Business models for e-journals : reconciling library and publisher requirements ?

The JISC commissioned Rightscom to carry out a series of interviews with librarians and publishers in order to understand the strengths and weaknesses of current business models for e-journals. Rightscom developed new business models and created dynamic working models for a selected number of them. Librarians and publishers agreed in such areas as the need for more funding to cover the increased output of research, the need for predictability and not restricting usage, but disagreed over the retention of print and the need for flexibility.


Background
It has been clear for some time that there are significant concerns in university libraries about current models for the purchase of electronic journals and the ability of libraries to meet the needs of their user communities in the provision of journals.Last summer the Journals Working Group of the JISC commissioned Rightscom to undertake an analytical study to help them to understand the underlying causes of those concerns in detail, to analyse the current business models themselves and to assess the potential for new business models.
Rightscom carried out a series of interviews with librarians and publishers in order to understand the strengths and weaknesses of the current models as viewed by the customers and the providers, and to gauge the reaction to possible alternatives.Some of the library interviewees were directors of library services and others had specific responsibilities for e-journals, providing a mix of strategic and budgetary perspectives as well as direct experience of negotiating deals, managing resources in accordance with licence conditions, and service issues.The publisher interviewees were selected from commercial, scholarly and open access organizations with a range of subject interests, and with representation from North America and Europe.

The view from the libraries
The fundamental concern for libraries is obviously to facilitate the widest access to the most appropriate resources for their user community, both for research and for teaching and learning, within budgetary and staffing/space constraints.E-journals (and other e-resources such as databases) are very popular with users, especially when they can be accessed in locations away from the library, unconstrained by its opening hours.As e-journals become more widely available, usage is increasing everywhere.In general, a move to e-only is supported in principle, but with significant provisos concerning cost (as VAT is imposed), and secure perpetual access to archives.The perceived importance of having their own archives varies considerably between libraries.
In terms of overall funding, it appears that many higher education institution (HEI) managements have been willing to make extra money available to facilitate greater online access to journals, but there is a widespread view that this phase is coming to an end.Recent favourable currency movements have also eased the pain of journal price rises.

Serials -18(2), July 2005
Look, Sparks and Henderson Business models for e-journals 157 Business models for e-journals: reconciling library and publisher requirements?
The JISC commissioned Rightscom to carry out a series of interviews with librarians and publishers in order to understand the strengths and weaknesses of current business models for e-journals.Rightscom developed new business models and created dynamic working models for a selected number of them.Librarians and publishers agreed in such areas as the need for more funding to cover the increased output of research, the need for predictability and not restricting usage, but disagreed over the retention of print and the need for flexibility.

HELEN HENDERSON Information Power Ltd
Bundling and the big deal -it all depends where you're coming from It is widely acknowledged that publisher bundling and big deals have greatly increased the number of journal titles available, but the way this works at the moment is seen to have some significant drawbacks.Views about the big deal vary considerably and while each institution is different, there are some discernible patterns, broadly according to type of HEI.Large Russell Group institutions with major science and medicine departments find that many of the bundled titles are not necessarily the right ones for their communities and there is a long tail of titles which is not used; the share of the deals in the total budget is squeezing out other purchases and some subject areas are losing out; and they can face heavy cancellation penalties, making it difficult to adjust the collections as research interests change.
The more social science-and business studiesfocused Russell Group institutions seem happier with the status quo.They seem to have fewer problems with overall budgets.However, they are still concerned about future price rises.
Post-92 universities used to have small print collections and therefore are more enthusiastic about the big deal as it has delivered a major increase in the resources available to their users at a very reasonable cost.But overall cost and continued affordability are issues for them as well.
Libraries in institutions that have to focus on a narrower set of research priorities in response to the concentration of research funding need to be able to adjust journal collections in a more focused way as well.This is an issue that will loom larger for more institutions in future.Their needs could be summed up as 'unrestricted access to a restricted set of materials; restricted access to the rest'.
Institutions with a high proportion of distance or part-time students share many concerns in common with HEIs generally, but licensing terms are probably more important.Though they often have highly-rated research, these libraries felt most strongly about student provision.
FE institutions make more use of Eduserv Chest deals for databases than of JISC e-journal deals, though some of the databases will include journal titles.FEs have needs which are often very specific and can frequently be more in the area of businessto-business than journal publishing.

