/or for the creation and maintenance of an institutional repository, using those resources to explore ways of supporting open publishing would seem to me entirely in line with what the institution likely expected when allocating that money to the library.) However, redirecting money allocated to you for the purpose of buying content and using it instead to underwrite open publishing programs constitutes a move well beyond “exploring possibilities.” I wouldn’t do that without first making sure that such a redirection is in line with the goals of those who allocate the funding.
At most institutions of which I’m aware, the appropriate venue for that discussion would be the provost’s office — library directors usually report to and submit their annual budget requests to the provost. And it sounds to me like you’re having conversations along these lines with yours already, which I think is very wise. If she supports you in using collections money to underwrite open publishing, then it sounds like you’re good to go.
Hi Rick,
Thanks for sharing your concerns about S2O. Your points seem to fit well for Read and Publish models which increase the financial burden on institutions, especially large university libraries, and remove the connection between library collections budgets and content access. I find S2O a welcome alternative that delivers predictable and “neutral” pricing and maintains the connection between subscriptions and content access.
You mention that OA business models “tend to put a much greater financial burden on large institutions that produce a lot of research”. That is the case for Read and Publish, as larger institutions are likely to publish more articles ergo paying more APCs. It is not the case for S2O pricing, which is likely based on subscription price paid for a package of journals at the start of the “big deal” with accumulated annual increases and title-add costs. S2O delivers a predictable, negotiated annual subscription cost that is most likely to be cost-neutral.
You warn of the possibility of institutions not participating – dropping their S2O subscriptions, relying on the OA for access. Of course, that could happen. If it does, then the publications are not made open. If the content is important for an institution’s teaching and research, it is in the institution’s interests to continue the subscription. Just as we subscribe to journals that flip their content to free-to-read after an embargo (or the reverse), we maintain our S2O subscriptions to ensure access to needed content. With Read and Publish, on the other hand, when my institution can choose to not participate, we retain access AND get cost savings.
While it is not the OA utopia, we all hoped for, S20 is a pragmatic approach that I support until something better comes along.
Athena
Hi, Athena —
You’re right, of course, that S2O does not impose the higher burden on high-productivity institutions that a flip to APC structures does. (Except to the degree that high-productivity institutions are paying higher prices for subscription content, which is often the case; but even here, S2O doesn’t, in itself, increase the cost burden of those institutions.)
As to this, though:
You warn of the possibility of institutions not participating – dropping their S2O subscriptions, relying on the OA for access. Of course, that could happen. If it does, then the publications are not made open. If the content is important for an institution’s teaching and research, it is in the institution’s interests to continue the subscription.
But as I pointed out above to another commenter, the institution that decides not to keep paying faces little or no risk. Either others will continue to pony up and the content will remain free, or the journal will fail to meet its support threshold and it will remain a subscription (or hybrid or APC-funded) journal, in which case the institution is no worse off than it was. It experiences the benefit of temporary savings and then goes back to the status quo ante. This may sound like a very selfish way of thinking, but we have to bear in mind that what colleges and universities need money for are not usually programs that can be considered “selfish.” It’s not like the subscription savings are going to fund a salary increase for the library director or the provost — they’re going to fund other content purchases, or other institutional programs that are intended to make the world a better place in other ways.
All of that being said: there’s nothing wrong with supporting S2O if you feel like doing so is in harmony with your institution’s goals and if you’re confident that you have the support of those who allocate your collections budget. My advice to publishers is to avoid S2O, because I don’t believe it’s sustainable in the long run; my advice to librarians is to counsel with those who allocate their budgets before contributing to existing S2O programs.
I’m just sitting here wondering why the fact that all science is innately OA is surrounded by so much confusion/contention? Yes, some journals charge a subscription fee for immediate access in the comparative infantile life of an article, but then there is the rest of eternity which far exceeds an initial 365 day subscription period (usually) whereby anyone anywhere can sit and read any manuscript any time up until the destruction of the universe.
Somehow immediate gratification has supplanted eternal access in all of this OA hysteria.
I’m just sitting here wondering why the fact that all science is innately OA is surrounded by so much confusion/contention?
Bob, I think the confusion is mostly on your end. To say that “science is innately OA” is to commit a category error; “science” can’t be OA. Publications (whatever the discipline) may or may not be OA, but if you have to pay for access to them either upon publication or in the long run, then there is no definition of OA that those publications fit.
I genuinely don’t believe that anyone else is confused on this point.
So I will agree I should have been more specific and said, “published science is innately…” which would be more accurate. Regardless, access on day 366 since publication (in most cases) if free/OA. If a reader didn’t discover the article until day 366 or not until day 8566 since publication and there is no paywall restriction, it’s by definition an OA article.
Bob, I think you’d have a very hard time finding an OA advocate who agrees that an article published on a toll-access basis and then made publicly available after an embargo period is “by definition an OA article.” Most, I think, would actually say that publishing it initially on a toll-access basis makes it, by definition, not OA. (And then you could get into a very interesting argument about whether it then becomes OA after the embargo period. The answer will depend, in part, on the terms of the license under which it’s eventually made publicly available.)
