The growth of open access (OA) via the payment of article processing charges (APCs) in hybrid journals has been a key feature of the approach to OA in the UK ever since the release of the Finch Report and its subsequent acceptance by the UK government, and the Research Councils UK (RCUK) OA policy with its stated preference for gold OA.
In order to help UK institutions manage the increase in costs resulting from the combined cost of journal subscriptions and APCs, Jisc Collections (in common with groups in some other countries) has been piloting ‘offsetting agreements’ that explicitly link subscription and APC payments, seeking to reduce one as the other grows.
In the UK, Jisc Collections has made some progress implementing such agreements with a number of publishers of various sizes. However, this progress has been against a background where the concept of offsetting has itself become increasingly contentious.
On the positive side, there is some evidence that offsetting agreements have curtailed and even removed the increases in costs to institutions and that, at their most effective, they can also support institutions in the administration and implementation of OA. However, such agreements are flawed through their implicit acceptance and strengthening of the current costly and opaque market for journal subscriptions. Meanwhile, many publishers decry the failure of research-intensive countries to adequately fund the implications of their policy decisions, pushing the responsibility on to publishers.
Nearly three years after Jisc Collections first started actively pursuing offsetting agreements, this paper will provide an update on the status of negotiations, reflect on the challenges and opportunities presented by such agreements, and consider the implications for the path of future negotiations.
The views expressed are those of the author and do not reflect the views of Jisc.
The Finch Report, its acceptance by the Government and the introduction of RCUK’s block grant in April 2013 to help with the implementation of its policy were all very influential in the growth of gold OA via the payment of APCs in the UK from 2013 onwards, especially in hybrid journals.
Whilst we in Jisc Collections had undertaken negotiations for journals and OA for a number of years, we had never sought to negotiate on the combined cost of subscriptions and APCs in ‘big deals’. For most of 2013 our activities were focused on mapping the existing landscape of expenditure on APCs and developing the concept of the ‘total cost of ownership’, also referred to as ‘total cost of publication’. This demonstrated the likely impact on (in particular) research-intensive universities of paying for APCs and subscriptions (which analysis suggested would be high since, in the absence of a global transition to OA and irrespective of anti-double-dipping policies, UK institutions would be paying for both for some time to come), and we commenced discussions with publishers aimed at raising awareness of the issue and putting in place mechanisms that would mitigate the impact.
Events in early 2014 quickly saw a much more concerted approach adopted to tackle this issue. The initial development was an open letter from David Willetts (then Minister for Universities and Science) to Dame Janet Finch following the first review of progress in implementing the recommendations of the Finch Report.
In it, Willetts made two statements that helped shape Jisc Collections’ activities. The first was to explicitly call upon the publishing industry to develop ‘innovative and sustainable solutions’ that would allow ‘a meaningful proportion of an institution’s total [APCs] with a publisher to be offset against total subscription payments with that publisher’. The second was to acknowledge the role of Jisc Collections: ‘Government welcomes efforts by Jisc Collections to develop sustainable funding models that establish a relationship between the payment of APCs (and the costs of administering them) and subscription fees for an institution’.
The letter demonstrated the support of policymakers for offsetting agreements, it made a clear link between the payment of subscriptions and APCs and the need for there to be a balance between them, it noted that administrative costs were an important consideration and, perhaps most importantly from a Jisc Collections perspective, it provided a strong mandate to pursue negotiations on behalf of the sector.
The next development was that soon after the Willetts statement Jisc Collections received the first analysis of data from 24 institutions on APC expenditure from 2011 to early 2014. This data set, incomplete though it was, clearly demonstrated the pronounced upwards trend in expenditure after the introduction of the block grant in 2013. From 2012 to 2013, expenditure more than doubled to nearly £4million and it was projected that it would more than double again in 2014 to nearly £10million.
It was against this background that in spring 2014 Jisc Collections approached all journal publishers with whom it had agreements to pilot some form of offsetting arrangement that would reduce or constrain either subscription or APC expenditure, or both.
