Over the past month, an unprecedented amount of data has been released that throws light on the flows of money in scholarly communication, both subscription and open access. While some of this information is depressing – there is so much wrong with the way the current system works – the very act of releasing the information is surprisingly heartening. These cracks in the publishing edifice are perhaps the first signal of a genuine shift towards price transparency. Transparency will not only throw light on the complexity of the system but will also be the means to foster real change and enable competition and market forces to act.
Opening up the big deal
On the subscription side, Tim Gowers has detailed the results of a Freedom of Information request (FOI) to the 24 Russell Group UK Universities about what they pay Elsevier for their ‘big-deal’ subscription package. An FOI is necessary because Elsevier has required institutions to sign a non-disclosure agreement (NDA). Needless to say, it is in the interests of Elsevier not to let on to anyone how much they are charging. In his insightful post, Gowers first outlines the background to his request (including the Elsevier Boycott, issues around double dipping and more recently Elsevier charging readers for their OA articles) before presenting a table of the results of the 19 Universities that agreed to release the information. This shows, as we might have expected, high variance in what different institutions pay for similar packages: “University College London pays over twice as much as King’s College London, and almost six times as much as Exeter”.
The real revelations, however, come from the correspondence he had with different institutions (especially their reasons for not giving Gowers the information), and with Alicia Wise from Elsevier and Lorraine Estelle from Jisc Collections, a UK Higher Education consortium (who are against NDAs). It turns out the system is far more complex than often assumed. Charging is largely based on ancient history – how much a given institution paid for individual print subscriptions in the 1990s. (Note also an older analysis by UCL’s David Courant that Gowers links to, which shows how the usage of Elsevier journals is high for a few titles but tails off very rapidly, with more than 200 journals not being accessed at all by UCL in one year.) Price negotiations with Elsevier have focused on the percentage rise each year for electronic access to the historic package plus relatively small fees for their larger bundled collections. Ultimately “the system ensures that what universities pay now closely matches their historic spend” even though the universities today are largely gaining access to the same content.
Gowers’ correspondence also exposes the ‘tricksy’ relationships between Elsevier and different institutions that help maintain this status quo. Some negative responses from the Universities, for example, contained paragraphs matching almost word for word the same arguments for not complying with his request, which he suggests points to a template answer that Elsevier provided to the institutions.
Although Elsevier and the UK is singled out here (with reason), Gowers also includes a preview of what some US universities pay, with information provided by Paul Courant, Ted Bergstrom, Preston McAfee and Michael Williams. This will be part of an upcoming preprint and will include other major publishers as well, such as Wiley and Springer, but is likely to recount essentially the same story.
In a significant move on the Open Access side, the Wellcome Trust released article level data on what had been paid to publishers for Article Processing Charges (APCs) from 2012-2013 in figshare. In a fantastic demonstration of reuse, the data were then posted to a public gdoc, where they were cleaned up and enhanced (e.g. adding the licencing information) through crowdsourcing. This was coordinated by Michelle Brook from the Open Knowledge Foundation with help from a host of others.
The subsequent analysis, summarised by the Trust, revealed that Elsevier received the lion’s share of the Wellcome’s APC spend and that hybrid journals charged the most for their APCs (which supports the Wellcome’s recent report suggesting the hybrid APC market is disfunctional). But the data also raised many concerns. Papers for which thousands of pounds had been paid were in some cases still behind a paywall or did not have the correct license.
Reasons to be cheerful
All these releases point to a trend for funders and research institutions to demand more transparent pricing for both subscriptions and APCs. Gowers’ 10,000 word post goes into the nuances, complexities and publishing norms that have led the subscription business model to where it is today – it’s not just Elsevier. Gowers acknowledges that getting rid of such an entrenched monopolistic system will not be easy and that the differential pricing he has exposed is a symptom of the system, not the problem itself.
The release of APC information by funders relatively early in the evolution of this Open Access business model, however, is a signal that those footing the bill for APCs are trying to avoid the mistakes of the subscription system and ensure price transparency and competition. Following Wellcome Trust’s lead, Cambridge University, Queen’s University Belfast and the Austrian Science Fund also released data for 2013 (all links to figshare). In the past few days Wellcome released additional data from 2010-2012.
The increased transparency that these data releases bring are very welcome, but we will need more complete, and more consistent data to really understand what how the cash flows are changing at the system level. The data being released today require a great deal of further work to process and integrate – Cameron outlines some of the problems of collectively curating data in a separate post – and they only represent a small part of the whole, even the whole of the UK. Regular, scheduled and consistent data releases will be needed if we are to manage the transition to Open Access effectively. The reporting of the first year of payments from RCUK institutional funds will be the next big step in this direction.
Ultimately as this data availability increases we can start to actively plan a shift from subscriptions to APCs. Many of the concerns raised about the move to Open Access publishing result from a lack of confidence about our ability to shift money from subscriptions to publication services. Understanding how much institutions are spending is a first crucial step.
For publishers these data release focus the attention of our customers – funders, researchers and institutions – on the quality of service they are receiving. The data release from Cambridge University is particularly sobering. The final column, showing whether a ‘problem free OA publication’ was received for the thousands of pounds paid in APCs, shows an awful lot of red for traditional publishers (OUP appears to be an honourable exception). Those publishers offering only an OA option did much better as you might expect. BMC, PLOS or Copernicus would be unlikely to accidentally leave a paper behind paywall. As the Wellcome post concluded “We expect every publisher who levies on open access fee to provide a first class service to our researchers and their institutions.”