Rick, I completely agree. Another pair of unfounded assumptions in the various mechanisms proposed to help shift to OA is that 1) funds are fungible across categories of library allocations (whether for collections or anything else), and 2) that library directors have the final say-so to approve all such reallocations. Neither is consistently true, especially in smaller colleges but also at research universities, especially when overall budget crunches get particularly bad at an institution.
At a meeting earlier this year, a librarian mentioned that she was not always aware that a journal or collection of journals had transitioned to S2O. I supposed the invoice comes regardless and depending on how invoices are paid at any given institution, this may escape attention. What happened next, according to this librarian, is the usage reports showed no usage, as is apt to happen when content is open and no login is required. She put the collection on the cancelation list.
I have always been concerned that provosts– or more likely CFOs at institutions– will figure out that a library is paying for something that is “free.” Loss of research money, 9 figure payouts to the government, caps on F&A, loss of international students, and god only knows what else is coming. The future does not bode well for voluntary institutional expenses.
Right. And trying simply to reframe the voluntary expense as something other than a voluntary expense does nothing to change the real-world dynamic. Reality always wins (though not always in the short run).
Thanks Rick….there are two issues–flipping to S2O now, which can be and has been done, and then building a sustaining model that recognizes or allows for growth of scholarly publishing over time–more papers and new journals, even accounting for the risks you note. One may say we’ve had enough, and don’t need that growth (as has been said for decades), but the reality is that legitimate growth–ignoring all the predatory mess and fake special issues, etc.—has occurred and I think will need to occur. New fields emerge that merit new journals (AI/ML) etc., and new needs like supporting notebooks and data/code curation, and publishers will need to invest in new technology and infrastructure. So imho, that’s the real challenge in this as a long-term model. A new journal based on APC business models can succeed if there is a demand, or if it can capture content from existing journals of other publishers. but how do you convince folks to support new S2O journals? Some S2O contracts allow this, but at scale?
Agreed. Vertical scale (accommodating an increasing amount of research output) and horizontal scale (maintaining support over time) are both real challenges for S2O. One thing that worries me is that I see publishers making long-term S2O decision based on very brief success experiences: “90% of our customers continued subscribing the year following our flip to S2O! The model works!”
My takeaways from some prior involvement with S2O partnerships at Knowledge Unlatched (Mar 2022-Dec 2023). These are just my brief observations, and they will not hold true for people with different experiences and insights.
– Publishers and their programs are not created equally or universally, nor is support for their S2O
– Access-mission aside, S2O is generally not a revenue growth model as much as a revenue maintenance model with risk of decline from ‘riders’ and general lapsing
– S2O requires above-average strong content to maintain Net support
– Like any “subscription” model, lapsed subscribers are difficult, if not impossible, to recover
– At the risk of oversimplification, reconsider S2O if you need support from institutions in China
– Publishers risk accelerating the ‘rider’ challenge if they don’t provide sufficient and ongoing evidence of value and return to individual supporters – it’s a relationship around value
One other point worth considering: S2O is not just about maintaining current subscription revenues, to do it right you need to also replace other revenue streams that the model eliminates, such as rights licensing revenues and APCs. So most publishers looking to move to S2O need to significantly increase either the number of subscribers to their journal or the subscription price of the journal in order to break even. IWA discusses this here:
https://scholarlykitchen.sspnet.org/2020/12/08/guest-post-subscribe-to-open-the-why-the-how-and-the-what-now/
I entirely agree David. Each publisher differs in what their numbers look like (and what value their content offers).
I think you’ve done a good job of articulating the risks, pressure points, etc. of the S2O model, here and elsewhere. I have also heard you make clear in other posts, presentations, and conversations that your own institution’s priorities and expectations lead you to the conclusion that S2O is not aligned with those priorities and expectations. What I don’t see is the evidence that the title would demand – that S2O is doomed.
Where are the other library deans/directors reporting that their delegated authority for stewardship of collections budgets presumes the sort of approval that you apparently need from their campus administrators? Or, for those who have the expectations that your administrators have for this level of consultation and approval, that their administrators see things as yours do? That’s the evidence I’d expect for a claim of doomed vs that S2O is one of the many models we’ll see continue among the plurality that exist and will be invented. If there are a significant/growing group of such library directors, that’ be useful to hear about.
(BTW, personally, I have significant concerns about S2O because I think it constrains growth, but that’s a completely different aspect, and even with that concern, I still think it is a good match to certain types of publications.)
Hi, Lisa —
Where are the other library deans/directors reporting that their delegated authority for stewardship of collections budgets presumes the sort of approval that you apparently need from their campus administrators?
I guess I should clarify: my line leadership hasn’t imposed an unusual level of oversight on our use of the collections budget. But because I’m conscious of the fact that our collections budget is allocated to us on the assumption that it will be used for the purpose it has served for 100 years — purchasing content or access to content — I want to be careful about unilaterally deciding to use it for a very different purpose. That caution led me to talk to my line leadership before making such a switch. Those conversations led me and my collections team to conclude that redirecting collections money towards underwriting OA would not be in line with our institution’s strategic posture. (I had similar conversations at my previous institution, with similar outcomes.)