Status of negotiations
At the start of 2017 Jisc Collections had offsetting agreements in place with the following publishers:
- De Gruyter: hybrid APCs for articles published in one year are offset against institutions’ expenditure on subscription and licence fees in the following year
- IOP (Institute of Physics) Publishing: hybrid APCs for articles published in one year are offset against institutions’ expenditure on subscription and licence fees in the following year
- SAGE Publishing (including the Royal Society of Medicine and the Institution of Mechanical Engineers): discount on APCs in hybrid titles, requires a code
- Springer: ‘flipped’ model, where UK subscription spend has created a ‘publishing pot’ to fund APCs. An additional transition fee covers access to subscribed content
- Taylor & Francis: institutions receive vouchers which offer a significant discount on APCs. The number of vouchers is based on expenditure
- Georg Thieme Verlag: complimentary APCs are included with subscription agreement
- Wiley: institutions are eligible for a tiered credit based on overall level of expenditure. The credit is added to an institutional account, which is drawn down against APCs.
In addition, discussions were ongoing with the Royal Society of Chemistry.
In order to evaluate their effectiveness, Jisc Collections is undertaking an ongoing evaluation of all of its offsetting agreements, with the first report (covering a subset of the agreements for which 2015 data was available) released in summer 2016.
Given that the RCUK block grant will continue to be available until 2018 and that the policy stance from government is still in favour of gold OA and still seeking progress with offsetting agreements, Jisc Collections will continue to pursue offsetting agreements with all journal publishers with whom it enters big deal negotiations.
However, whilst in the first round of negotiations Jisc Collections was not too prescriptive about the type of agreement put in place, the evidence gathered over the course of the last three years (2014–2016) on the effectiveness of different offers will be used to improve on existing agreements. Furthermore, wider discussions around offsetting agreements and the evolving attitude of institutions towards them will influence the approach taken.
In the next section, the paper will identify some of those discussions, both positive and negative, that are influencing the development of offsetting agreements.
A bigger big deal?
Why then are offsetting agreements regarded with such suspicion? At its simplest, the reason is concern around the potential costs and implications arising from their relationship to OA in hybrid journals and the big deal.
Whilst OA in hybrid journals may be popular with UK authors (around 70% of paid APCs in the UK are in hybrid journals), they are much more controversial with libraries, research managers and OA advocates. The reasons for this are well documented – briefly, they emphasize the association of gold OA with the payment of APCs, those APCs tend to cost more than APCs in pure gold journals, there is a higher cost of administration compared with pure gold and green OA, there are concerns over what one is actually purchasing, how one ensures that there has been no double-dipping – and, more generally, there is a sense that hybrid is not very progressive and poses the same market issues associated with the subscription journals market.
An offsetting agreement magnifies these tensions by coupling hybrid with that other bête noire of scholarly communications, the big deal, or rather what now resembles a bigger deal.
If an OA model is meant to replace the subscription model, why does expenditure on both APCs and subscriptions continue to rise so inexorably? It does not matter that many of the offsetting agreements are pilots and have not yet been fully reviewed. The fact that the expenditure lines are not flat or falling in either case makes offsetting agreements resemble nothing more than an ‘advantageous lock-in for status quo publishers’.
The preponderance of these status quo publishers is another of the issues with hybrid and offsetting. A review of the publishers receiving the bulk of the expenditure on APCs suggests that far from posing a threat to those status quo publishers, it is a very profitable additional revenue stream and the same publishers who dominate the subscription journals market dominate the OA market as well. As noted above, this has been used to argue that the current incarnation of the market for APCs is displaying the same market dysfunctions as the traditional market for scholarly journals.
Transparency and cost allocation
By building on, or taking existing big deals as the starting point for negotiations and cost, offsetting agreements must also contend with two further issues concerning traditional big deals: the lack of transparency on the commercial terms and the challenge of cost allocation across individual institutions.
The issue of transparency reveals itself in a number of ways. The absence of information about the terms of the agreements suggests that there is something clandestine occurring, thus undermining trust in the motives and outcomes of such negotiations. As David Crotty wrote in early 2016, ‘Whereas every deal brings with it a triumphant set of press releases, public details on the financial mechanics of these deals are non-existent. Most seem to be happening under the same sorts of non-disclosure terms that commercial publishers require of their big deal subscription customers. This, on its surface, seems to fly in the face of the increased transparency and openness that lie at the heart of the OA movement’.
