Following the publication in June last year of ‘Finch’, the report on Accessibility, sustainability, excellence: how to expand access to research publications, prepared by a mixed group of researchers, publishers and librarians under guidance of Dame Janet Finch, CBE, we suddenly saw many scholarly publishers, formerly known for their reluctance towards the concept of open access (OA) publishing, embrace the concept as a sort of ‘egg of Columbus’. Open access suddenly became the salvation of scholarly publishing, and hybrid journals would pave the way to this unexpectedly golden future.
In this article, some of the findings of the Finch Report will be explored from a very practical point of view, especially highlighting the financial risks for universities and university libraries when embarking on the lifeboat of ‘hybrid OA’. This contribution builds on a presentation I gave during the European ICOLC-conference that took place from 15–17 October 2012 in Vienna.
Why were these publishers (and by the way, the British government) so happy with ‘Finch’ and why was this considerably less so in the case of the libraries and the universities? Well, at the end of the day, it is of course, as usual, all about money.
Prior to answering the question, I should like to emphasize that the costs of subscriptions – and in the OA model, the costs of article processing charges (APCs) – do not compensate for the costs of research, nor for the costs of peer review itself, but just for the costs of making a qualitatively authoritative selection of peer-reviewed and peer-approved articles available in a recordable version. This may seem like forcing an open door, but again and again one still reads that universities (e.g. funding bodies, governments or taxpayers) have to buy back their own research, because researchers are stupid enough to give their valuable assets for free to a greedy publisher. This is nonsense for more than one reason. Universities (at least research universities) in our part of the world usually spend about 1% of their overall budget (and therefore an estimated 3% of their research budget) on ‘acquisition’, not only for journals, but also for books and bibliographic and other databases. By doing so, they not only buy access to those products for their faculty and their students, they also contribute to maintaining a competitive and authoritative platform on which their own research can be published and disseminated worldwide.
“Open access suddenly became the salvation of scholarly publishing …”
So why did the publishers embrace the conclusions of ‘Finch’ so wholeheartedly? I believe this was because this was the very first time that a commission of this stature, representing almost all major stakeholders in the value chain of scholarly publishing, so unambiguously made clear that open access shall be ‘gold’ open access, respecting and even defending the interests and the claims of the scholarly publishers in the debate, and almost rejecting the claims of the ‘green’ open access lobby – in my opinion, rightly so. This recognition of their role in the process and the recognition that that role is vital, inevitable and indispensable for the prosperity of research make it much easier for them to open their minds and to recognize for their part that knowledge and information want to be free. Herman Pabbruwe, CEO of Brill Publishers in The Netherlands, and host of a meeting of Dutch publishers, librarians and policy makers with Dame Finch in November last year in Amsterdam, stated in his introductory speech that open access to scholarly publications is unavoidable, maybe not so much for moral or economic reasons, but because modern technology makes it so natural. I believe he is right. Especially in an academic climate in which intellectual freedom and free exchange of intellectual insights is so common, so characteristic for the species itself, it is hard to believe that ‘information’ in the long run will not liberate itself from the barriers to access. So indeed publishers had better embark on the same trip …
“… it is hard to believe that ‘information’ in the long run will not liberate itself from the barriers to access.”
But now about the costs and why (apart from the green OA movement) some librarians were less enthusiastic in their reception of the Finch Report. Even ‘Finch’ admits that British higher education (HE) will suffer from price increases under a gold OA regime. It is not difficult to understand why that is so. Scholarly information in some form is used by a wide array of organizations and individuals that do not contribute (or hardly so) to the production of scholarly information: lawyers, doctors, engineers, consultants, bankers, pharmacologists, amongst others. Today, they still contribute to the costs of the dissemination of knowledge. In an OA regime, they will stop doing so, meaning that all the costs will be loaded on the shoulders of the producers of knowledge (who already have to bear the costs of the research from which the knowledge has arisen). Finch calculates that this will take an extra £40m to £50m per year in the UK, depending on the level of APCs and on the uptake of OA in both the UK and the rest of the world. It is not quite clear if this, according to Finch, is merely a transitional item or that these additional costs are considered as structural. But unless the forces of competition are much more effective in an open access regime than in the current toll access regime, and will bring down the margins on scholarly publishing dramatically, one can easily see that in an open access regime, higher education will pay the price for all those who will stop paying a price. What's more, this price increase will not be divided evenly over all HE institutions, but will gather cosily at the country's most prominent research universities.
“… higher education will pay the price for all those who will stop paying a price.”
From toll access to open access
Publishers, in their euphoria about their newly found solution to the serials crisis, claim that hybrid OA will help us to make the transition from toll access to open access. For if they compensate their (increasing) income from APCs in the (decreasing) price of the subscriptions, we all play even, don't we? Well, actually, we don't. The only ones who play even are the publishers – if we rely on their honesty in living up to their promises (and why would we not give them the benefit of the doubt …?). Many librarians, especially those at non-research universities, may feel a light breeze of moderate price decreases. Librarians at research universities will not even feel that breeze, because they will be blown over by the storm of increasing APCs they (or at least their universities) will have to pay. Of course, this has to do with the uneven division of gains and losses that result from the New Order. And this is even more so in the transitional period during which the New Order is only gradually and slowly replacing the Old Order. A little arithmetic will help to make my point clear. Let's take the (real life) example of the Dutch research universities and their consortium, UKB.