I strongly suspect that my situation is typical — that if the great majority of academic library directors and collection development officers were to have conversations with their administrators similar to the ones I’ve had with mine, the outcomes would be very similar. I suspect this for the reasons I laid out in my post. But because I a) strongly suspect this but b) have no direct evidence in support of my suspicion beyond my own (limited) experience and the academic and economic dynamics I’ve described above, I urge my colleagues who are inclined to participate in S2O programs to first discuss their plans with the administrators who allocate collections budgets to them and figure out whether doing so would be in alignment with the institution’s wishes. I’ll go out on a limb and predict that in some cases, the administrators will say “That sounds like a great plan. Go for it.” I’ll go further out on a limb and predict that a) this will be a very small minority of responses, and b) in those cases, support will decrease over time as the S2O ask continues to grow (which it invariably will).
In short, I believe S2O is doomed because I see no reason to believe that underwriting open access publishing will rise to the top of the fiscal priority list for a critical mass of academic institutions, over time.
You seem to assume that the conditions under which you’ve been allocated your collections budget is that which others labor under. Personally, I doubt that’s the case given the wide range of things I have seen collections budgets used for even before OA. But, I also wonder if you’d get a different response if you said “I’d like to ensure that our scholars have continued and uninterrupted access to this resource, and it is possible that in doing we will contribute to the resource being made freely available to everyone, ok with you?” Though, perhaps your institution/users/you don’t value continuous and uninterrupted use in the way others do.
To be clear though, if you find free riding and then dealing with having to restart a subscription is worth the risk, it seems very logical to take that risk. But, I don’t at all see compelling evidence to concur that the model is doomed (though application of the model in some cases may be unsuccessful in opening content and thus a risk for the publisher).
You seem to assume that the conditions under which you’ve been allocated your collections budget is that which others labor under.
I do assume that when a university allocates a collections budget to its library, it expects that money to be spent on building collections, not underwriting open publishing. (And no, I don’t agree with those who say that those two things are essentially synonymous.) That’s been the case at every institution I’ve worked at, and it’s the case at every institution of which I’m aware — the library isn’t just given a big lump of money and told “Do what seems best with this,” but is rather given a lump of money for running the library organization and facility, a lump of money for collections, and (in most though not all cases) a lump of money for personnel. Up until 15 or 20 years ago, there was no question what the collections money would be used for; now there are proliferating options, most of which are proliferating below the awareness of university administrators — hence what I see as the strategic importance of more granular conversations with administration if the library wants to use that money differently. But if there are institutions at which such budgets come without any strings or expectations as to how they’ll be used, then those institutions have more freedom than others to use the budgets as they wish. I would caution my colleagues against simply assuming that when money is allocated to them for buying content, the allocators would be equally happy to see it used for very different purposes.
But, I also wonder if you’d get a different response if you said “I’d like to ensure that our scholars have continued and uninterrupted access to this resource, and it is possible that in doing we will contribute to the resource being made freely available to everyone, ok with you?” Though, perhaps your institution/users/you don’t value continuous and uninterrupted use in the way others do.
As I’ve pointed out in response to other commenters, in most cases I don’t believe declining to support S2O means taking a significant risk of interrupted access. If (based on the details of a particular case) that risk is significant, then of course that would factor into the library’s decision about whether to participate. It’s not a binary question of whether the library does or doesn’t “value continuous and uninterrupted use in the way others do” — it’s going to be a complex calculation based on the importance of the content to the institutional mission and the risk of actually losing continuous access. But if I’m paying into an S2O program because failing to do so would likely result in an interruption in access for my institution, then I wouldn’t feel any need to coordinate with my boss; what I’m doing in that scenario is using the collections budget precisely for its intended purpose. (And I also might question exactly where the “O” is in that particular S2O model.)
To be clear though, if you find free riding and then dealing with having to restart a subscription is worth the risk, it seems very logical to take that risk.
The risk of having to restart a subscription is a pretty low-impact one — and if the former subscription price of the journal (or journals) was high, the risk/benefit calculus leads even more towards a free-riding posture. Though the term “free rider” seems a bit tendentious when the whole purpose of the model is to make the content free for everyone. Having once paid for a journal doesn’t, in my view, magically create a moral obligation to underwrite its open publication, particularly when that money is being freed up to advance my institution’s mission more directly and effectively.
“If there are institutions at which such budgets come without any strings or expectations as to how they’ll be used…” — I just want to make sure it is clear that I wasn’t suggesting I know of any libraries without any strings or expectations; however, I do know of ones where, based on what you’ve said, there appear to be different ones than the ones you’ve experienced.
Also, I don’t mean anything negative by free rider – just descriptive (but I understand it is often used negatively so probably should have phrased that differently).