Reaction to new models
Libraries dislike restricting access to users and reject models that involve budgetary unpredictability and might involve them policing usage or even cutting it off if budgets ran out.They tend to oppose anything that would require researchers or students to pay individually for usage.However, it is partly a question of the level of charge and how peripheral the resource is.Pay-per-view models were not dismissed out of hand and some libraries believed it would be beneficial if some of the unpredictability and risk could be minimized.More flexibility in journal deals is a key issue for almost everyone, as circumstances for HEIs are changing more rapidly than in the past.There was general enthusiasm for a model that would allow libraries to tailor the bundles of titles they receive more closely to the interests of their community and to make some substitutions during the life of a deal.
There appears to be managerial enthusiasm for institutional repositories, partly to manage Research Assessment Exercise (RAE) submissions and showcase the institution's research output.While libraries are supportive of open access in principle, there is a good deal of scepticism about whether academic researchers are willing or able (in view of their focus on the RAE) to change their publishing patterns.

The view from the publishers
For most of the publishers interviewed, continuing their business at current or better levels of profitability (in the case of society publishers, this implies the level of contribution generated by publishing) was essential.Few could envisage continuing to have the support of their investors, shareholders or societies if they failed to deliver profits or surpluses that were close to or better than they currently were.Several of the publishers interviewed believed strongly that the solution to current problems lies in more funding for library acquisitions.They pointed out that the increasing volume of research was generating more and more papers for publication, and that therefore growth in journals was inevitable.Several of the publishers interviewed expressed frustration at the current situation, and believed that they could be doing much more for researchers if institutional budgets were to be expanded.
Publishers also stressed the level of investment that has gone into e-journals and one explicitly identified this as the reason for a one-off price rise that would not need to be repeated.
Publishers feel that their costs are hard to reduce.Each paper needs refereeing, and although referees are not in general paid, the administrative costs are directly related to the number of papers received.In some cases, these are handled directly by the publisher; in other cases, they pay expenses to editors to support the cost of additional administrative staff at the editor's institution.
A key point of difference between publishers and libraries was in attribution of the demand to continue with printed journals.Many libraries viewed the publishers as wedded to print-based pricing models; many of the publishers interviewed claimed that they would like to abandon print in the near future but that libraries were resistant to such a change.VAT has to be charged on electronic publications and this has an impact on library decisions.
Publishers would like predictability as much as libraries.They realize that pay-per-view models are difficult for everyone.All publishers were concerned about models that would encourage libraries to constrain usage: this is another shared viewpoint between libraries and publishers.In the case of the publishers, as well as being concerned about the impact on research itself, they were also concerned that the constraints of such models would lead to journals not being used effectively and thus to a perception that they represented poor value.Most publishers wanted to see some sort of relationship between usage and price.
Many of the publishers interviewed claimed to have neutral views on the question of open access: if an author-pays model could be made to work, they would be happy with it if it generated revenues and profits equivalent to their current model.
Publishers are also generally accepting of consortial models, although none felt that they did particularly well financially out of them.Most of the publishers interviewed felt that existing deals such as NESLi2 suffered from the fact that they are opt-in models and take-up was not always as high as hoped.This made it difficult for them to offer more substantial discounts.Many of the publishers interviewed therefore suggested that a UK-wide deal covering all institutions was a good way forward.It would secure access for all researchers, and secure a revenue stream that would allow publishers to develop their services.A small number of publishers, though, had definite objections to a national licence model.They felt that it would be very hard to determine centrally how resources were allocated and which titles might be included.

Building new models
A 'blue-sky' approach was deliberately taken for development of the initial group of models, with no consideration for the practicalities of implementation.The models that were developed for the project were derived from this initial group and are described below.
For each of these models a series of spreadsheets was prepared which allowed the models to be tested.A table allows variables to be entered, such as usage, base price, per article prices, institution size and accelerators.In this way it was possible to see the effects of the different variables on the initial price and ongoing costs.These models are intended for internal JISC use in evaluating publisher offers and proposing alternatives.

National licence
The national licence is similar to the original Pilot Site Licence Initiative model and assumes a single national payment to publishers for limited access to all their content.It allows a basic need for universal access to be met, but allows institutions the discretion to add such additional services as they feel are worth paying for, such as archives and print.

