SPARC Europe response to the UK Government Department of Business, Innovation & Skills Committee inquiry into the UK Government’s Open Access policy

SPARC Europe welcomes the UK Government’s positive interest in Open Access (OA), the commissioning of the Finch report on expanding access to research publications and the Research Councils’ active engagement in developing policy.

  1. SPARC Europe (Scholarly Publishing and Academic Resources Coalition) is pleased to have the opportunity to comment on the issues below. SPARC Europe welcomes the UK Government’s positive interest in Open Access (OA), the commissioning of the Finch report on expanding access to research publications and the Research Councils’ active engagement in developing policy.The Government’s acceptance of the recommendations of the Finch Group Report
    ‘Accessibility, sustainability, excellence: how to expand access to research publications’, including its preference for the ‘gold’ over the ‘green’ open access model
  2. SPARC Europe’s response to the Finch Report is a public document and details our reservations about the direction and conclusions of the report. Before we explain these we wish it to be noted that we have serious concerns in two areas – there are factual inaccuracies in the Finch report and there are a number of assumptions or conclusions that are not evidence-based. We point these out in order to flag up points that the Committee may wish to note in its own considerations and because these contribute to our analysis and response to the Finch study and its conclusions.
  3. Factual errors. The report:a)    Implies that ‘Green’ Open Access is always embargoed (delayed by publisher embargoes on permission to put papers into repositories). This is incorrect: 60% of journals allow immediate opening-up of the peer-reviewed version of the article and a further 27% permit the opening-up of the submitted version. 16% even allow the published version (the Version of Record) to be placed in a repository immediately upon publication.b)    Reports that the impact of repositories (Green OA) is limited except in those disciplines that support large-scale subject repositories. In fact, repositories are providing seven-eighths of the 40% of the UK’s research outputs that are Open Access, and the amount of Green OA is greater than Gold OA (openly-published in journals) in every field except the life sciences[1].c)     Reports that Green OA cannot be given a liberal licence to allow text- and data-mining. This is incorrect: there are many examples of liberally-licensed Green OA material and these are growing.d)    States that most Open Access journals charge article-processing fees. This is incorrect: the majority of OA journals (70%) do not charge a publication fee.e)    States that publication fees are paid by authors. This is misleading, since most publication fees are paid by the institution (24%), by the funder (59%) and only in 12% of cases by the author. There are implications here for the public purse.
  4. The lack of an evidence base for some of the important assumptions made, and some analytical errors: a)    Concludes that Green OA with short embargo periods may damage publishers. This ignores the complete lack of evidence for this and the substantial body of evidence against it, which we detail below in this response.b)    Concludes that repositories have a role merely in preserving the literature and for archiving theses and grey literature. This is a strange conclusion given that virtually all repositories were built specifically to provide Open Access to the journal literature and in the UK have succeeded in providing this at way above global average levels.c)     Uses an above-average value for publication fees in the calculations of the cost to the UK of Gold Open Access. The value used in the study was £1500-2000 (based on the publication fees being paid by the Wellcome Trust for authors to publish in Gold OA journals in the life sciences). The global average is USD 907 (c£570)[2]. This is important because the overall Finch recommendations are for the UK to move to all-Gold Open Access (i.e. pay for every article published) and Government-level acceptance of such a high value for publication charges may lead to publishers putting their charges up to this level if they feel the money is there.d)    Does not take into proper account the existing policies from governments and research funders. There are over 50 funder policies[3] on Open Access, all of which align except for detail and all of which have a primary focus on Green Open Access using repositories. The Finch conclusion – to focus on Gold Open Access – is an outlier (though the aim is, apparently, to ‘lead the world’).e)    Assumes that potential users will want the Version of Record (VoR). Potential users may prefer the VoR but this requires Gold OA for all articles – an expensive option for the UK. For most potential users, the Green version is perfectly adequate.f)     Assumes that innovative SMEs all need liberally-licensed research outputs so that they can carry out text-mining. While this is the case in some sectors (e.g. biotechnology), for the majority of sectors the need is simply for humans to have access to articles to read them.
  5. The main reservations we have with the conclusions of the Finch report are these:
    • The Finch study is disappointing in not focusing properly on providing Open Access but instead on providing Extended Access
    • The UK is already in the forefront in providing Open Access, but the way it has achieved this success is neither recognised by this study, nor used to build further similar progress
    • Recommending paying for ‘hybrid’ Gold Open Access without securing full re-use rights in exchange for the payment is not the advance we were hoping for, and will hamper academic and commercial research progress. To some extent this problem has been overcome by the subsequent RCUK policy
    • The overall recommendations are out of line with those of every other policy from every other nation, region or institution. This is not brave leadership, but a serious mistake. It is very disappointing given the pattern of initiatives elsewhere in the world focused on achieving Open Access in an aligned manner, and especially in the light of the UK’s commanding global position on Open Access.
  6. Given these inaccuracies, SPARC Europe cannot fully endorse the conclusions of the Finch report, nor support the specific shape of the subsequent RCUK policy based upon the report. No national policy should be based on flawed or missing evidence and we exhort the Committee to pay attention to the facts when considering what its own recommendations might be.
  7. There is a considerable body of opinion that feels that the Finch study gave overmuch weight to publisher fears (evidence-free) about their businesses and sought to promote a route to Open Access that put publisher interests above those of the research community and the public that funds research. SPARC Europe concurs with this in general, though our view is nuanced: we believe that it is possible to develop a policy position that achieves Open Access within a reasonable period, supports innovative publishing services and new business models, delivers a competitive market and returns value to the taxpayer. What the Finch report concluded and recommended does none of these well.