Rick Anderson raises fair concerns about the sustainability of Subscribe-to-Open, but I think his conclusion that the model is “doomed” overlooks both history and current practice.
First, our scholarly ecosystem has long relied on voluntary, collective commitments. Libraries sustain HathiTrust, ORCID, arXiv, LOCKSS, and consortial subscriptions without being compelled to. These succeed because institutions recognize that their missions are better served by maintaining shared infrastructure than by retreating to purely transactional deals. S2O is no different.
Second, Anderson frames S2O spending as shifting from “mission-critical” purchases to discretionary “support.” But that framing misses a crucial point: universities benefit most when their scholars’ work circulates as widely as possible. Large research universities in particular gain disproportionate visibility and citation impact when journals flip open. In that sense, continued investment is not optional philanthropy; it directly advances institutional goals.
Third, redirecting funds back to pay-to-read or pay-to-publish models would undercut those same goals. Subscriptions limit readership of their own faculty’s research, while APCs restrict who can publish, disadvantaging early-career scholars and underfunded disciplines. Both approaches clash with institutional commitments to broad dissemination. S2O, by contrast, spreads costs fairly, avoids inequities, and reduces administrative burden.
Finally, the track record so far is encouraging. Programs at Annual Reviews and Berghahn have shown stable renewal rates, not collapse. The predicted “free rider” problem has not materialized at scale.
Anderson is right that S2O requires trust and community commitment. But that’s not a flaw—it’s a feature that aligns it with the very ethos of scholarly communication. If universities mean what they say about impact, equity, and global reach, then supporting Subscribe-to-Open isn’t charity. It’s strategy.
First, our scholarly ecosystem has long relied on voluntary, collective commitments. Libraries sustain HathiTrust, ORCID, arXiv, LOCKSS, and consortial subscriptions without being compelled to. S2O is no different.
S2O is in many ways similar to the examples you’ve cited, and for that reason I wouldn’t encourage my colleagues to direct collections money towards those things without making sure that those who allocated the money are supportive. Again, though, it’s important to consider whether membership in (or underwriting support of) these initiatives provides any direct benefits to the sponsoring organization, and if so, what kinds of benefits they are. In the case of HathiTrust, for example, there have historically been benefits offered only to supporting members, some of them having to do very specifically with access to the HathiTrust collection. That might make a HathiTrust membership a more natural fit for the collections budget than, say, supporting ORCID. ArXiv represents more of a pure support scenario: it’s clearly a content offering, but there are not (to my knowledge — correct me if I’m wrong) any enhanced access benefits or services for financial underwriters, so contributing to arXiv doesn’t build the contributor’s collection in any direct way.
Second, Anderson frames S2O spending as shifting from “mission-critical” purchases to discretionary “support.” But that framing misses a crucial point: universities benefit most when their scholars’ work circulates as widely as possible.
For the sake of argument, let’s allow that that statement is categorically true. The question, then, is whether declining to underwrite an S2O program significantly risks undermining the institution’s researchers’ ability to circulate their work as widely as possible. The likelihood of that risk being high enough to justifying the cost of underwriting a particular open journal or suite of journals strikes me as trivially low.
Third, redirecting funds back to pay-to-read or pay-to-publish models would undercut those same goals. Subscriptions limit readership of their own faculty’s research, while APCs restrict who can publish, disadvantaging early-career scholars and underfunded disciplines. Both approaches clash with institutional commitments to broad dissemination.
To the degree that an “institutional commitment to broad dissemination” exists, that commitment will certainly come out in the library’s conversations with campus administrators about redirecting collections funding towards the underwriting of open publishing. And if that commitment is really there, and represents a high enough institutional priority, then the library and the administration will come quickly to agreement about that budgetary redirection. But it’s also possible that having a limited readership will seem to administrators like a reasonable tradeoff for the other good things that the university could do with the money involved.
Finally, the track record so far is encouraging. Programs at Annual Reviews and Berghahn have shown stable renewal rates, not collapse. The predicted “free rider” problem has not materialized at scale.
The available data is still way too limited for even cautious conclusions, on two dimensions: time and breadth. Too few publishers have tried S2O, and the ones declaring victory are (in my view) doing so way, way too early — and by no means are all of them are declaring victory. David Crotty provides examples of collapse above. In any case, sustainability is not demonstrated by one (or even three) successful rounds of renewal.
Anderson is right that S2O requires trust and community commitment.
I can’t accept credit for being right about that, because I never said it.
Putting on my chair of the ORCID Board hat, I want to make sure that it is clear that there are benefits that ORCID members have that are not available to non-members. Whether those are a fit with the “collections” budget will of course depend on whether your individual institution would see certain kinds of access to data as access to a content or if they’d see it as a component of providing access to content (e.g., by populating an institutional CRIS system).