The 13 Dutch research universities in the UKB consortium all belong to the top 100 or 200 of the world (although none of them can compete with the best British universities). They serve some 250,000 students and employ some 30,000 faculty. They spend around £50m on acquisition, of which about £35m for journals. They publish an estimated 30,000 scholarly articles a year, about 25% of the British scholarly production. Indeed, the scholarly scene of Dutch and British research universities on a national level is pretty much equal, the Dutch part being 25% of the British, which quite nicely reflects the difference in our total populations.
“The wider the gap between a university's own performance in OA publishing and the rest of the world's, the bigger the problem.”
If we take the average APC of £1,450 that Finch et al have used for their modelling as a starting point, what would happen? If Dutch universities published their articles in an all OA world, they would spend £43.5m for journal publications, about 25% more than they spend currently. If Dutch universities published 10% of their articles in open access, keeping pace with the rest of the world, the following would happen. First, the universities would keep paying 90% of their subscription fees of £35m = £31.5m. Second, they must pay 3,000 x £1,450 = £4.35m. The sum is only slightly higher (2.5%) than the current spends on just subscriptions. The real problems occur when Dutch universities publish a significantly larger portion of their articles in OA than other countries; in that case, the decrease of their subscription spends will not keep pace with the increase of their spends on APCs. Suppose Dutch universities publish 50% of their output OA, while the rest of the world stops at 5%. Subscriptions then will take £33.25m, APCs will take another £21.25m, totalling £54.5m, or 55% over current spending on subscriptions. The wider the gap between a university's own performance in OA publishing and the rest of the world's, the bigger the problem. Apart from other organizational, administrative and transactional problems that eventually will surround large-scale application of APCs, this financial discrepancy will prove to be absolutely prohibitive for making the progress in enhancing access that the Finch report is aiming for. Even if the British government catered for the full £60m or so Finch is calculating as additional costs, it would not be enough by a long way to cover the extra costs that research universities will be confronted with in the case of hybrid OA.
Two ways out
Given the desirability for whatever reason to change the business model for scholarly publishing from toll access to open access, how can we get round these financial cliffs?
First of all, we (universities, librarians, researchers, government) must accept that OA is substantially more expensive than subscriptions for HE in countries that contribute above average to the production of scholarly information. However, as Finch stipulates, it is (in the UK) after all just a marginal effect of some £40m on an overall research budget of at least £5,500m. This would imply that the costs of dissemination would increase from about 3% to about 4% of the overall research budget. If we really ‘believe’ in open access, then we should accept this as a fact of life. But we should stop promoting OA as a solution to the serials crisis, as some librarians and other OA advocates still do.
“… we should stop promoting OA as a solution to the serials crisis …”
To avoid hybrid OA disrupting the universities' budgets more than necessary, we must furthermore discuss with the major scholarly publishing houses a connection between the costs of subscriptions (licences, big deals) and the costs of hybrid OA publishing. As far as I can see, this connection may essentially take the form of one of two models.
In the first one, the licence that gives (toll) access to the journal or package will include a clause that allows researchers employed by the licensee to publish their articles and contributions for a heavily (75 – 90 %) discounted APC in that journal or package. Of course, we should accept that in this case, the price of the subscriptions worldwide will come down only very moderately, because the income from APCs itself would be (very) moderate.
The second possibility is more challenging and adventurous for both parties. In this model, universities (or consortia) would no longer pay for access to the journals, but would get access for free. At the same time they would agree to paying the full APCs for all publications of employees of that university (or consortium) in the journals of the publisher. This would mean an immediate transition from the toll access model to the open access model on the level of a single institution or consortium and a single publisher. The rest of the world would still be paying subscription fees, and again these fees would not drop, because the publisher is not making extra money from OA: OA income here simply replaces subscription income. In this case, which from a perspective of change is much more challenging than the former one, neither the university nor the publisher would be sure that costs (university) and income (publisher) would more or less reflect the existing situation. However, universities/consortia have a track record with publishers, which means that we know about how many articles employees of a university/consortium usually publish in the journals of the publisher. So it would be possible to make informed estimations on what would happen if such a deal were to take shape. To reduce uncertainties for both parties, both models could be applied in combination – for instance, 50% model 1 (which would reduce the price of the licence with 50%) in combination with 50% model 2 (which would reduce the APC with 50%).
Both models can be applied on the level of a single institution, the level of a consortium, or even on a national level. The UK is well placed to perform large-scale experiments, because the whole system of HE in the UK essentially is controlled by the government in a comprehensive way.
“… the costs of dissemination of research will inevitably rise for all research-intensive institutions in the (western) world.”
Hybrid OA – a way to go?
As I have shown above, hybrid OA is potentially a very disruptive way of implementing open access as the predominant business model for scholarly publishing. In a way, it combines the negative financial impacts of both current models and therefore it is not a suitable mechanism for change. Therefore it is absolutely essential to create a connection between the two systems, subscriptions and APCs, to form a sort of communication channel at the individual university and/or consortium level. Only under that condition, may hybrid OA be a way to go. But even then, the costs of dissemination of research will inevitably rise for all research-intensive institutions in the (western) world.