    Rights of use and re-use in relation to open access research publications, including the implications of Creative Commons ‘CC-BY’ licences

     

  8. Openly accessible publications may be specifically licensed to permit barrier-free re-use for multiple purposes, including text-mining. Text-mining is a technology that is proving to have particular potential in applied biomedical and chemical research, though it is likely to be increasingly important in most disciplines.
  9. Where publishers are paid for Gold OA they are selling a publishing service. Purchasers of that service should select the level of service that they deem appropriate and where public money is being used, as in the case of the RCUK’s intention to pay for Gold OA, the purchaser should decline to pay unless the product fully suits its needs. RCUK should require the form of licensing that it deems appropriate (it is minded to require a CC-BY licence) in return for Gold OA payments.
  10. It should be noted – and was not in the Finch report – that Green Open Access material can also carry a liberal licence such as CC-BY. Increasing amounts of Green OA material is licensed in this way[4]: liberal licensing is not limited to Gold OA.

    The costs of article processing charges (APCs) and the implications for research funding and for the taxpayer

     

  11. The Finch report’s calculations on the cost to the UK research budget of moving to an all-Gold Open Access scenario used relatively high article processing charges (APCs) of £1500 or more, based on the charges that the Wellcome Trust currently pays publishers for articles from its funded research in biomedicine. Biomedical journals have the highest APCs of all disciplines. The real current average charge levied across all OA journals that charge APCS (most do not) is £571[5].
  12. By using a high figure, the Finch study raises the expectation of what the UK might be prepared to pay in APCs, opening up a potential distortion of the market as publishers with moderate or low APCs raise their APCs in line with that expectation, and exposing the UK’s research budget to unnecessary and certainly unwarranted demands with respect to publication costs. This move is a grave disservice to British research and to other nations whose researchers will also to have to pay any higher APCs levied by publishers.
  13. Moreover, the Finch recommendations and the subsequent RCUK policy do not differentiate between fully OA journals and ‘hybrid’ Gold Open Access: they take the view that both are equally beneficial and APCs from either are eligible for payment by UK research money. Hybrid Gold Open Access may be viewed as a mechanism for transition and clearly is as far as the Finch study goes. The alternative view is that it may be a mechanism that will actually slow progress – and cost a great deal in the process – as publishers delay converting their journals to full OA so as to enjoy for as long as possible revenues both from subscriptions and from APCs for a proportion of articles (a process known as ‘double dipping’ into the public purse).
  14. To make matters worse, the Finch report does not contain any assurances that hybrid Gold access will be provided with liberal re-use licence conditions in return for the money paid for it. RCUK is apparently trying to make suitable arrangements with publishers but compliance with RCUK’s demand that where it is paying for Gold OA then the articles will be CC-BY-licensed is by no means guaranteed: publishers have good business reasons to fear such liberal licensing terms and some, or even many, may resist it for some time to come. This is not a good deal for UK research, innovation or taxpayers.
  15. Conversely, pure Gold Open Access is now no longer an experiment. Full Open Access publishers have become mainstream players, demonstrating that business models to deliver pure Gold Open Access can be successful and sustainable ones for journal publishers, including in the humanities (and including for monograph publishing, for which there are now some workable and innovative new business models in play). These publishers have also shown that it is entirely possible to use open licensing without this threatening their business: in short, they are facilitating the kind of dissemination that is optimal for research and for society and are making a commercial success of it, creating jobs into the bargain.
  16. In addition, this model has the potential to introduce proper competition into the market, which will result in better processes at lower cost to the taxpayer. Directing substantial amounts of research funding into hybrid Gold publishing will thwart this competitive drive and undo a lot of good progress: if money is available to pay for Open Access publishing, we suggest it should be confined to paying for the real value created by pure Gold Open Access.The level of ‘gold’ open access uptake in the rest of the world versus the UK, and the ability of UK higher education institutions to remain competitive