Thank you, Rick, for expressing a viewpoint I have been thinking for so long, but have been discouraged about saying it openly because the social pressure to be “pro-OA” is almost overwhelming among academic librarians. I have held firm (if impotently as the power rests with my boss, the library director) that in our public university funded primarily by taxpayers and student fees, our collections budget is entrusted to us librarians with the understanding that we will maximize researcher (students, faculty) access to the resources they need to do their work. Whatever lofty statement the official mission expresses about community etc., everyone knows that if we started treating any part of the collections budget as our little slush fund for supporting non-profit projects of any kind (including publishing) of our choosing, there would be an outcry and very likely a loss of those funds. Private institutions may be very different in this regard. As an acquisitions librarian, I will say that more specifically, we have a fiduciary responsibility to spend our money on the needs of our university far above anyone else and any other long term professional ethical value. I’ve gotten the impression that the term “fiduciary” is not in the vocabulary of most librarians.
Melissa – as I have said, I’m just an uninformed onlooker, and I appreciate that your duty is to your employer, the university, and your students. But if all universities publish the material you need on an open access basis, then your students will have access to it without you having to buy it.
David, the problem is that those students (and the faculty) need not only access to material produced by their institutions — what they need much more of is content produced elsewhere. Buying access to that content is the purpose of the collections budget. It’s essential to understand that no library can afford access to everything its constituency needs; so when a journal becomes OA and its publisher asks the library to keep paying so that it can stay OA, the library has to ask itself whether it should keep paying for content that is now available at no charge, or instead use that money to start paying for other content that it was not able to buy previously.
First of all, thanks Rick – and the many commenters! – for a thought-provoking discussion of a tough challenge. For those of us committed to both open publishing AND moving away from the inequities of APCs, there aren’t easy answers. As we work on this at PLOS we share the concerns about offering our publishing services only as a public good. An appeal to values-based acquisitions practices only gets you so far – and that’s likely to be all the more true as higher education funding in many regions comes under increasing pressure. But I do have questions about a couple of assumptions:
1. We talk about the challenge of a system that puts more burden on research-intensive institutions. But surely that’s appropriate in a model where institutions are paying for publishing services, rather than access – communicating the research is a core part of the research process? Of course, there are questions about how this might work in practice – for example, is this still in the library budget – but my question is about the principle.
2. I think we need to clearly separate Diamond and S2O. As others have noted above, one of the founding principles of S2O was baking in some degree of private benefit that would encourage libraries to continue subscribing. You seem skeptical about that – do you think it’s impossible, or just that we haven’t found the right model?
Hi, Alison —
Great questions, thanks. To attempt a couple of answers:
We talk about the challenge of a system that puts more burden on research-intensive institutions. But surely that’s appropriate in a model where institutions are paying for publishing services, rather than access – communicating the research is a core part of the research process?
In principle, I agree that it’s fully defensible for institutions that use more publishing resources to pay more into the publishing ecosystem. (Of course, by that same logic, it was defensible for larger institutions with more readers to pay publishers more for access to their content.) Also in principle, I think we can agree that communicating research results is vital. (Whether it constitutes a “core part of the research process” is a topic for another discussion, I think, one that would need to define carefully what we mean by “core part.”) But obviously, what makes a system sustainable is not whether it’s defensible in principle, but whether it can be sustained in practice, which is almost entirely a question about money. I think that remains a very open question for S2O — and for some other models, of course.
I think we need to clearly separate Diamond and S2O.
I’m skeptical of that separation; to me, S2O seems like a particular mechanism for supporting Diamond journal publishing. (I’m defining “Diamond” as a model that levies no charges on either readers or authors, and relies instead on institutional subventions. S2O generates one kind of institutional subvention. I realize that may not be a universally agreed-upon definition, but I’m not dogmatically advancing a particular definition of Diamond in this piece.)
I’m also skeptical of the ability of S2O models to incorporate sufficient exclusive incentives to retain paying subscribers. I’m not asserting that it can’t be done, only saying that the solution isn’t obvious and the decidedly mixed record of S2O success we see so far suggests that finding that balance is very difficult. One problem is logical and structural: to the degree that the incentives involve access to content, you’re no longer really publishing openly; to the degree that the incentives involve things other than access to content, you’re undermining the logic of using collections budget (and therefore the much-vaunted workflow consistency) to support it. And if you’re not using the collections budget, it becomes even less clear where the money will come from. But another problem is simply finding incentives that will be as mission-critical to the sponsoring institution as brokering access to research content is. (Access to publishing opportunities is arguably just as mission-critical, but once you start offering free publishing in return for a “subscription” fee you’ve basically departed the realm of S2O and entered that of transformative agreements, which of course have their own problems, both philosophical and practical.)
Dumb question Rick, but how does the S20 model differ from, say, donating to our locate NPR stations? When we donate, we’re paying for “free” content so real news stations can stay on the air and provide content to everyone, ourselves included. And enough people believe in this model to make it viable (in combination with sponsor support, grants, etc.). Isn’t S20 pretty much the same thing? To me, anyway, it seems like an equitable way of supporting open content, provided that the people forking over their support dollars view this expense not as optional but as part of their commitment to supporting knowledge—no different than funding travel for conferences, updating software, or buying books for the shelves. Maybe the issue you’re identifying is less about the model than the ability of librarians to sell this model during times of austerity?
