The third post in our small series on open access, publication shifts on the horizon and how it all matters to IR and social science, this time by David Mainwaring  (Pablo’s post was first, then Colin Wight’s, and following David came Nivi Manchanda, Nathan Coombs and our own Meera). David is a Senior Editor at SAGE with responsibility for journals in politics and international studies. So he oversees journals like Millennium and European Journal of International Relations amongst others (which is how we know him), and can thus offer a close reading of movements within the academic publishing industry. Images by Pablo.

---

Open Access is the talk of the academic town. The removal of barriers to the online access and re-use of scholarly research is being driven by a cluster of technological, financial, moral and commercial imperatives, and the message from governments and funding agencies is clear: the future is open. What is much less clear is exactly what sort of open future social scientists would benefit from, let alone what steps need to be taken in order to transition away from the existing arrangements of scholarly communication and validation. Here the conversation is in its relative infancy, characterised at this point by a great deal of curiosity, anticipation, confusion, and the shock of the new. What it needs to move towards is a recognition and coordinated response to the fact that although social science may share the same open access goal as the STEM disciplines, the motivations for travelling down that path are not identical, and the context – especially in terms of research funding – is significantly different. The roundtable discussion at the Millennium conference at the LSE on 20th October was an attempt to explore these issues specifically from an IR perspective; further such events (such as those being run by the AcSS and LSE this autumn) are to be warmly welcomed as a means of building broader understanding of the issues among social scientists and facilitating strategic thinking.

Social Science and the Open Access Debate

The long-running debate about how scholarly research communication should be funded and transmitted has been, and remains, a discussion conducted primarily by those working in STEM. Most of the blogosphere’s best-known voices on open access, the likes of Mike Taylor, Michael Eisen, Peter Murray-Rust, Björn Brembs, Cameron Neylon, Kent Anderson, Stephen Curry and Tim Gowers, have backgrounds in STEM research or publishing. (Notable exceptions are the philosopher Peter Suber, who heads the Harvard Open Access Project, and self-archiving advocate Stevan Harnad, a cognitive scientist). To dip into the often heated debates on open access can leave you with the strong impression that, despite the occasional nod to social science and the humanities, the frame of reference is proper, rigorous, natural scientific research, the kind carried out in a laboratory that leads to medical advances and the development of new technologies.

By and large social scientists – and arts and humanities scholars, to whom many of the points raised in this piece apply equally – have had a back seat in this conversation, and the development of open access awareness and capabilities has been slow. Many leading social science journals continue to be distributed in print form due to subscriber demand well over a decade after the launch of their online editions. The American Political Science Association’s 2009 book Publishing Political Science devotes just two out of over 250 pages to open access, and fewer than 10% of the nearly 13,000 signatories of the ‘Cost of Knowledge‘ boycott of Elsevier were social scientists, despite the company’s position as one of the world’s leading publishers of social science journals. That’s not to say that there is nothing going on: the Social Science Research Network has acted as a site for open paper sharing since 1994; there are active ‘open’ movements in disciplines including IR and economics; and the Directory of Open Access Journals lists more than 1,600 social science titles. To date, however, very few of the latter have been able to break into the higher echelons of profile or reputation within their fields.

Social Science and Open Access Mandates

Over the last eighteen months a series of events, including George Monbiot’s polemic in The Guardian and the defeat in the US of the Research Works Act, stirred for the first time a significant consciousness among social scientists about open access. This awareness increased dramatically – in the UK at least – with the publication in June 2012 of the final report of a committee set up by the Government to examine access to published research findings. The 17-member body included social scientific representation in the form of its Chair, sociologist Janet Finch, and committee member Adam Tickell, an economic geographer. The majority of the recommendations made by the Finch Committee were swiftly backed by the Government and transformed into policy statements by the main UK higher education funding bodies. At their heart is a clear commitment to OA backed by mandates applicable to anyone wanting to publish the results of taxpayer-funded research in academic journals. Almost at the same time the European Commission announced a similarly-intentioned package of measures relating to its Horizon 2020 funding programme, and other governments around the world have subsequently introduced mandates of various forms.

Social scientists in the UK reacted cautiously to the publication of the Finch Report. Support for the broad principles of open access was coupled with concern and uncertainty about the viability for social science of the preferred – ‘Gold’ – model of OA, and the apparent lack of a clear transition plan. The Gold model enables research to be immediately accessible online via article processing charges (APCs) paid on the acceptance of articles. UK universities will receive block grants, carved largely from the existing research budgets, and will then be responsible for their allocation to eligible research projects. Of course, this depends on two things: firstly, whether the journal the grant-eligible research is destined for offers a Gold OA option; and secondly, whether the university wishes to spend a portion of its OA allocation on that particular article. The Government kept open an alternative open access option for publicly-funded research that for either reason could not use the Gold route. Green open access refers to the deposit by the author of a post-print version of their work in an open repository, usually after an embargo period, while the ‘version of record’ remains behind a paywall.

