A recurrent topic among faculty and librarians interested in infrastructure reform is the question of whose turn it is to make the next move. Researchers rightfully argue that they cannot submit their work exclusively to modern, open access journals because that would risk their and their employees’ jobs. Librarians, equally correctly, argue that they would cancel subscriptions if faculty wouldn’t complain about ensuing access issues. And so the discussions usually keep turning around and around, not getting anywhere.
Both groups are correct, of course, and so it boils down to which of the two obstacles is easier to overcome.
As tenured faculty myself, the only way to solve the problem on our side is to remove or at least decrease the pressure to publish in subscription journals. Many have realized this, for instance there is DORA, the REF explicitly excludes impact factors for evaluation and my colleagues and I have published a review of how top-ranked journals publish the least reliable science. I would argue that we have taken all the steps we currently can reasonably take. However, there is still a long and tedious road ahead of us where many important factors are out of our control. For one, the hypercompetition for positions and funding, due, in large part, to allocation of funds on a project rather than ongoing basis, leading to an overproduction of early career scientists and cash-strapped labs, is largely beyond faculty control. Second, even if public statements eventually were to manage to effectively devalue journal rank, in an atmosphere of hypercompetition, the obvious risk-averse strategy is to keep assuming that journal rank is instead used on an informal basis: having a paper in a top journal would never hurt, while one in a ‘lesser’ journal could. Hence, as long as there are journals and hypercompetition, faculty will continue to risk their own jobs, their employees’ jobs or the lights going out in their labs if they refuse to submit to subscription journals.
In brief, the obstacles for faculty radically changing their submission strategy are a) human nature, b) government-induced hypercompetition and c) the existence of journals.
While all three obstacles can (and perhaps should) be overcome in principle, there is probably a decent argument to be made that c) may be easier to master than either a) or b). Which brings us to the obstacles for eliminating journals by simply starving them of money. As all agree that paywalls have to go, subscriptions will have to run out at some point. The discussion is no longer about “if”, it’s about “when”. As it appears, for the above reasons, unlikely that scholars will substantially change their ways within my lifetime, I would argue that right now is the time to let subscriptions run out as the logical next step. I have two reasons for my argument, one is technical and one is logistical.
The technical argument, spelled out in more detail elsewhere is that we now have a whole slew of solutions at our disposal to create something jokingly referred to as “legal Sci-Hub”. Much as the actual Sci-Hub, it would provide convenient but patchy access to the scholarly literature, only this version would be legal (with, as now, actual Sci-Hub filling the patches not covered by legal Sci-Hub). As the current subscription-based system is also patchy, but often inconvenient, there is little reason for faculty to complain if they instead get something similarly patchy but maybe more convenient. Moreover, as the change entails an upgrade that comes with massive savings, faculty need not even be involved in the decision, as hardly any of them would know any of the involved technicalities anyway – faculty hardly know the publishers behind their journals, let alone what a link-resolver, DOAI or LOCKSS is. The only thing required is, of course, to inform them that there are a few upgrades running behind the scenes and that they are asked to pardon the dust for a while. In the same letter, faculty would also be informed that the upgrades come with such savings, that they will receive a whole host of new services that such a modern infrastructure will enable.
The logistical argument involves looking at the number of people who have to be convinced to act. If researchers are asked to act next, this would entail convincing just under 8 million researchers at ~10,000 institutions to risk their jobs. What are the number of infrastructure experts who need to be convinced that now is the time to bring their institutional infrastructure from the 20th into the 21st century? A large fraction of subscriptions are by now bundled up in so-called “Big Deals“. These Big Deals, in turn, are often negotiated by regional, national or supra-national consortia. These consortia constitute existing collaborations between institutions in order to solve an infrastructure problem: paying for increasingly overpriced subscriptions. There are currently about 200 of these consortia organized in the “International Coalition of Library Consortia“. Hence, one would have to convince the leading individuals at only a few hundred organizations that now is the time to shift their infrastructure from subscription-based solutions to modern solutions. Clearly, this would not cover all subscriptions, but it would be so widespread, that every institution on the planet would contribute in some way or another and hence all institutions would at least receive notice of the upgrade. Starting with these consortial Big Deals should bring in several billions (US$ or €) annually, which would easily cover implementing the more advanced components one would expect from a modern scholarly infrastructure during this transition period. Any individual researcher who would still like to read a subscription-based journal (who would submit to these journals to make them worth reading?) can still subscribe with their own funds, just as they can still write their manuscripts on type-writers or dictate them to their secretary, if they so choose. Subscriptions simply do not belong to the canon of solutions any more, that tax-funded public institutions can afford to subsidize.