This lack of transparency poses further complications for those attempting to implement and monitor such agreements. The 2015 review of the implementation of the RCUK OA policy made numerous comments (for example, page 11) about the need for transparency and the challenges of gathering data to give firm evidence as opposed to opinion-based footing for the review. It is notoriously difficult to find comprehensive information on subscription expenditure, even if information on APC spend is increasingly widely available, for example, via the OpenAPC initiative, which means it is difficult to tell whether there have been any cost savings.
Jisc organized a series of workshops in order to help institutions get the most from the agreements, one of the reasons being that the agreements negotiated ‘were significantly different from one another that there was confusion across the sector, leading some institutions to ignore the deals altogether and others to become increasingly frustrated with the management of the schemes’. It would appear then that the very complexity of the agreements being negotiated could make it difficult for institutions to take advantage of the offers. To an extent this reflects the novelty of agreements and the fact that in many cases publishers and institutions are developing systems as they go, sometimes as a response to, rather than in advance of, the negotiated agreements. In the case of the Jisc workshops, publishers themselves attended with the aim of improving not only their understanding of the agreements, but also their own understanding of institutional requirements.
A further challenge arising from the basing of offsetting agreements on existing big deals is the issue of cost allocation between institutions. It is a source of frustration to anyone involved in the licensing of e-journals that pricing is based on historical print spend, where the amount spent on access to e-journal collections is based on levels of expenditure on print journals in the mid-1990s. Over time this has led to wide variations in the amounts that ostensibly similar-sized institutions pay for access to the same e-journal collections. As knowledge and understanding of this issue has grown, it has become increasingly difficult to justify, with an increasing perception of ‘winners’ and ‘losers’. There is increasing desire within the UK to introduce cost allocation mechanisms based on more transparent measures suitable for a sector that has moved from print to electronic and changed considerably over the last 20 years (for example, a 2014 review of Jisc Collections’ NESLi2 e-journal licensing initiative found considerable desire amongst library directors for Jisc Collections to accelerate moves away from historical print spend).
This variation in spend is such that making the move is difficult in practice, especially with the most expensive agreements, and as a result, despite its deep unpopularity, historical print spend persists as the basis for most big deals.
However, if one moves into an APC market based on publication output rather than subscriptions and the consumption of content, the use of historical print spend becomes even more difficult to justify. Whilst funders have made some (considerable) funds available to assist with the transition to OA, those funds do not cover all of the cost (up to £80 million according to some estimates) and are not guaranteed on an ongoing basis. This raises the affordability issue for the most research-intensive institutions, many of whom have adopted ‘green first’ policies internally. Most acutely, though, it increases the need for the members of consortia to consider how they will approach the issue and balance the legitimate concerns of their members.
The nature of this challenge is further complicated when one considers that, were there a global transformation to OA of the type proposed in a white paper by the Max Planck Digital Library (MPDL), then cost reallocation would not just have to happen between institutions in one country, but across many countries. Now, this is not impossible – under the subscription model, there is a de facto allocation of costs across countries based on a variety of causes such as historical expenditure, changing patterns of research output and publisher sales targets and, furthermore, the MPDL paper states that there should be budget surplus arising from such a transition and it could be that such a surplus might be used to smooth the impact of the transition between institutions in different countries – but it is undoubtedly challenging.
Such a flip also assumes that all countries agree on this as the best path to OA, when in fact there are very different attitudes and approaches to OA in different countries. Offsetting agreements may make sense in European countries with a more centralized and national-level approach to negotiations backed by policies supportive of gold OA. Such an approach may not make sense elsewhere. Recognizing this challenge, the University of California Libraries’ project, ‘Pay it Forward’, sought to understand ‘whether a large-scale conversion to open access scholarly journal publishing funded via APCs would be viable and financially sustainable for this class of large North American research-intensive institutions’. It concluded that a fully APC-based model would cost such institutions more than their current library budgets, even if it recognized that additional sources of funding could cover the gap. It also highlighted the crucial role of authors acting as ‘informed consumers’ to introduce competition and market pressures. Reading the report, one notes the extent to which North American institutions are operating in a context that is not conducive to offsetting agreements as they have been implemented in Europe.
If the challenges above give a sense of the objections from the perspective of universities and OA advocates, it is also important to note that many publishers themselves have a number of issues with such agreements, as outlined below.