This preference for Gold over Green has divided opinion and it puts the UK out of step with the mandating policies of many other countries, including Ireland and Australia, resulting in potential asymmetries of access. Proponents of Green OA decried a missed opportunity to further develop and invest in the capacities and infrastructures of the UK’s repositories. Certainly this route struggles currently with low levels and consistency of depositing, coupled with relatively poorly-developed international infrastructure. Perhaps more problematic is the paradox that a well-funded and adequately infrastructured Green model could well lead to the cancellation of subscriptions in the slower-burning social sciences, especially if, as the Research Councils UK has signalled, the embargo period is eventually reduced to six months from 12 to bring it into line with STEM. This loss of revenue, as I set out below, could have significant consequences for the operation of peer-review and other activities, and needs to be factored into the conversation.

The UK Government’s preference for APCs is based on the fact that the ‘Gold’ model is felt to be the one most able to maintain the ‘high quality of services’ provided by the current system while allowing for immediate and permanent open access to publicly-funded content. In addition, by mandating a CC-BY licence for APC-funded articles (as opposed to a CC-BY-NC licence for papers made available via the Green route), the Government is deliberately opening up research for re-use and commercial exploitation through activities such as text-mining with the goal of delivering broader social and economic benefits.

The STEM disciplines already boast a successful open access culture based largely on variations of the Gold model. Within this there is much experimentation and great diversity, with many different models operated by both traditional publishers and relative newcomers such as the not-for-profit Public Library of Science (PLoS). These include the $99 pay-to-play model being pioneered by PeerJ; the $1,350-a-go peer-review-light of the hugely successful PLoS One (which generates a financial surplus on 70% acceptance rates); the traditional, highly selective peer-review services offered by higher priced ($2,000-3,500) APC journals such as PLoS Biology or the BioMed Central family; or the CERN-backed library consortium funding model that has replaced subscription fees for a small cluster of high-energy physics journals (the SCOAP3 project). All of this is supported by generous levels of research funding, from sources both public and private (roughly ninety pence in every pound of RCUK funding goes on STEM disciplines, and social science certainly cannot compete with the sizeable private funding reserves of the likes of the Wellcome Trust).

In light of these funding disparities, can any of these different models offer a solution for social science? As Martin Coward among others has pointed out, the Gold route in the form and price point set out in the June announcements provides at best a partial solution for the HSS disciplines where average journal may have fewer than 10% of accepted papers supported by Gold-enabling funding (at the more critical or theoretical end of social science, this number may be closer to zero). This funding culture, coupled with the stiffer challenges involved in demonstrating social scientific impact, has led to fears about the relative lack of attractiveness of social science to university administrators allocating their pots of OA-mandated cash. Fears about marginalisation go alongside more generally applicable concerns of a Gold culture: the additional strain on static research budgets during the transition from the subscription model; the costs to universities of administrating APC budgets; and the impact on early career, interdisciplinary or independent researchers.

The hope of the RCUK is that over time the transparency of the Gold pricing structures will have a deflationary effect, with APCs falling from their current levels in accordance with market forces and with librarians able to negotiate their remaining subscription costs proportionately downwards. Yet it seems equally likely that the opposite may be true for the leading journals in any field, where prestige will allow them to set their prices, while the rest race for the bottom to attract custom, potentially damaging the quality of peer review. Overall, while the Finch Committee report acknowledged many of the issues affecting the social sciences (and, perhaps to an even greater degree, the arts and humanities), its primary concession in a Gold-dominated approach was to allow publicly-funded HSS research an extended Green embargo period of 12 months with the expectation that in the medium-to-long term APCs would prevail right across the research spectrum.

The Open Access Future in the Social Sciences

If some form of open access to research output is the goal for social science, which option should be chosen and how can it be achieved? How do we get from where we are to a new model of research communication with minimal collateral damage to the rest of the academic ecosystem of review, prestige, validation, branding, funding and promotion? Or should social science actively tug at that string to initiate a broader process of controlled disruption to current practices? If you’re minded that some of these other areas of academic life need an overhaul as well then the net widens to bring in impact factors, academia’s ability to communicate to external audiences, rankings and hierarchies of journal prestige, systems of peer-review, the precarity of early career researchers, the funding of learned societies, and so on. Wherever you position the discussion’s boundary rope, it is hard to escape the feeling that social science is being swept along in a game whose rules have been set by the STEM community. It wants the same goal, but doesn’t have the money to play. So what’s right for social science? The answer probably lies beyond straight Green or Gold, and will necessitate sector-specific innovation rather than the adoption of an off-the-peg solution from STEM.