  17. All this said, we do question the over-weighted ‘Gold’ thrust of the Finch report and the RCUK policy. Economic modelling[6] has indicated that the cheapest way for the UK to become an ‘Open Access nation is to transition through ‘Green’ OA to a final ‘Gold’ OA world. The general benefits of OA (greater usage and impact of research, better reach to SMEs, professional, practitioner and education sectors and the general public) accrue throughout the transition, but the costs of transitioning are much less. This delivers value back to taxpayers at the same time as achieving the goal. We do not understand the thinking behind spending precious research funds on publishing activities when OA can be delivered more cheaply and as effectively in other ways. There is a policy opportunity with respect to Gold OA, especially the opportunity to help create a better market for publishing services, but we do not agree that the RCUK policy approaches it in the best way.
  18. Some 40% of the UK’s annual research outputs are already freely available, a far greater proportion than the world average (20%). The Finch study does not acknowledge at all the fact that the vastly greater part of this content is provided through ‘Green’ Open Access, delivered through a nationwide network of interoperable repositories that is second to none in terms of technical and policy advances. Considerable investment in this arena has been made both by research-focused institutions (all of which have Open Access repositories) and research funders (some of which have their own repositories), and policy developments by both have resulted in an exemplary leading position for the UK in this respect. Yes, there is more to do here, but it is policy development and implementation that needs attention – and that comes free.
  19. The latest data show that 35% of the UK’s research outputs are freely provided through the Green route (compared to 5% through the ‘Gold’ route)[7]. And the UK has more policies that are ensuring this Green progress than any other country, size for size.
  20. It is mystifying, then, that the achievement of UK repositories and their promise for the future is completely unacknowledged in the Finch Report, and that further building on this foundation is not recommended. Instead, repositories are consigned, with little further consideration, to a role as providers of access to grey literature and data, and of preservation services. While repositories are certainly flexible enough to fulfil these roles, and are already doing so to some extent, the primary imperative for their establishment in every case has been to provide Open Access to the research literature. With proper policies in place, repositories can deliver Open Access extremely effectively.
  21. We would like to see explicit recognition of that fact and of the potential that repositories embrace, and strong support for repositories in furthering the aim for increased accessibility. We note also that the report suggests that Green Open Access always involves embargoes, which is untrue (see paragraph 3a). Embargoes are not an inherent characteristic of the Green system: they are publisher-inflicted, where they occur, and can be rectified by proper attention during policy-making to copyright and the timing of deposit.
  22. The report also suggests that this type of Open Access cannot provide liberal re-use; also untrue. Most repository software packages offer the depositor a choice of licensing terms to use for the deposited item. It is perfectly possible for authors to elect to use a liberal licence: that many do not yet opt for this is not an intrinsic flaw of the Green route but a result of lack of awareness and understanding, something that can be remedied over time by careful advocacy.
  23. On the specific issue of Gold OA in relation to policy, our analysis of worldwide Open Access policies of research funders is presented in summary form in the table below (we have not included our institutional policy analysis here because it is out of scope). This is an analysis of the 49 funder policies from the Registry of Open Access Mandatory Archiving Policies (ROARMAP) that have been formally implemented. It is clear from that analysis that the UK (RCUK) stands alone in policy terms in requiring Gold OA where it is offered by publishers. In every other case, where Gold OA is explicitly part of a policy it is included as a choice for authors, not a required option. This includes the forthcoming policy from the European Commission for Horizon 2020, the framework programme for 2014-2020.  Forthcoming policy from the US to cover federal agencies’ research will be either the same or will be Green-only.
  24. Type of policy

