Can We Subscribe to This, Please? : Realizing a Core Journal Collection for the University of Sussex

Library was finding it difficult to respond to demands for increased journal coverage. The library received constant requests for new titles and complaints that the collections were not supporting research and teaching at Sussex. It was obvious that one of the major obstacles to achieving a core journal collection for the University was that the existing budgetary model and method of reviewing subscriptions was no longer fit for purpose. An additional factor was the significant decrease in library staffing, especially professional 'faculty' posts. There are currently 2.8 professional staff directly supporting Learning & Teaching and Research, but no one at that level in the resources (or acquisitions) department. An objective was set to devise a new model that provided a flexible way of maintaining and developing an appropriate collection of journal titles. At the same time it was essential to preserve a transparent method of ensuring equity across departments in selection and retention. This system had to be managed and maintained with a limited staffing resource. The resources budget at this time was based on a University-determined formula that assigned an amount for books, journals and (most) online resources based on a teaching/research ratio in conjunction with a figure for full-time equivalent (FTE). ■ A teaching weighting was calculated for each department using the total undergraduate and postgraduate FTE for the department expressed as a proportion of the total undergraduate and postgraduate FTE. ■ A research weighting was calculated on the basis of the total research postgraduate and faculty FTE for the department as a proportion of the total research postgraduate and FTE faculty. Following a teaching/research ratio of 37.5/62.5, the budgets were weighted respectively to produce departmental allocations. The weightings were then combined to form a final weighting for each department. The resources budget less top-slice, which paid for generic resources, most standing orders and official publications, was then multiplied by the weighting for each department to produce the budget allocations. In 2005, like many HE institutions, the University of Sussex Library was confronting the problems of resourcing journal provision to support its teaching and research. These included increasing costs (with wide discrepancies of price across the subjects), the potential increase in access of content (such as NESLi2 consortia, deals, etc.), greater interdisciplinary research and a requirement for more non-traditional academic titles.This was coupled with an expectation from students and staff that the Library should be responding to the …

By the beginning of 2005, the University of Sussex Library was finding it difficult to respond to demands for increased journal coverage.The library received constant requests for new titles and complaints that the collections were not supporting research and teaching at Sussex.It was obvious that one of the major obstacles to achieving a core journal collection for the University was that the existing budgetary model and method of reviewing subscriptions was no longer fit for purpose.
An additional factor was the significant decrease in library staffing, especially professional 'faculty' posts.There are currently 2.8 professional staff directly supporting Learning & Teaching and Research, but no one at that level in the resources (or acquisitions) department.
An objective was set to devise a new model that provided a flexible way of maintaining and developing an appropriate collection of journal titles.At the same time it was essential to preserve a transparent method of ensuring equity across departments in selection and retention.This system had to be managed and maintained with a limited staffing resource.
The resources budget at this time was based on a University-determined formula that assigned an amount for books, journals and (most) online resources based on a teaching/research ratio in conjunction with a figure for full-time equivalent (FTE).
■ A teaching weighting was calculated for each department using the total undergraduate and postgraduate FTE for the department expressed as a proportion of the total undergraduate and postgraduate FTE.
■ A research weighting was calculated on the basis of the total research postgraduate and faculty FTE for the department as a proportion of the total research postgraduate and FTE faculty.
Following a teaching/research ratio of 37.5/62.5, the budgets were weighted respectively to produce departmental allocations.The weightings were then combined to form a final weighting for each department.The resources budget less top-slice, which paid for generic resources, most standing orders and official publications, was then multiplied by the weighting for each department to produce the budget allocations.

Can we subscribe to this, please? : realizing a core journal collection for the University of Sussex
In 2005, like many HE institutions, the University of Sussex Library was confronting the problems of resourcing journal provision to support its teaching and research.These included increasing costs (with wide discrepancies of price across the subjects), the potential increase in access of content (such as NESLi2 consortia, deals, etc.), greater interdisciplinary research and a requirement for more non-traditional academic titles.This was coupled with an expectation from students and staff that the Library should be responding to the explosion in availability of online journal content.It was concluded that the existing budgetary model and method of reviewing subscriptions was becoming very cumbersome and not up to the task.
It became essential to formulate an innovatory method by which the Library could be confident that its journal collections supported the teaching and research work at the University.This article will detail the steps taken in the last two years to deal with this.