Concentration on established larger publishers
As noted earlier, APC expenditure appears to be concentrated on larger, established and often commercial publishers, with some 70% of APCs being spent in hybrid journals. (The University of Cambridge has reported that the figure is even higher there, at 85%.) This has not gone unnoticed or unchallenged by smaller, society and pure OA publishers, who are unsurprisingly concerned that if budgets are constrained and an increasingly large proportion of those budgets is being spent with the biggest publishers, what does that mean for their own sustainability? Might a move to OA end up with the unintended consequence of increasing concentration of publishing with a smaller number of publishers?
Is ‘offsetting’ appropriate or feasible for all publishers?
A further criticism is that, whilst a generous offsetting agreement may be feasible for publishers with a large amount of subscription revenue and wide subscriber base, what of those publishers that may have a comparatively low level of subscription revenue, but have been successful in attracting a large number of authors paying APCs? Given the constraints of the journals market and the priority given to constraining any price increases in consortium negotiations, if these publishers deeply discount APCs they are giving up one of the few mechanisms available to them to boost revenue (and even introduce some movement into the marketplace?).
Some publishers who are approached by consortia for offsetting agreements express the view that they have to offer three levels of discount. First is the ‘discount’ they already offer on the collective price of journals as part of the big deal, second is the global price reduction on subscription costs they implement in order to fulfil their anti-double-dipping policies and, finally, they are now asked to implement a third local offset for those research-producing institutions faced with increased APC costs. Withdrawing any one of these will lead to protests from at least one segment of their customer base depending on that particular segment’s attitude or policies around gold OA, thus placing them between a rock and a hard place.
Inadequate funding of OA policies
This brings one to the accusation levelled at some research funders that, having made OA a requirement of their funding, they have not made adequate funds available to institutions to comply with those policies. In evidence to the review of RCUK’s policy on OA in late 2014, Steven Hall of IOP Publishing commented that there is a ‘real challenge in properly making the RCUK gold policy work, in that the default position of most universities is towards green – for very obvious reasons in terms of levels of funding available currently to support gold’. According to this rationale (and it should be noted that IOP Publishing does offer a very generous offsetting scheme), it is not the responsibility of publishers to fund the implementation of funder policies and the burden has shifted too far on to publishers. Though of course, others will note the persistent financial health of many publishers despite (or maybe because of) OA policies, the comparatively high cost of APCs in hybrid journals and the additional costs being borne by institutions as counter-arguments demonstrating that the burden is being shared and that were it not for the high cost of hybrid APCs, money from funders might go further and institutions might be less concerned about the implications of following a gold over green policy themselves.
Continuing growth of the subscription model
Finally, some publishers note that, whilst the volume of OA material published globally is growing year on year, subscription content is also growing. A report for the Universities UK Open Access Co-ordination Group found that ‘the number of articles published globally in journals with an immediate OA publishing model (fully-OA journals, and OA articles in hybrid journals) grew faster between 2012 and 2014 than articles in subscription-based journals, although both increased in absolute terms’. So, whilst OA material accounts for a growing proportion of the overall amount of published material, it does so in a context where the volume of material published globally is increasing annually. OA may not therefore be replacing subscription content and if there is a transition, it is to a market based around at least two business models rather than just one; both of these need to be paid for.
One might note in response that the big deal model long since broke the link between purchasing an absolute number of journals and the price paid, in order to maintain some semblance of affordability for institutions when faced with an ever-increasing volume of research output. One might expect similar pressures to come to bear on any models requiring payment for both subscriptions and APCs.
There are doubtless other arguments and counter-arguments around offsetting agreements, for now it is perhaps important to note that a publisher’s view of them will owe much to their size, portfolio and position in the existing market place.
Given the number of challenges facing offsetting agreements, and the list above is not comprehensive, why is it that consortia continue to enter into such agreements?
It is important to note that those involved are very much aware of the various challenges and are reviewing the effectiveness of the agreements on an ongoing basis. The fact that many are billed as ‘pilots’ not only reflects publisher wariness, but that of consortia as well. In the UK, Jisc Collections worked with institutions to draft Principles for Offset Agreements, against which the sector might review them. By the same token, the Efficiency and Standards for Article Charges (ESAC) initiative issued a Joint Understanding of Offsetting from a number of European consortia, laying out some of their concerns and requirements for the development of such agreements, including the renaming of offsetting agreements to ‘transitional’ to emphasize that such agreements should become unnecessary as soon as possible and should not be considered a permanent feature of the negotiation landscape.