Condensing these estimates to an (admittedly very simplified, but rather pointed) dichotomy, asking researchers to submit exclusively to OA journals entails convincing just under 8M individuals at ~10k institutions to risk their jobs, while asking librarians to let subscriptions run out entails asking maybe 1k individuals at ~200 organizations to do their job.
While I wouldn’t dare to imply that any of these two options were trivial or even just easy, from my faculty perspective, the second option seems to be the technically and logistically less difficult one.
This insight of course does not entail shifting responsibility for change away from me personally or any of my colleagues, towards librarians. Librarians and faculty are fundamentally on the same side in this issue and share the same interests in a scholarly commons. As such, there is no way around symbiotic cooperation if we strive to overcome the significant social hurdles set up by a self-reinforcing system towards a major shift in the overall research culture. We all need to change and adjust and here I have laid out just a few of many reasons why libraries are in a unique position to now take the second step – after faculty have started to boycott publishers like Elsevier, created OA journals, are submitting more and more of their manuscripts to them, created preprint servers and are constantly increasing their preprint-deposition. At this point, there is very little further action I can see faculty can take. Librarians have already embraced the open access movement many years ago by creating repositories at almost every institution, a vital component of a modern infrastructure. Librarians have a many centuries-long tradition of providing services for faculty. Now is the time for maybe their most critical service yet, completing the modernization of our scholarly infrastructure, a necessary next step to halt further commodization of scholarship and to keep working towards a scholarly commons.
The end result of this proposal would be:
1. Funding bodies give money to scientists, no strings attached.
2. Scientists do research, then transfer exclusive copyright to journals.
3. Libraries refuse to subscribe to those journals, denying their faculty legal access to essential articles.
I’d suggest a better solution is to change step 1. Neither faculty nor libraries have the power to solve this problem. The power rests with funding bodies, where the money is. The root cause is that funding bodies don’t require public access. As long as that is true, asking faculty and libraries to sacrifice their own interests out of altruism will continue to give glacial progress.
An approach that gave rapid and huge progress was the NIH public access policy. What’s needed is to expand that policy to all disciplines and all public funding in all major countries. Approaching 100% green OA would then finally make it feasible to cancel subscriptions.
Thanks a lot for your comment Thomas. I’m not entirely sure what your precise arguments are, so I hope I haven’t misunderstood any of them. Reading through my comments below, I’m also not sure if I got my points across as well as I intended to, so please let me know if I’m making any sense at all 🙂
The end result of what proposal? To kill journals by canceling all subscriptions? Are your points describing the consequence of that proposal or the current status quo?
In the following, I’ll assume in your comment you extrapolate what would happen if we canceled all subscriptions now:
WRT #1: Right now, funders have to mandate OA as grant recipients feel forced to publish TA. If TA ceases to exist, there is no need for such mandates any more.
WRT #2: Without journals (how would they find money to survive? Who would submit/review/edit to/for journals without any subscribers?) there is nowhere to transfer copyright to. Those publishers who haven’t transformed into service providers like Ubiquity or Scholastica will simply go bankrupt and cease to exist.
