As long as there has been open access (OA), there has been talk of a global “flip” of research journals away from the subscription business model. The difficulties in coordinating an enormous number of stakeholders with different interests have continued to make this unlikely. However, a recent paper from the Max Planck Digital Library claiming that, “An internationally concerted shifting of subscription budgets is possible at no financial risk, maybe even at lower overall costs,” has once again fueled talk of a flip. Has this paper discovered a golden ticket to global OA sustainability, or is it based on flawed assumptions?
Long-time green OA advocate Stevan Harnad has written at length about the improbable nature of a global overnight flip to Gold OA via an organized system of membership deals, and about the adverse selection such a system would create:
The very next day the system would destabilize, with cash-strapped institutions cancelling their “memberships” to journals that their users needed to use but in which their authors published little, preferring instead to pay for publishing by the piece for the few articles they publish in them.
This would in turn destabilize the sustainability of yesterday’s subscription revenue streams via memberships, which would mean that membership fees would have to increase for the non-defecting institutions to sustain all publishers’ net revenue, which would in turn mean that institutions would be paying more for memberships than they had been paying for subscriptions.
Much of the drive toward a flip is based in the EU and the UK, where public higher education is highly centralized at the national level. This creates the notion that there exists a global pool of funds that could be diverted away from subscriptions and toward OA fees. But the difficulties in coordinating action between self-interested parties becomes even more evident when one thinks about how libraries are funded and subscriptions are paid for in the US, still the major producer of scholarly articles worldwide.
I frequently ask US librarians where their subscription budget comes from and the responses vary widely, but the most common answers are tuition, student fees and some portion of grant overheads. Because tuition and student fees are collected by individual institutions, there’s no big pool of funds that can be diverted centrally from one purpose to another. Such a flip would massively increase the financial burden on productive institutions, while freeing non-productive institution from any responsibility in funding research access.
If I’m running a small teaching school and can save money by cancelling subscriptions, my Dean is going to be much more interested in spending our students’ tuition fees on our students, rather than sending that money off to Harvard to help their poor professors publish papers.
US universities are increasingly cash-strapped, which makes any coordinated give-aways like this unlikely. And having major contributors to the literature like the US, Japan and Australia choose the Green route puts a damper on any global move to Gold OA
But a recent paper from three members of the Max Planck Digital Library suggests the whole thing could be done immediately and at a cost-savings. Their thesis is that each individual library could stop paying subscription fees and instead divert those same funds toward article processing charges (APCs) for their campus authors, and that doing this could happen within current library budgets, requiring no additional funds from outside, and no pooling of funds between institutions.
As Rick Anderson recently pointed out, there’s a difference between advocacy and analysis. Reading this paper, it’s clear which this is. The authors clearly state that they are trying to advocate for a cause:
The purpose of this paper is to assert the necessity of a large-scale transformation of the current subscription journals to an open access business model and that this transformation can be achieved only by an equally large-scale transformation or liberation of the library acquisition budgets.
Thus this paper cannot be read as an objective piece of research. It is clearly a passionate argument meant to influence opinions, rather than a carefully designed research project with appropriate controls. Unfortunately, in order to meet a preordained conclusion, the data has to be put through unrealistic contortions.
The key problem in the analysis is in the chosen numbers used for article processing charges (APCs). The authors start with an unproven assumption — “it is safe to assume that in a predominantly open access publishing system the average article costs would not nearly be as high.” No actual evidence is offered for why this is a given.
The authors then seek a number to use as an average APC in order to calculate what a given institution would spend on publishing papers as compared to subscribing to journals. The problem with this approach is that it assumes every university hews to the average. It is likely that University of Oxford researchers publish in a different set of journals than researchers at a rural community college in the United States. Researchers at MIT may lean more toward publishing in science journals as compared to researchers at a liberal arts college like Davidson. Stanford authors may prefer journals with higher rejection rates and production values, hence higher costs. Yet by assigning the average cost for all universities to an individual university’s expenditures, it is assumed that all researchers at all institutions are identical.
If Library A spends $1,000 per OA paper and Library B spends $4,000, the average spend is indeed $2,500. But Library B is still spending $3,000 more per paper than Library A.
To get the number used (€2,000, or approximately $2,248), the authors look at data from German universities, the Austrian Science Fund and the SCOAP3 Coalition. Data from the Wellcome Trust is mentioned, but dismissed because it does not fit the chosen narrative.
Wellcome’s numbers don’t count because they include author spend on hybrid OA journals (apparently 75% of Wellcome’s authors prefer hybrid journals to fully-OA journals). Given that the hybrid journals chosen are often high-end journals, with high rejection rates, the resulting APC spend is higher than for policies where publishing in hybrid journals is not allowed. To come up with a lower average APC, the authors have to dismiss the notion of author choice in where they publish their work.
In order to make their numbers work, the authors describe not a simple flip to Gold OA, but rather a complete reordering of the culture of academia. We know that higher-end journals cost more to produce. We know that eLife spends around $14,000 per paper, and that PLOS continues to lose money on their selective journals despite charging authors nearly $3,000 per paper.
The proposed fully-OA world described in this paper would require the elimination of these high-end journals, a fact that the authors simply brush aside with little discussion, which is rather shocking given the centrally-important issue of author preference, the continuing importance of journal brand as a filtering mechanism and the degree to which academic career structure remains dependent upon the publication process. One may certainly object to any of these realities, but pretending they aren’t there is just absurd.
The underlying assumption of this argument, then, is that flipping to universal OA is easily achieved: all it requires is a complete cultural and administrative restructuring of academia. But even if this cultural revolution could be delivered, the numbers still may not work, as evidenced by the announcement this week from PLOS that publication costs are spiraling upward, resulting in a necessary increase in APC rates for even the megajournal PLOS ONE.
Magical thinking in the guise of objective analysis does no one any good. For those looking for real-world, practical ways to improve access to knowledge, it sets up unrealistic expectations; OA efforts based on such unrealistic assumptions will inevitably fail. That hurts everyone — advocates, publishers and readers alike — all of whom will have to deal with the unintended consequences of any policies based on flawed research.
Access to the research literature continues to improve. Every day, more people can access more information than the day before. The problems haven’t all been solved yet, by any means, but the progress currently in evidence is a good thing. There are all kinds of interesting experiments going on throughout the scholarly communications world, new business models, new ways of publishing research. Let’s continue to experiment and find new ways of improving what we do, driving real world progress rather than hoping for magical overnight solutions.