Dumb question Rick, but how does the S20 model differ from, say, donating to our locate NPR stations?
It’s not a dumb question at all. That’s actually an analogy I’ve used regularly when discussing the pros and cons of S2O. In one way, the models are very similar: NPR’s viability depends, to a significant degree, on listeners being willing to underwrite its work voluntarily. (And the fact that only about 10% of listeners donate is something that S2O advocates should seriously consider.)
But there’s also a fundamentally important difference: it is that NPR listeners are donating their own money, which is theirs to do with as they wish. Libraries that donate allocated funds to support open publishing are using money that does not belong to the library, but actually belongs to the library’s sponsoring institution and was allocated to the library for specific purposes. That’s why wise library leaders will check with those who allocate those funds to them before using the funds for purposes very different from what was intended.
If we are to look at S2O and similar models the same as supporting NPR, this raises another question, at least for public institutions. Are librarians legally allowed to support this using public funds? This will vary from country-to-country and state-to-state, along with what any added benefits are that can be justified. However, I suspect in many cases public institutions are not actually allowed to redirect public funds to “donate” to a cause, no matter how worthy (which can be even more of an issue if the journal is published by a for-profit). And if the funds can be used this way, it may have to be through a very different process than a typical library acquisition.
To be clear, I am not saying S2O is the same, but I think it is a good analogy for the reasons Glen mentioned above. It might be similar enough that librarians should review their procurements rules and regulations if there is no clear or only a de minimis direct benefit to the institution.
Glenn, there’s a significant difference from the NPR model. If an NPR station operated under a parallel S2O approach, the station would set a threshold and, if that threshold wasn’t met, access to the broadcast would be limited to those who donated. Instead, NPR scales its activities to the budget that it raises, with presumably a floor below which it wouldn’t be able to operate.
One thing is for certain here, libraries will need auditable and accurate Open Access usage statistics from all publishers. We will be launching soon a way for libraries to acess OA usage metrics via theIPregistry.org. This will help in any decisions made about OA because you will see exactly how much OA is being read for each publisher and you can make better informed choices, like those you make with Counter stats on paywalled content.
I find it fascinating that with so many comments, no one seems to have mentioned two concepts that would be directly relevant: “freeloading” and “tragedy of the common ground”. (I haven’t had time to read all of the comments, so Alt-F searched and didn’t find them.)
In one sense, the entire OA enterprise has been about deliberately providing “freeloading” “read” opportunities for those who otherwise couldn’t afford access, but freeloading by institutions that can afford it has always been a concern. I have many times over the years asked my own team as well as others on listservs if they have any kind of threshold in mind for either individual journals or entire packages for cancellation. For example, a threshold might be that once a hybrid journal reaches 70% OA articles, the paid sub could be cancelled. Absolutely no librarians that I’ve seen either have set such a threshold lower than 100%, or they aren’t admitting to it in public.
The “tragedy of the common ground” obviously comes into play in the debate here about the prospect of journals being flipped back from open to closed due to lack of continued paid subscriptions.
I actually haven’t noticed ANY journals flipping back. From some comments here I get the impression that there are some, and I would appreciate example titles.
We are all of course going to watch what happens with ACM’s flip of their entire journal portfolio over the next year or two. We are among many libraries that, sorry ACM, will pay very close attention to our usage and faculty publishing to make that decision in a couple of fiscal years.
We also have an example here in Canada of another approach a publisher might take. NFB (National film board) has been selling higher ed subscriptions to their streaming videos for a long time. This year they just switched to “free to all” (I can’t call it OA because it’s not any kind of OA license). But, and this is a big but, they have said they expect all previous subscribers to continue, and have threatened to use their knowledge of our IP ranges to block all users from our IP ranges from access. Yes, there are workarounds, but they may not be legal and our institutions are no more going to try that than they are likely to recommend faculty use SciHub.
Melissa, no need for apologies regarding ACM’s upcoming flip to 100% OA on January 1, 2026 and paying close attention to faculty and student publishing activity to determine value. Only time will tell if ACM’s model, which is very different from S2O, is sustainable over the long term (i.e.- years, decades, longer, etc.), but with over 2,750+ universities on board (most with 3-5 year Agreements) and more joining every week, we are optimistic about the model’s near term prospects for success (3-5 years). I share many of Rick’s concerns about the S2O model and have been fairly critical about the model since its introduction. What concerns me most about S20 is the long term sustainability of the model for exactly the reasons Rick has given in his various posts. If an individual library cancels a subscription to a particular journal title and that title continues to remain OA via S2O, the canceling library loses nothing and gains budget to use elsewhere. This is not the case with ACM’s model. If an institution cancels their ACM Open license, the immediate impact is that affiliated authors at that institution would need to pay the APC to publish in any of ACMs 70+ journals and hundreds of technical conferences, since ACM Open covers everything we publish. That institution would also lose access to the Premium version of the ACM Digital Library, which includes a long list of premium features and content, as well as the largest bibliographic index covering the field of computer science. Our historical usage data shows that for many of our university customers, the value lost and impact on faculty and students in computer science would be significant. One thing we learned very early on is that authors do not want to pay APCs, but at the same time they feel enormous pressure to publish in the top journal and conference venues, so ACM Open helps them to continue publishing with ACM, get the tangible benefits of OA publication (1) Funder compliance (2) increase downloads and (3) increased citations. For institutions that publish very little with ACM, the costs are typically lower with ACM Open than they have historically been for just DL “Read” Access. For those that publish a lot, there is an increased cost, but the value proposition is there. But, as Melissa pointed out, only time will tell if this is long term sustainable. That will depend to a large extent on the research community and the library community.