Offsetting agreements are often only one of the initiatives that consortia and institutions undertake to support the transition to OA. For example, Jisc Collections, working with colleagues across Jisc, has agreements for pure gold publishers, OA monographs and infrastructure services, not to mention discussions on green OA requirements that it holds with publishers. There is thus recognition that hybrid journals, offsetting agreements and APCs more generally are only one path to OA and will not be appropriate in every discipline, country or format and certainly not in the longer term.
Offsetting agreements might then best be viewed as an ‘attempt to tame the hybrid phenomenon and turn it into a productive transitional strategy that leverages the efficiency of existing institutional relationships with publishers’. As such, they represent a pragmatic response that acknowledges the current primacy and persistence of the big deal, but does not accept it as the end-point or any type of ideal. Indeed, some of the criticisms of offsetting agreements seem designed to distract from this point and suggest that there is broad support for big deals, with only the offsetting agreements having their legitimacy and value questioned.
However, by using the Principles mentioned earlier as a framework, one can see some evidence that opportunities for a productive transition do exist, even if realizing those opportunities will be far from simple.
Does the agreement contribute to the transition to OA?
There is evidence that the right type of agreement can radically increase the volume of OA with any particular publisher. The UK Springer Compact agreement has seen in excess of 3,000 articles made open access in its first full year. Figure 3 compares articles made OA in 2015 and 2016, clearly demonstrating the ‘flip’ that has occurred. It will be interesting to see whether the increase in publication by UK authors with Springer in November 2016 compared to November 2015 becomes a trend indicating a preference for the model.
Whilst not the same as ‘affordability’, the review undertaken of Jisc Collections’ offsetting agreements found that in 2015 the five agreements reviewed saved £1.2million over the amount institutions would have had to spend were no agreement in place. However, one should note that it is not clear whether these articles would have been made OA if there had not been an agreement in place. So, whilst the agreements may have encouraged more articles to be made openly available, it is not clear if money is being saved, or costs avoided.
Since October 2015 the UK Springer Compact agreement has provided access to content and seen more than 3,000 articles with an APC cost of over €6million published by UK authors for the same amount of money spent previously on gaining access to the same content and publishing around 500 articles openly.
The IOP Publishing offsetting agreement, which is now starting its second round, allows institutions to recover 100% of their APC expenditure.
However, a concern has been expressed that in order to achieve these ‘savings’, institutions are signing up to agreements that demand a greater initial financial outlay. The timing of many of the agreements in the UK followed the initial surge in expenditure in 2014. Many institutions have entered agreements as a way of constraining costs that had already grown substantially. Thus offsetting agreements have at least improved affordability, even while questions remain about the overall level of expenditure.
Although the figures and hypothesis are contested, the MPDL paper discussed previously calculated that there was sufficient funding within the current system to transition to OA without additional expense. Whilst there are clear challenges over cost allocation in such a transition, offsetting agreements can provide a practical stepping stone between different states. Helping to ensure affordability, whilst supporting a feasible transition.
Further evidence will be needed to see whether offsetting agreements have truly limited the growth in expenditure overall.
Ease of administration
The differences between various offsetting schemes in the UK had been a source of frustration to institutions, as reported earlier. Initial feedback suggested that the systems to support the different schemes were at best untested and at worse non-existent.
Administration has also been an issue, as shown by the fact that Research Libraries UK (RLUK) saw the need to create an Open Access Publisher Practices Group and, until November 2016, Jisc ran the Open Access Good Practice project to help institutions implement OA and get the most out of agreements. Discount and voucher schemes have been particularly unpopular. However, it should be noted that institutions and publishers have made efforts to improve processes and develop systems that support administration. Dashboards providing up-to-date information of the type offered by Wiley, and the simplicity of the Springer and IOP Publishing schemes, have received positive feedback.
One can also point to the inclusion in the agreements of provisions supporting compliance with funder requirements. For example, under the Springer Compact agreement, Jisc was able to show that 100% of articles had been given a CC BY licence, in line with RCUK and Charity Open Access Fund (COAF) policy. This saves institutions a considerable amount of time and effort and shows a significant improvement on the 66% compliance reported by the Wellcome Trust in their review of their OA spend for 2013–2014.