    Number of policies

    Green OA mandate

    36

    Green OA mandate with a Gold OA option (payments for APCs allowed from grant or claimed from funder)

    12

    Gold OA mandate with a Green OA option

    1*

    *RCUK

  25.  This independent policy stance by RCUK could be seen as a brave lead, or it could be seen as a lonely walk into the unknown without any understanding of the likelihood of finding any other nations marching in step.
  26. Whichever, it is going to be a costly business for UK research-performing institutions and there are problems and dangers in the implementation. Early estimates from large research-intensive universities suggest that additional staff will be needed to manage the system of block grants that RCUK intends to award to institutions to pay for Gold APCs. Researchers are anxious about the process of allocation of funds (block grants) that RCUK admits will not be enough to cover APCs for all articles published from RCUK research. Smaller publishers, including many learned societies, are worried about achieving their ‘fair share’ of the block grant money when there will be competition for it and the large publishers are likely to offer ‘deals’ to universities in order to secure large proportions of the available funds. Indeed, the largest publisher has already made a move in this direction (see paragraph 31 below).
  27. To reiterate a simple point, the money RCUK intends to make available for APCs is coming from the research budget. RCUK has declared that it will make £100 million available for this purpose over 5 years, money that would otherwise directly fund research. What this means for the competitiveness of UK HE institutions, struggling to acquire research funding and to stay at the forefront of international research efforts, is self-evident.

    Support for universities through funds to cover article processing charges
  28. Even before the Web made the case for Open Access compelling, there was a market problem in academic publishing. The market is far from perfect because there is no proper competition between products and users of the products (researchers) are not the purchasers (libraries are) so pressure on prices is reduced and price elasticity of demand has been low. As a consequence, journal prices have risen inexorably over the decades.
  29. Open Access provides the opportunity for publishers to change their business models to provide free access to readers and to cover their costs by making a charge for their services rather than for their products. Such as system should naturally adjust the market by putting researchers who are – as authors – the customers for publishing services into the position of purchaser. This will create a more competitive market as publishers will need to compete on service quality and price.
  30. Unfortunately, the proposed system of block grants to institutions will hinder the development of a more competitive market. Authors will be one step removed from the purchase decision by dint of an inevitable allocation system within each institution.
  31. Moreover, opportunities for further monetisation of publicly-funded research are already being grabbed. One very large publisher is attempting to do a national-level deal with libraries to acquire a large share of the £10 million that David Willetts handed out to 30 institutions to ease the path towards implementing the RCUK policy of paying for Gold OA. This publisher has made an offer to the UK’s research libraries to make the RCUK-funded articles it published last year (2012: note that this has nothing to do with Open Access; it is retrospective back-catalogue access) freely available for sums of hundreds of pounds per article (varying depending on how many of the circa 3000 articles involved the libraries can or will pay for). Note that this publisher has already been paid, in the form of subscriptions, for handling these articles. Note, too, that opening up the articles is portrayed as being a complex task of ‘conversion’, one worthy of a charge of several hundred pounds per article.
  32. This example clearly illustrates that the proposed block grant approach will simply turn the current ‘Big Deal’ system for subscriptions into a Big Deal system for article processing charges. There will be many losers, just as there are from the Big Deal subscription system, not least small publishers that cannot compete with the might of the publishing giants, and certainly including the British taxpayer because a better, competitive market will not result[8].

    Other points

     