Whether the thousands of institutions that have already signed on to ACM Open will be able to continue supporting our model over the long term will depend to a large extent on the value proposition. Budgets and funding will always be an issue for libraries and one of the things we’ve had some success with for the ACM Open model is facilitating cost sharing between libraries and departments (which are getting a large part of the value from OA publication and access to the premium version of our Digital Library through ACM Open). We have seen numerous examples of departments pitching in to share costs and this has directly led to multi-year agreements being signed. Relying solely on library budgets to fund models like ours over the long term is risky.
It is very refreshing to hear from a publisher who has such a practical understanding of the concept of “value proposition”, which is exactly what Rick’s original post was about, and thus has designed their Open plans with realistic understanding of how decisions are made, rather than assuming librarians will just want to “help out” with their budgets, as if collection management is a moral undertaking rather than a financial one.
The only wrinkle I see is that I think most of us have very silo’ed budgets in our universities. Saving faculty/research money (not having to pay APC fees) doesn’t in any way cause that saved money to end up in library budgets, unless the library administration and dept/univ research arms make an explicit effort to have discussions and even negotiations about the matter. This is certainly not the first time in my 30+ academic library career that I’ve seen an idea that could save the university overall money but because it changes whose budget pot the cost comes out of, it didn’t work. In this case, library subscription costs versus faculty APCs. I know some libraries are actually directly involved in managing APC “grant” budgets internally, but for those of us who aren’t, there’s some hard extra internal bargaining work to be done.
Thank you Rick and others. So much to think about here. I note and agree with the problems cited, but somehow I have always thought of S2O as being, in principle, the most equitable route to OA. If I may recount an analogy, as a child I lived in a small town in Iran (Bojnurd) with a population of 30,000. With no TV in the town, the sole entertainment was the one cinema which we attended regularly. We could afford to go but many couldn’t. The enterprising owner had heard that in the US there were outdoor cinemas, so he built another cinema next door but without a roof, saving a lot of money! It was very popular and I still have great memories. The residents of the neighbouring houses, who would traditionally dine on the roof in the hot summers, suddenly had free entertainment while they dined – and they had no problem watching the same movie for a week! So we paid for the running costs and they “freeloaded” (to quote Melissa above). There was no friction between the two parties – we would always pay to go and the guys on the roof couldn’t afford to take the whole family, so a win–win. I guess the fact that we had comfortable seats was the “added benefit” that we got and we continued to pay and everyone was happy. I admit this is not a perfect analogy!
Regarding S2O and some of the comments, I think that more communication with different stakeholders is important. Perhaps more regular communication from the publishers to libraries as well as authors emphasizing the importance of continuing their subscription and underlining the extra benefits – if any.
Finally, an inherent problem I have been grappling with is the practicalities of S2O. As it is an annual process, there is always the added admin overhead of when to decide to flip for the year, even though as has been stated, once flipped, journals generally stay open. Often the decision to flip is not made until 2–3 months into the year in question. This means licenses need to change retrospectively and resubmitted to indexing platforms – having good updatable metadata helps. I wonder if publishers can take a risk earlier in the process and bite the bullet, taking the risk that the S2O limit might not be reached, but obviating the need to update and resubmit metadata.
Thanks very much for the thoughtful comment, Kaveh.
I’m going to quibble with the outdoor cinema analogy, though: as with the NPR model discussed above, one important difference between your situation and S2O is that everyone who financially supported the outdoor cinema arrangement was using his or her own money. But academic libraries that support S2O are doing something very different; they’re taking money allocated to them for the purpose of buying access to content, and using it for a different purpose. To extend your analogy, it’s as if you and your neighbors had been given money by the municipal government to plant gardens to help beautify the city, and decided instead to use that money to support a free outdoor cinema. Both goals may be worthy; the municipal government might even have supported that redirection of funds. But if you and your neighbors want to continue receiving funds for public projects, wouldn’t it be wise to check with the municipal government before using the allocated funds for a purpose very different from what was intended by those who allocated them?
100% Rick. I had that in mind but forgot to write it. So it underlines your principal point. I wish life was simpler. 😉
I don’t know Rick. Institutions make decisions all the time about how to spend money that isn’t technically “theirs”—where to invest, what programs to support, what discounts to offer, etc.—and many of these decisions are value laden. So if S20 spending aligns with the values a university embraces and serves their community well, how does this differ from, say, a university deciding to cut history programs and offer more computer science? I didn’t ask for this. We hand over money—via tax dollars, grants, investments, etc.—and entrust our institutions to spend it wisely, using their best judgement.
