Readers of The Scholarly Kitchen may have noticed a recent guest post on the blog of the Higher Education Policy Institute (which should not be confused with the Institute for Higher Education Policy or the Higher Education Research Institute) written by three administrators from the University of Oxford: Patrick Grant (Pro-Vice-Chancellor for Research), Tanita Casci (Director of the Research Strategy & Policy Unit), and Stephen Conway (Executive Director of Research Services). Titled “Open Excess: Remove Open Access Burden from REF,” the piece argues that “the latest REF [Research Excellence Framework] proposals for Open Access… depart from the generally positive direction of the forthcoming REF 2029.” How so? “The proposed [open access] OA requirements for research outputs risk being unaffordable and excessively bureaucratic, and will bias the selection of outputs that we submit for assessment.”

Importantly, the authors acknowledge the possibility of providing “pragmatic answers to the technical questions to try to contain or selectively delay some of the excessive burden of implementation.” But they urge serious consideration of a deeper question: “Why… have a REF OA policy at all(?)”.

A young businessman with an overstrained expression his face is holding an oversized stack of coins on the back.

The authors raise four concerns about the proposed policy:

  1. “The costs and the bureaucracy of implementing the expanded policy will be unaffordable.” (They say compliance would require “processing and checks on upward of 15,000 separate outputs” annually at Oxford alone, and estimate the institutional cost of that work at £20M.)
  2. “OA costs are not commensurate with value.” (“An output that is OA is no more excellent than one that is not.”)
  3. “A focus on OA detracts from opportunities to promote open research practices.”
  4. “REF OA rules unhelpfully distort the REF itself (by constraining) the pool of eligible outputs that institutions can put forward.”

While all of these arguments will, I hope, become grist for interesting and useful discussions about REF 2029 and the future of government-mandated OA in the UK and elsewhere, I was particularly struck by the first one, which surfaces an issue too frequently elided in debates about OA mandates: the cost of enforcement.

Two things are important to note here.

First, the authors of this post oppose neither OA nor OA mandates (they explicitly recognize “our national obligations to OA”). So they’re not saying the “OA burden” shouldn’t exist; they’re saying the burden should be borne by others. Their position is that OA requirements “are more logically driven through UK funder mandates attached to individual research projects” than by REF requirements imposed on research institutions. They assert that “the removal of an OA policy from REF 2029 will, in a simple single act, excise complex and costly bureaucracy for all concerned” – but of course, their proposed solution does not “excise” the costs; it reassigns them. If the responsibility for monitoring and enforcing compliance is shifted from the funded institution to the funding agency, then the costs of a monitoring and enforcement regime move with it. Now obviously, in the proposed scenario a funder could reduce bureaucratic cost by taking a casual approach to enforcement, making compliance mandatory in theory but enforcing it only gently or sporadically in practice – in other words, reducing enforcement costs by reducing enforcement. This “nonmandatory mandate” approach is available to funding agencies; it would not, however, be available to research institutions if they are charged with enforcement, and therefore must be concerned about the compliance of all their potential REF submissions.

Second, this situation raises an issue that has not, to my knowledge, yet been seriously addressed in discussions of mandatory OA policies: even if (as many argue) the cost to the system as a whole of universal OA publishing would be no greater than the cost of toll-access publishing, the cost of implementing and managing mandatory OA policies – in other words, the administrative cost of removing choice from authors – is an entirely new one, conjured by the requirements of ideological orthodoxy. On the other hand, to the degree that authors’ agency is preserved, these costs don’t arise because no enforcement is required.*

And all of this, of course, leads logically to a truly simple (and, I think, admirably effective) way of actually avoiding these costs altogether: leave authors free to decide for themselves where and how they will publish their research results. Effective coercion is expensive. No compliance requirement means no need for enforcement – and therefore real cost savings.

Effective coercion is expensive.

Up until fairly recently, research institutions, funding agencies, and governments have generally maintained a position of neutrality on the question of where and how research may be published. While disciplinary norms have certainly constrained the choices of individual tenure-seeking researchers, such norms have generally been self-imposed; it’s not the government, the funding agency, or even the institution that says “You must publish in Journal X in order to get tenure,” but rather the researcher’s colleagues and peers who will pass judgment on her tenure or promotion bid, based in part on their assessments of her publishing performance. Say what one will about that system, it at least requires no costly enforcement bureaucracy; “enforcement” of these norms is, for better or worse, an integral part of an existing and long-established tenure and promotion process. (Though there are unusual examples of countries, notably China, that have imposed centrally defined publishing criteria for their faculties.)