Transparency of agreements
On the question of transparency, the picture is again mixed. The fact that the agreements are based on big deals does impact on the transparency of how much individual institutions are paying for the mix of subscriptions and APCs. Jisc’s own analysis of subscription and APC data found that, ‘It is difficult to assess the true cost of APCs paid for through offset and voucher schemes and these are often recorded at £0’.
Whilst the Springer Compact agreement is welcomed in terms of supporting a transition, affordability and ease of administration, it could be said to be lacking in transparency on the breakdown of institutional spend between subscriptions and APCs. However, the level of reporting provided as a result of the agreement and the fact that this information is being made openly available, means that there will be more information than ever available on exactly what the outputs of the agreement were and what was paid for them.
Offsetting agreements can greatly increase the volume and speed at which information is being made available. For example, the UK, Netherlands, MPDL and Austrian Science Fund are all making their data from the Springer Compact agreement available via the OpenAPC initiative.
The level of transparency is still not as great as it should be but the more data is made available, the more useful it becomes, the more it becomes an essential tool in justifying the continuation of the agreements and the more unacceptable it becomes not to make the data available on subscription expenditure.
If nothing else, experience so far suggests that a market based around APCs would be more transparent than the traditional subscription market and offsetting agreements would appear to be contributing to the release of information.
Implications for negotiations
These are examples from one round of agreements in one country and are therefore far from conclusive. All of the agreements have issues, albeit some more than others, and will need improvement.
From the perspective of future negotiations, the experience to date, the need to achieve and demonstrate value for money and feedback from institutions, funders and others would suggest the need for an emphasis on agreements that:
- are genuinely transformational and contribute to an affordable transition to OA and, where this is not the case, there will be pursuit of agreements that incorporate a strong provision for green OA
- recognize the diversity of the sector and the differing levels of interest and need for agreements that incorporate gold OA – so, there will be a need for more options that meet differing institutional priorities
- are capable of monitoring and demonstrating compliance and which require greater transparency on expenditure and the outcome of that expenditure for authors, institutions and funders
- are increasingly concerned with active service-level agreements with detailed reporting and a focus on administrative efficiency
- are focused on examining and testing cost reallocation mechanisms.
In addition, consortia are broadening their portfolios so that they cover related services, such as ORCID, that support the efficiency of offsetting agreements and transition to OA more generally.
The case for offsetting agreements is clearly far from proven. Such agreements have significant problems associated with them and they can undoubtedly provoke strong reactions. Even those consortia and institutions that negotiate offsetting agreements are wary of them and of entering into new ones on account of fears that they will not save any money, are inefficient and costly to manage or implement, and as time goes on seem to offer meagre opportunity for underpinning a widespread transition to OA. However, unlike many models, proponents of offsetting agreements recognize these challenges and flaws from the outset, consciously setting as a measure of success the eventual need to no longer pursue them because they will have achieved their goal and, if not, being willing to exit the arrangements.
Furthermore, it has been shown that, done well, they increase the amount of material made OA in a way which is practical, affordable, transparent and easy to administer. Thus they provide an opportunity for agreements that are qualitatively superior to agreements solely based around subscriptions to journals.
Increasingly, I have come to believe that the major failing of offsetting agreements lies in their assumption and continuation of the norms that govern the negotiation and implementation of existing big deals. Having been conceived as a genuine attempt to undermine and move beyond such arrangements, it is unfortunate that they have far too easily come to be regarded as ‘business as usual’ and even contradictory to the objective of open access. In reflecting on the various criticisms, the most cutting is around the lack of transparency in many of these arrangements. Those involved in such negotiations are undoubtedly acting in good faith and seeking the best possible arrangements for the institutions on whose behalf they act. All of this hard work and effort is undermined by the absence of transparency, which simultaneously prohibits a proper exposition of the benefits and a full response to criticism beyond ‘Take my word for it’. Rectifying this will be essential if they are to play any role in a sustainable transition. Funders could play a major role here by insisting on greater transparency if their funds are to be used as part of such arrangements.
If offsetting agreements have achieved anything, then, it is to raise the profile of and stimulate widespread interest in matters that were all too often overlooked, avoided or considered irrelevant, be it communication with authors, transparency, reporting, effective service-level agreements and what a properly functioning scholarly communications system could look like – even if it is only because no one wants it to look like an offsetting agreement.