WRT #3: Like I wrote, amount of access is not really much different after you cancel subscriptions – just more convenient than now for some institutions (if they do it right). Moreover, within a few years, all back archives will be 100% accessible, as we will have enough money to, e.g., just buy all the back archives from the dying publishers. The new infrastructure takes care of publishing any new articles OA for about 2% of the current costs. Hence, even if for some institutions there would be some small decrease in access (could happen), it would likely be hardly noticeable and only for some fields. Moreover, this decrease would be offset by invaluable new time-saving services the new infrastructure enabled and it would be over within a relatively short time frame. In principle, this transition could be over in about three years, a common timeframe for a Big Deal contract. If we implemented the “legal Sci-Hub” next year, we could have cancelled all subscriptions by 2020, leaving me 15/16 years to enjoy a proper 21st century working environment before I retire 🙂
WRT your suggestion: Funders have no power over, e.g., me at all: My university provides me with staff and a budget so I don’t have to write grants any more (other than for exceptional projects). For all other colleagues: OA mandates is what, e.g. Stevan Harnad and many others in the OA movement have asked for since about 1995. Now it’s 2016 and look how far we have gotten. You mention the NIH mandate. If we chose, as you suggest, to simply continue what we have been doing for the last 21 years, maybe in another 21 years we have such or similar mandates everywhere, such that at least the fraction of colleagues that need grants are forced to waste time with posting duplicate copies to servers, while people like me still have the fortune to publish where they want, without any extra duplication of our workload. Not to talk about the growing resistance of many scholars who argue that such mandates are infringing their academic freedom. So we’re not approaching 100% green OA even if in another 21 years maybe we have gotten most funders to adopt some version of green OA mandates. Do you even need 100% OA to cancel subscriptions? With 30k journals, each academic probably has less than 50% access anyway, statistically speaking. Given that in some fields we have more than 50% green/gold OA publication already, there wouldn’t even be a difference if you cancel now or in 21 years. All these arguments don’t even include the problem of green embargoes, which in mathematics for example stand at a whopping 5 years! I’ll retire in 20 years, so that strategy, even if it worked, is definitely completely irrelevant to me and I would not waste my time on it. In fact, I won’t waste my time on any strategy which would entail me getting to use a modern infrastructure for less than ten years. IOW, any strategy that doesn’t lead to a complete elimination of subsidized subscriptions within the next 5-10 years is too slow.
Finally, why do you seem to imply that asking libraries to cancel subscriptions is against their self-interest? Haven’t libraries been the custodians of your work for centuries? What could be more in their self-interest than to remain such custodians even in the digital age, rather than sidelined and made obsolete by parasitic publishers? A modern scholarly infrastructure, based in libraries and paid for by the same funds that now pay subscriptions, is probably the one thing that will ensure the survival of our libraries.
Hi Björn,
thanks for the detailed reply. Yes, I was referring to your proposal that libraries cancel all subscriptions.
“How would they {TA publishers} find money to survive” under your proposal?
Via this step: “download individual article and pay the fee”
https://bjoern.brembs.net/2016/05/why-havent-we-already-canceled-all-subscriptions/
Your proposed system still gives away publicly-funded research, no strings attached, then buys it back. The collective insanity at the heart of the system remains intact.
Publishers also take a cut of ILL fees via the Copyright Clearing House. These options are much more expensive per download than subscriptions in almost all cases, which is why libraries still subscribe. If libraries could save money by switching to individual downloads & ILL, they would have done so.
*”Funders have no power over, e.g., me at all” – your funder, the German government, has absolute power over you. They are just not exercising it yet. Tenure doesn’t exempt you from the law. If they passed a law requiring public access to publicly-funded research, recipients would provide it.
I strongly agree that existing piecemeal mandates aren’t nearly big enough, and have come much too slowly. But they have been done, and they work. What’s needed is national mandates covering all publicly-funded research.