Hi, Glenn —
The question isn’t so much whether S2O is aligned with the university’s values — though that is one important question. In this particular context, the question is whether the library is using the money for the specific purpose intended in its allocation. Everything that we do has to be values-aligned; but not every value-aligned activity is appropriate for the library, or is an appropriate use of the library’s collections budget. (For example, it would be entirely in harmony with my university’s mission and values for the library to hire another religion librarian. It would not be appropriate for me to use the collections budget for that purpose.)
To build on Kaveh’s analogy, consider that some cinemas and movies attract more favorable interest than others. Much depends on the expected or known value of an S2O program. Some S2O offerings are more attractive and marketable than others. The reception of one publisher’s program can differ a lot from the reception of another publisher’s program — as they should, since they are different content offerings.
Rick, you’ve repeated this (or something like this) so often: “academic libraries that support S2O are doing something very different; they’re taking money allocated to them for the purpose of buying access to content, and using it for a different purpose” that I have to ask … if not your collections budget, which budget (operations?) does your library use to pay for discovery layer/catalog, A&I services, your OCLC cataloging bill, etc.? None of these are “buying content” — but rather facilitating discovery of content — and all of which I have seen paid from collections over the years of my experiences in different libraries.
Hi, Lisa —
In BYU’s case, those things are mainly paid for centrally by campus IT. To my knowledge, I’ve never worked in a library where those kinds of services and memberships were paid for out of the collections budget — in my experience, they would typically be paid for out of operations or technology funds. (Which, of course, is not to dispute your experience of seeing them paid out of collections budgets elsewhere. In those cases, I would imagine that the library and the allocating administration have mutually agreed that this is an appropriate use of collections funding.)
Wow, that’s amazing that campus IT pays for the library catalog, abstracting/indexing services, and OCLC! Maybe you’ll do a post sometime explaining how that was pulled off. I bet a lot of us would be pretty excited to have the campus paying for Scopus, Web of Science, etc.
It’s not quite as exciting as it might sound. There are a number of big-ticket software and technology-related products/functions that are paid for centrally by campus because, years ago (and in a few cases recently), the library agreed to transfer that chunk of our budget to central IT. It benefits us because it means that although we lose a chunk of budget, future price increases are absorbed by central IT rather than by us.
Most of these agreements were struck before my time, but to my knowledge, none of these transfers have come out of the collections budget, but rather they’ve all come from operations. Certainly the couple of times we’ve done this since I arrived, the money has not come out of collections.
I appreciate the extra info. I’m pretty shocked your A&I services came out of operations and not collections previously. But, I guess it’s a good reminder that, though there may be patterns, there is also great variety in academic libraries!
Oh, sorry — correction: we certainly do pay for abstract databases and indexes out of our collections budget, because they constitute collections content in and of themselves as well as being tools for making effective use of content. (When you initially said “A&I services” I actually read it as “AI services” — insert headslap emoji here.)
A response to Rick’s Sept 10 response to Elizabeth (since this platform doesn’t allow responses nested that deeply): Sad though it is, there are publishers of reputable scholarly journals that do NOT guarantee post-cancellation access rights (PCA, also know as perpetual access rights, PAR). Although most of them do offer a few years’ backfile access with current subscription, it is entirely possible that a library that paid the full price for a journal in 2023, cancelled in 2024 because of OA hopes, then started another subscription in 2025 might have totally lost access to 2023’s issues.
Good point, Melissa. What you’re describing would be something less than S2O, of course, since in that scenario the content putatively published under an “open” model would not really be open at all.
Hello Elizabeth, Melissa, Rick, David
This discussion underlines how helpful developing long term mechanisms for supporting new equitable OA models like S2O is.
The Royal Society’s S2O model offers flat pricing throughout three year agreements, ie zero increase in Y2 and Y3. We expect most S2O supporters to be our existing subscribers, so this is a genuine saving for our loyal subscriber base. There various other ways long term agreements can be made mutually beneficial.
On the specific discussion about legacy access, if a customer subscribes to our 2025 volumes, supports S2O in 2026, then stops, they will keep legacy access to 2025 and (like everyone) will have access to 2026’s CC-BY OA articles.
Our perspective is that libraries will support S2O if the return on investment is good and it aligns with their priorities. That approach has seen us through 360 years of journal publishing and we see no reason to change it now.
Thanks for this Rod. A concern I have is where it is only individual journals that are being subscribed to on a year by year basis, via journals agents (for example). Not every institution will have the funding to subscribe to ‘big deals’ that cover 3 year periods – if that is what you are referring to with your comments on Y2 and Y3.
Thanks Elizabeth. Yes, we recognise that libraries have different requirements – one year agreements will suit one, multi-year agreements another (and we offer both). Publishers need to keep evolving models like S2O so that they work well for libraries and publishers alike