    Embargo periods

  33. It is worth noting that the Finch report states that embargoes of less than 12 months are ‘unreasonable’. It bases this on the claim by subscription publishers that immediate Green OA threatens their businesses, yet no evidence to support this claim has ever been produced. There is no evidence at all that short embargoes – or even no embargo at all – damage publisher subscription businesses (what has caused relentless attrition in journal subscriptions over 30 years is the equally relentless above-inflation increase in subscription prices).
  34. On the contrary, the two main society publishers in high-energy physics have publicly stated that they have seen no damage to their businesses from 20 years of Green OA (where there is no embargo at all) in this discipline. The National Institutes of Health (US) policy, a ‘Green’ Open Access policy that has been in place since 2007, requires authors to deposit their articles in the OA repository called PubMed Central – and the publishers of thousands of journals also voluntarily submit their own articles to PubMed Central, not something they would do if it damaged their businesses[9]. And Nature, the most prestigious scientific journal of all, says, “We have, to date, found author self-archiving [Open Access] compatible with subscription business models, and so we have been actively encouraging self-archiving since 2005.”
  35. It would seem not unreasonable then to conclude that Green Open Access does not (yet) harm subscription publishers: and the recently-completed PEER project, a publisher-led, EU-funded study to gather evidence of the effects of Green Open Access, concluded exactly that[10].
  36. RCUK should therefore allow a maximum embargo period for natural sciences, engineering and medicine of 6 months, and for humanities and social sciences of 12 months. This is still not Open Access, but it is a step towards it.Improving the levels of Open Access in the UK
  37. RCUK revised its original Open Access policy in 2012. The original policy, implemented by all seven Research Councils by 2007, was the most successful funder policy in the world in terms of the amount of OA material it was achieving. [It was not the most successful policy of all: that honour continues to belong to the University of Liege in Belgium (collecting 83% of its annual outputs in its OA repository) and, since the whole Belgian HE system is now adopting the same policy conditions in conjunction with the Belgian national research funder, FNRS, Belgium is expected shortly to overtake the UK as the country leading the world in providing OA. This is a properly joined-up national policy scenario.]
  38. RCUK can retain its lead by continuing with its original policy but enforcing it better. In 2012, both the Wellcome Trust and the NIH in the US announced that they would be strengthening the policing of their OA policies to ensure better levels of compliance. RCUK should do the same instead of changing the whole basis of the policy, confusing and alienating the UK research community and shifting research funding into further spending on publishing.
  39. RCUK was already working on technical solutions for measuring and assessing the impact of its funding programme, and this included a plan to harvest RCUK-funded outputs from institutional repositories (IRs) into a central RCUK database that would provide the material for analysis and assessment. This was a simple and elegant way for the impact of British-funded research to be assessed. Now, the reduced focus of the new policy on Green OA will mean that RCUK-funded research will not all gather in institutional repositories, as was the original plan, but will appear variously across IRs, centralised repositories such as UKPMC (now PMC Europe) and multiple publisher websites. This seems a step backwards, forcing more complicated technical processes to be developed in order to measure the impact of British research, and inevitably delaying the implementation of such a system.

SPARC Europe is an organisation that promotes change in scholarly communication for the benefit of research and society. http://sparceurope.org/
This submission was written by Dr Alma Swan, Director of Advocacy Programmes
a.swan@talk21.com   01392 879702
7 February 2013

 



[1] Data from Björk B-C, Welling P, Laakso M, Majlender P, Hedlund T, et al. 2010. Open Access to the scientific journal literature: Situation 2009. PLoS ONE 5(6): e11273. doi:10.1371/journal.pone.0011273  http://www.plosone.org/article/info:doi/10.1371/journal.pone.0011273 and from Gargouri et al, 2011 (unpublished; personal communication from Yassine Gargouri, Université du Québec à Montréal)

[2] Solomon DJ and Björk B-C (2012) A study of open access journals using article processing charges, Journal of the American Society for Information Science & Technology  63 1485-95  DOI: 10.1002/asi.22673 http://www.openaccesspublishing.org/apc2/preprint.pdf

[3] Registry of Open Access Repository Mandatory Archiving Policies (ROARMAP): http://roarmap.eprints.org/

[5] Solomon & Björk (2012), op cit.

[6] Houghton J and Swan A (2013) Planting the Green seeds for a Golden harvest: clarifications on ‘Going for Gold’. D-Lib Magazine, Jan/Feb 2013 19(1/2) http://www.dlib.org/review/dlib/january13/houghton/01houghton.html doi:10.1045/january2013-houghton

[7] Data collected for a study on behalf of the OAIG (Open Access Implementation Group), to be published shortly.

[8] Note: three days ago (04.02.2013) a preprint was published by a US economist on the adverse effects of big-publisher market power in an Open Access scenario: Odlyzko, A [University of Minnesota] (2013) Open Access, library and publisher competition, and the evolution of general commerce http://www.dtc.umn.edu/~odlyzko/doc/libpubcomp.pdf

[9] The NIH Public Access Policy. Director’s report, February 2012: http://publicaccess.nih.gov/public_access_policy_implications_2012.pdf

SPARC Global
LIBER