PPV converting to subscription
In this model, the institution may have subscriptions to some titles from a publisher, but uses pay-perview (PPV) to access other titles on an ad-hoc basis.Usage is based on a per download cost with a threshold at which sufficient usage has been made to convert to a subscription.The publisher would be able to set this at a premium above the standard subscription.

Pay-per-view pre-purchase
For smaller institutions, or those with very varied interests (see Cranfield's ILL experiment 1 ) which as a result will never reach a PPV threshold for some journals, it might be possible to buy blocks of discounted PDF downloads.This model is already offered by some publishers.

Core + peripheral
The publisher offers a set of 'collections' which include all their titles in a specific discipline.They then provide access to non-subscribed material (the rest of their titles) on a discounted pay-perview basis.Another option is for a library-selected set of 'core titles' replacing the publisher selection.

Open access -author pays
This is an open access model based on payments by the author on publication.Optionally, the institution may also pay a subscription: in this case, payment by any author at that institution will be discounted.

Open access -hybrid model
The author can choose whether to pay for publication, and make the article immediate open access, or not pay and have the article available only to subscribers.
These models were presented at a series of workshops at the Annual Conference of the UKSG in April 2005, which produced valuable feedback, particularly about which models would be preferred for trialling, and which would be easiest to trial -not necessarily the same ones!

Feedback from the workshops
Some of the key points made by workshop delegates are summarized here, under the relevant models.

National licence
Some delegates felt that a national licence could be politically problematic, in terms of top-slicing funding to pay for a package that not all academics would view as valuable.A national licence deal could also reduce choice.Though this was not the most popular option, it was seen as the second easiest model to trial.

PPV converting to subscription
This was much preferred to the unbridled pre-paid PPV model, and was the second most popular model overall.This has to be seen in the context of other models, though -it is an adjunct, not a total solution.There are also issues about the amount of premium applied to the subscription price by the publisher.

Pay-per-view pre-purchase
Although COUNTER is allowing usage metrics to be collected from different sources and consolidated, delegates considered that there is still little information on what individual researchers are buying and these PPV costs are lost in the system.Some delegates believed that the Elsevier TULIP experiment some years ago showed that PPV might not work in a large general environment.Corporate and niche libraries, however, could take advantage of it as in many cases they are transferring costs to the end-user.
PPV was generally seen by delegates as risky, complicated and expensive.Both of the PPV models were seen as relatively difficult to trial.

Core + peripheral
This model, along with some other ones, was considered by some delegates to be potentially too complex for agents' systems to handle.The administration costs for libraries could also be high for this type of model.
The issue of what constitutes the core is a crucial one, and the general view expressed was that it would have to be selected by the library, not the publisher.
On the positive side, it was felt that the core + peripheral model would allow the library to manage their relationships with departments better.It was the preferred option, and also seen as the easiest to trial, perhaps because it is a modification of the current bundled deals.
Trials for the first four models do not change the fundamental way in which journal content is paid for (i.e. by the subscribing organization or individual) but reflect a different way of assessing the value and basis of the charges.However, the two open access models present a different problem.The models are radically different, and publishers may not be prepared to experiment with existing subscription-based titles.
There are risks involved in introducing new models: ■ New models, especially forms of pay-per-view, may cause changes in behaviour by users, libraries and publishers that cannot be entirely predicted.
■ Certain key titles will always be able to command premium prices: these may be hard to integrate into wider-ranging deals.
■ For publishers, some new models (not just open access) represent changes in cashflow that will in general affect their businesses unfavourably and may eventually lead to the model becoming unworkable.

What new models can do -and what they can't
It should be obvious that the current situation presents a fundamental challenge: libraries clearly feel a strong need to reduce or at least contain costs; publishers feel an equally strong need to maintain current revenues.There is no business model that can solve that issue.What modelling can do is present the opportunity to make more informed decisions that potentially offer libraries ways to get better value for money out of their journal acquisition budgets.At the same time, the models should help publishers in understanding where their resources might best be used to help sustain their operations through delivering services structured in a way that meets the needs of their customers.
We believe that all the models stand a chance of success.Implementation, however, will be critical.