JANE HARVELL Research Liaison Manager
The Library, University of Sussex Individual journals were allocated to the relevant department based on the subject area of the title; costs were shared when a number of departments demonstrated interest in the particular title.Existing subscriptions were reviewed annually with departments using lists to confirm whether they wanted to continue with a subscription, along with their budget allocation.This had grown out of a previous system which involved faculty and doctoral students voting for journal subscriptions 1 .
This resource allocation model and method of reviewing and subscribing to journals was becoming unsustainable for a number of reasons:

Lack of continuity of provision
It had always been difficult to allocate titles to particular departments and grew to be more so as research and teaching became more interdisciplinary.There were several instances where one department wanted to cancel a title whilst the others, who shared the cost, wanted to continue subscribing.There were examples of titles which had been cancelled by a department one year and then, after a gap, had been taken up by another a few years later.The increase in the cross disciplinary nature of journal titles meant that it was no longer possible to confidently identify who was using the journals that were being purchased.

Impact of the 'big deals' and the explosion of online content
Teaching and research at the University of Sussex covers a vast range of disciplines and the Library is committed to supporting this wide range of subjects.There is an acute awareness that the NESLi2 deals would represent the best value for money in terms of journal coverage.However, it was almost impossible to allocate costs, such as the noncancellable titles in the existing model.When trying to apply a transparent 'broad brush' subject approach to allocation, disagreements arose amongst the departments on the relevance of the collection to their particular area (and who should carry the cost).

3.The 'non-cancellation' print clauses
A common element of the electronic big deals publishing model is the 'non-cancellation' print clause.This caused problems during the annual review of journal subscriptions as great tranches of the listings were identified as non-cancellable.This difficulty was often compounded by the timings of deals.The fact that many consortia and national negotiations are concluded so late in the year meant that intra-institutional consultation was rushed or non-existent.

Inconsistency of research funding
Having an element of the departmental budget allocation based on research levels was problematical as these elements were variable.
Predicting accurate journal costs for a department at the time of budget allocation was difficult, often resulting in many departments having small pockets of money remaining at the end of the financial year.These surpluses would usually be spent on non-journal resources.
In addition, the ever increasing difference in journal costs between subject areas meant that, due to the high costs of their journal commitments, some departments had no 'book budgets' to speak of.This meant money had to be vired around departments -all very time consuming in terms of liaison with academic staff.It is also difficult to track and indeed, in the end, to justify.

Library staff resources
An annual journals review was very staff intensive, placing a great deal of pressure on both academic liaison and resources staff for a concentrated three-month period.

6.Volatility of pricing levels
It was very difficult to get accurate costs for subscriptions and to set precise budgets for the annual reviews.Journal and e-information budgets were allocated to around 40 different departments.This gave the academics the impression that the amount was final -where often it was not.
In response to these six issues, it was decided during 2006/07 to top-slice the subscriptions budget (e-books and electronic resources included).This released funds caught up in departmental allocations, giving us greater flexibility to make the decisions on acquisition based on need.(Invoices are still nominally allocated at a departmental level for the purposes of TRAC/FEC accountability.)Journal requests from inter-library loans data and journals for which SFX provided turn-away statistics were evaluated.Evidence from this data enabled the library to take advantage of virtually all the NESLi2 journal deals and increase the quantity of online core content across all subject areas.
As a result there is now a substantial subscription budget to manage in the Library and a responsibility to provide a transparent and equitable methodology for creating a core journal collection which supports teaching and research requirements as far as possible within the constraints of our resources budget.
A number of administrative mechanisms have been developed to help achieve this.
1.The Library resources budget supports both subscriptions and (one-off) book purchases.In the face of increased requests for new titles and costs (and an increasing need to accommodate subscription models for e-book purchasing), the book budget is protected from being squeezed by subscriptions by setting a limit on subscriptions.The allocation (for all types of commitments) must not exceed 75% of the entire resources budget.This also protects the book budget from any potential cuts in the resources budget as a whole as many big deals require a commitment to maintain current subscriptions.Without such a limit, any cuts would need to be made from the book budgets.
2. Current subscriptions and suggestions for new journal subscriptions are now reviewed using a series of criteria.
Under the previous budget model, where departments had fixed allocations, any new subscription could only be made by cancelling an existing title to release sufficient funds.If the title was interdisciplinary, negotiation with other interested departments was required to allow the costs to be distributed.
The new model is based not only on need but also has a built-in methodology for judging whether the title is core to Sussex teaching and research priorities.It is a system of evaluation which helps both the selection of new titles and the monitoring of current subscriptions in a way which is manageable by fewer staff.This methodology is based on a system of scoring against a set of criteria that alters slightly depending on whether the title suggested is a new subscription or newly published.Each of the criteria is weighted and a score assigned.Taken individually, each metric has its flaws; some may be more subject biased, some may not be entirely accurate, but as a whole they provide us with a clear picture of the value of adding or retaining the title.
The criteria:

Titles used by Sussex academic faculty for submission in the 2008 Research Assessment Exercise (RAE)
It was agreed that if research had been submitted in a publication as part of the RAE process then that journal was deemed to be a core subscription.This works well with the selection of newly requested titles, however, it is recognized that this metric will soon become outdated.In the future it is likely that the institutional repository will be interrogated for this type of data.