Given current per article pricing of about US$40 and current publisher revenue of about US$10b annually, there would have to be about 250 million paid downloads. These downloads, of course, would have to be be largely to newly published articles, as we’d be using LOCKSS to get access to all the already subscribed to content in the historical back-archives (and newer articles are probably read more?). However, there will be a few articles not covered by LOCKSS or ILL, Portico, etc. As about 25% of all 120 million currently published articles are OA right now, this leaves some unknown fraction of the remaining 90 million articles not covered. It’s probably reasonable to assume our libraries collectively covered about 90% of these articles at one point or another (otherwise the journals would never have existed, if nobody had ever subscribed), leaving 9 million articles not covered. Half of those again would be covered by Sci-Hub, leaving about 4.5 million already published articles. Given that only between 12-82% of articles are read, let’s take another 50% of these articles that people will actually want to read so much they’ll pay for them, leaving us with 2.25 million articles any one at some point would want to read and pay for. This is likely too high an estimate, as this is the total number and I’m counting it per year.
Add to that those articles that will still be published in these journals that nobody has subscribed to any longer, after subscriptions have run out. Currently, there are 2 million articles published every year. Let’s be generous here as well and assume nobody will change their publication practices at all and will continue to submit the same proportion of their articles to journals that nobody subscribes to (highly unrealistic, as the services we will provide from the saved subscription funds will make scholarly communication so much more efficient than this 17th century journal system). A recent study has found that 50% of newly published articles become accessible at or quickly after officially entering a journal now, leaving 1 million articles uncovered. Given that part of the effort will go towards making preprint submission automatic, it should be possible to reach about 90% deposition rate within a very short timeframe. This would leave your scenario with a publisher target of 250 million downloads from a total volume of about (rough, but probably more than generous estimate) 2.45 million uncovered articles, only 200k of which are actually new articles. Thus, essentially, in order for “The collective insanity at the heart of the system remains intact” to be correct, you would have to assume an average of 100 downloads for every single uncovered article that someone may potentially want to read every single year. I find this rather unrealistic. I’d find one average download per article already unrealistically high, but even if that happened, I wouldn’t call a 99% drop in revenue “intact” by any stretch of the imagination. Obviously, if the publishers raised per articles fees in compensation to ~US$4k per article, even fewer people would pay to download.
As you can see in the post about canceling subscriptions and the “oa.cancellations” Harvard tag, libraries do cancel subscriptions and save money, some of them explicitly using ILL and article downloads. However, at a larger scale, without proper use of LOCKSS, of course, ILL and download alone are often not good enough, I agree.
Of course, you are also right in another point: if we convinced politicians worldwide to force faculty to make their work accessible, access would happen. However, there are two main reasons why this is an even less attractive strategy than just convincing faculty to risk their jobs:
1. With ~US$1b in profits every year, Elsevier alone has more access to politicians (and effective $$$ access, as the RWA has shown) than us scholars
2. If (the current guard of largely neoliberal) politicians only pass laws about access, the result will most likely be APC-based gold OA (as you can hear from any politician you might want to ask, just look into the newspapers), which is even worse that what we have now. So if I would rank the three options of who to convince, from most difficult/undesirable to easiest or desirable:
1. All politicians of the planet (number? 6600?)
2. All faculty of the planet (just under 8 million)
3: ~1000 librarians
Finally, I also agree that national mandates would make work accessible, but given that access is such a tiny problem of our infrastructure, by the time you got these mandates rolled out on a world-wide scale, I will not only have retired long ago, but the remainder of our infrastructure problems will have brought publicly-funded scholarship to its knees long ago.
Thus, in summary, I cannot see how single article downloads to such a small fraction of the literature (~2-3%, tops) should ever compensate to the tune of US$10b and while your alternative strategy options might help solve the access problem, they are a) too slow to be attractive for me b) risk making the overall situation even worse than it is now as an unintended consequence.
P.S.: On bad days, I even share your cynicism that my faculty colleagues will continue to submit to journals where literally nobody can read their articles, simply to have the journal name on their CVs (like physicists and mathematicians) – and that prospect worries me greatly. I guess if that really came to pass, we deserve all the horror that’s bound to result from such behavior. 🙁