In other words, the costs discussed in the Oxford administrators’ post arise entirely from abandoning a position of institutional neutrality on publishing protocols, and would be eliminated by preserving authors’ agency, letting them exercise their own judgment in deciding how, where, and under what conditions to publish. Of course, allowing authors to make their own publishing choices would undermine the goal of a universal transition to OA. But maybe it’s time to balance more rigorously the real-world benefits of such a transition against the real-world costs, rather than simply insisting upon its necessity on an ideological basis.

I suggest that what needs to be rethought is not the locus of enforcement burden, but the fundamental idea of a national (or global) “obligation to OA.” Maybe, just maybe, OA is not the One True Religion of scholarly communication, but rather a model of publishing that – like any other model, including toll access – presents a mix of costs and benefits, of upsides and downsides, and should therefore be treated as one in a panoply of useful but imperfect publishing models, all of which can and should be permitted to flourish in a diverse ecosystem of scholarship.


* Theoretically, if universal mandatory OA were to create genuine cost savings within the system, for example by eliminating shareholder profit from the equation, those savings could be redirected to enforcement without additional marginal cost – but here it’s important to bear in mind that much of the money currently fed into the scholarly communication system has its origins in academic institutions that might well see such savings as an opportunity to do valuable and important things other than either underwriting or enforcing mandatory OA schemes. A lot of the money that is currently “in the system” is there precisely because of the current characteristics of the system; change those characteristics in fundamental ways, and it’s entirely possible the flow of money into it would change too. Large research libraries, for example, don’t typically have multi-million-dollar collections budgets because their host institutions want them to invest millions of dollars in OA publishing programs; they have those budgets because their host institutions want access to high-quality scholarly content, much of which is (still) expensive.

Rick Anderson

Rick Anderson

Rick Anderson is University Librarian at Brigham Young University. He has worked previously as a bibliographer for YBP, Inc., as Head Acquisitions Librarian for the University of North Carolina, Greensboro, as Director of Resource Acquisition at the University of Nevada, Reno, and as Associate Dean for Collections & Scholarly Communication at the University of Utah.


6 Thoughts on "Oxford Administrators Want OA Policy Removed from REF 2029. I Have an Even Better Idea."

Rick, I am a bit confused by a piece of your analysis here re shifting rather than cutting costs. Is it the case that funders in the UK have no OA policy compliance monitoring except what happens through the REF? That would surprise me to learn (though there are indeed many things that surprise me about OA policy!).

What the Oxford administrators are objecting to specifically is the REF OA policy, which puts the responsibility for compliance management on institutions that will be submitting research outputs for REF consideration. I don’t know whether UK funders have any other compliance regimes in place.

It seems to me that what this all comes down to is the very foundation of economic theory: there is no free lunch! Someone has to figure out just who should pay for that which has little demand in a small market. So far what has been learned from this 20 or so year experiment is that publishing is expensive, that a few want to read a few articles, and that enormous amounts of time which equates to money is being spent to solve the costs associated with publishing an article. It is interesting that not much time is spent discussing the subscription model. The subscription model problem regarded price which was and is being addressed by the market.
I have not seen data on the impact of OA on the advancement of science. We have seen a growth of OA journals and the reallocation of funding away from research to author fees.
I wonder would OA journals survive in the subscription model world?

Just to be clear, though: what’s at issue in this particular case is not the cost of publishing. It’s the cost of enforcing an OA requirement.

I don’t find it surprising that individual institutions try to offload the burden of OA to some other authority: they would, wouldn’t they. Unfortunately, however, someone (or some authority) has to check for OA, because if left to their own devices, most authors would not bother to ensure their content is OA. Tenure and promotion might be highly motivating for academics, but ensuring their content is OA is, sadly, not. That’s why institutions spend time and money (in the UK, via institutional repositories) chasing academics to try to ensure their content is visible, even though this was a condition of their funding. “Letting authors exercise their own judgement” has not resulted in publicly-funded research being consistently visible to the community. The Nelson/OSTP Memo recognizes the same problem exists in the US, even if the specific solutions proposed may be different.

“Letting authors exercise their own judgement” has not resulted in publicly-funded research being consistently visible to the community.

Indeed. And of course, this begs the question I raise in my final paragraph: given that every mode of publishing (including OA) offers a mix of costs and benefits, to what degree should we be trying to enforce universal OA, and to what degree should we be allowing other modes of publishing to exist and flourish alongside it? In other words, to accept that all authors are obligated to publish openly is to say that no other kind of publishing should be allowed. I’m not sure that’s a premise we should accept uncritically.

Comments are closed.