Locally-cited references
The case for considering a local needs approach to journal selection and retention was presented clearly by Price 2 .Web of Science (WOS) is used to gather information on whether faculty at Sussex have published in a journal.This use of WOS as a bibliometric indicator follows a currently proposed use of a citation analysis tool for the Research Excellence Framework (REF) 3 .Monitoring of the development of the REF may result in the use of other tools to gather information in addition to the WOS, e.g.Scopus.

A high Impact Factor
Impact Factors are imperfect as a tool for measuring research quality.The principle problem is that Impact Factors do not reflect either the quality or the number of citations of an individual article 4 .However, the library has taken the view that there is no reason why an Impact Factor measurement cannot be run alongside complementary modes of assessing research quality 5 .
In order to score in this area the title must appear in the most recent top 100 list of either the Science or Social Sciences edition of the ISI's Journal Citation Reports.

Inter-library requests (ILRs)
This criterion is used for new requests only.ILRs are examined for the 12 months previous to the request for the new title.Multiple requests from the same person are also taken into account as repeated use for a perceived singular piece of research would not necessarily indicate that this would be a valuable addition to the collection.

Citation on reading lists
New subscription requests are checked to see if they feature on a referenced reading list and if an article has been digitized under the Copyright Licensing Agency (CLA) licence.Existing subscriptions are checked as to whether the title is linked to any reading lists.

High online usage
This is applicable to existing subscriptions only.Statistical information is gathered for those titles that have COUNTER-compliant usage data.In addition, and in order not to discriminate against those titles that do not comply with COUNTER, data is also collected from the SFX journal management system.
An average use is calculated for each set of data and those that fall above this average (in either of the sets of data), will score in this criteria.The pilot phase of this project has shown that useful data can be collected here as long as the comparative use within a single discipline is examined.It is acknowledged that it is unreliable to look at online journal usage across subject disciplines.

A significance factor
This gives the flexibility to reflect the variable use of journals across disciplines.For example, Maths titles are used only for research and not for teaching.When any are requested or reviewed, a significance factor is applied to counteract the low score in teaching.Similarly this will be applied if a title is available in print only or if it is a new publication.

8/9. Essential for research/Essential for teaching
These criteria require input from the requestors and the teaching and research faculty.When a new title is requested, the library asks which course(s) it will be used for and/or whether there is substantive evidence that a subscription will support the ongoing research activities of the University.
When reviewing subscriptions, these criteria are filled in by departments, providing evidence of need and, importantly, future needs.

Costs
The final criterion is the actual cost of the title.The model was originally based on some of the fundamental principles of zero-based budgeting.For this reason methods of acknowledging costs are looked at when considering new subscriptions and reviewing existing titles.It has given some interesting data on the average cost of subscriptions by subject area which have been fed back to the academics.
Although the cost of the subscription is not currently judged as a critical factor, the affordability of the title is nonetheless crucial in any decisionmaking.A high cost could sway a borderline decision, for example.Methods of integrating this metric more fully into the core selection and retention criteria for journals are being investigated.

Conclusion
The library has just completed a year-long pilot phase of this new methodology for evaluating new and existing subscriptions to ensure they are core to teaching and research at Sussex.New requests have been scored and evaluated for the subscriptions of seven departments during the pilot.Titles are still allocated to departments in order to help review the collections.Many of those titles reviewed are still non-cancellable but it is felt that it is still important to identify those that could be cancelled, if the option was available.It means that the library can respond quickly to cancellation clauses in future publisher deals.
Since 2006 Blackwell Publishing, Taylor & Francis, Springer and Wiley collections have been added to existing subscriptions along with a number of other major collections.In June and July 2008, a survey was carried out on current and future use of the Library by teaching and research staff at Sussex.The results have provided convincing evidence that academic colleagues appreciate the recent expansion of journal provision and are keen for this to continue.In a question about priorities, journal provision came out top of the list.Their satisfaction in particular with the online delivery of journals and other resources was a strong endorsement for the University's policy to move towards e-only journal provision.