One place to start would be to consider what the current subscription model provides for social science. (Which, to place it in a financial context, costs in the UK around 10-15% of the total RLUK-reported £190m library spend on journals and databases, based on the relative number and pricing of social journal journals.) The list below is probably not exhaustive and the items on it may not be viewed as ‘positive’ from all angles. Furthermore, it’s possible to imagine how each of the points below could be achieved through or another variant of open access. But it’s a worthwhile means of thinking through the desired components of an open access model that works for social science:

Revenue: This is at once the most obvious and the most thorny of issues, and it underpins all of the other points below. Money from subscriptions underwrites the editorial operations that are the heart of the scholarly communication system as we know it. While it’s true that very few journals’ editorial operations function on the basis of subscription revenue alone – the financial, temporal and infrastructure contributions of universities can be significant – very few would exist without it. Then there’s oft-discussed question of learned societies, many of whom fund their diverse networking, career development and community support activities in large part from journal-generated revenue.

Sustainability: It’s mercifully rare that a subscription-based social science journal goes to the wall. In my own fields of politics and international/area studies I can think of very few examples, and none involving top or even middle-ranking titles. Without a steady revenue base, some open access journals may find their survival resting on the support of a particular department or the input of a particular group of academics.

Ordering and Branding: Google Scholar is without doubt a wonderful thing, but the effects of journal branding, ranking and pre-publication peer-review are to remove the heavy lifting of sorting through the literature for quality and importance from the shoulders of academics. This is backed up by sophisticated and integrated platforms developed and maintained by publishers and content aggregators like JSTOR that facilitate the location and connection of research.

Production and Investment: While it is undoubtedly true that the technology that enables the online publication of individual journals is cheap, readily available and pretty simple to use, the basic work involved in converting thousands of accepted papers across hundreds of journals to polished, typeset, tagged and linked final articles is time-consuming and often mundane work. At present, this workload – along with the investment in the infrastructure of scholarly communication – falls on a third party rather than on time-pressed academics or universities.

No barriers to entry: Anyone can submit their research for consideration regardless of their seniority, discipline or funding position.

We’re not short of suggestions for alternative models. Some of these involve changes to the current subscription system, for example the introduction of price regulation, or the development of a ‘fair trade’ economy where reviewers receive payment for their part in the process. Other proposals are in some respects more radical, envisaging a future where the current arrangements have to a greater or lesser degree been ripped up and reconstituted. Among these are a raft of variations on what is sometimes referred to as ‘Platinum’ OA, in other words peer-reviewed journals that do not charge APCs, thus making them free both to submit to and to read. Proposals in this vein include the reinvigoration of the not-for-profit university press (as discussed by Pablo on this blog and librarian Debby Shorley here) and the re-conversion of learned societies into publishers, through to the out-and-out reclamation and self-organisation of the publishing system by academics themselves.

Alternatively there might be space for a social science mega-journal operating along the lines of PLoS One, with a light peer-review process to check that the methodology is sound and that the results are written up to an acceptable standard. SAGE Open and the Social Sciences Directory provide contrasting early views of what such journals might look like, using lower price points than their STEM counterparts. Or then there’s the super-repository route, perhaps boosting the capabilities of existing services such as the Social Science Research Network and using the ‘publish then sift’ model to establish relative value.

Which of these routes (or which blend of options) you favour depends on where you stand on three fundamental questions:

Firstly, what you feel the cost and complexity of publishing is/should be, and consequently how scalable the various options above are beyond the level of the individual journal or small publishing operation.

Secondly, which set of benefits you want the new model(s) to deliver beyond freely accessible content.

Thirdly, to what extent you wish to address issues around the (re)organisation of academic life beyond the specific question of research communication.

Without doubt, changes to the system of social scientific research communication will have implications for roles and relationships of all parties, including societies, publishers, journal brands, librarians and researchers. One of the major transitional questions for social science is how to ensure that money saved from subscription costs will remain within the sector and not allocated away to the bright lights of the high-impact STEM disciplines.

It is very clear that finding a workable way forward depends on collective action on an international scale, and on the ability to foster an informed interdisciplinary discussion to develop a position and voice with which to engage key players and challenge the STEM focus of the debates. In addition to government-led mandating policies there are several potential catalysts for change, including the use of national research assessment programmes (such as the UK’s 2020 Research Excellence Framework) and the development of open access polices at university level (for example that set out by Harvard earlier this year).

What these drivers can achieve for social science is currently unclear, given that the only viable route available to most researchers is self-archiving in an imperfectly-developed Green architecture. The Harvard announcement was made with much fanfare, and seems workable for many STEM researchers given that large sections of their community seem likely to transition to open access on a Gold model over the next five years. But when their IR department is looking to hire or promote, will they privilege a CV of publications in open access journals over one boasting contributions to International Organization, World Politics or EJIR? Without a concerted effort by social scientists to acknowledge, and tackle the vexed questions of funding, sustainability, validation and infrastructure development the rate and nature of transformation is likely to be incremental and highly variegated. In October PLoS Director of Advocacy and biophysicist Cameron Neylon wrote of open access that ‘…there is still much to be done and the challenges remain large, but the remaining questions are largely ones of implementation, not principle.’ Social science is certainly waking up to the principle, but with less capacity to deal with disruption it lacks a clear implementation roadmap. As a result, the ‘remaining questions’ are in fact more foundational ones about priorities, options and consequences.