Introduction and overview

In 1992, the author celebrated with the British Library's (BL's) French agent the Library's reaching the £1 million target for ILL sales in France – maybe only one banker's bonus today, but vital for the BL's financial health at that time. The volume of ILL requests continued to rise at the BL, often at 20–30% a year until levelling out in 1999 at about four million requests a year; it then declined to about 1.5 million by 2011. The BL is still the largest document supplier in the world, mainly because this decline is a pattern reflected worldwide – except for the US, which will be discussed later. What caused this dramatic decline and is ILL still relevant today?

The decline of ILL

The global decline of ILL over the last ten years is well known, at least in respect to article copies, so only a brief note of the reasons is necessary:

  • the growth of ‘big deals’ has given access (if not necessarily ownership) to a greater number of serials
  • retrospective digitization of serial titles led by Elsevier in 2000
  • user expectation of immediate availability, thus many users are not prepared to wait for an ILL
  • an assumption that ‘if it isn't on the web, it doesn't exist’.

“The BL is still the largest document supplier in the world, mainly because this decline is a pattern reflected worldwide – except for the US…”

There are also countervailing forces:

  • the high price of serial titles and big deals leads to cancellations and hence unsatisfied demand
  • improved ILL services
  • more users and more research leads to more demand for ILL
  • more material being found as a result of improved searching techniques – and more being found that is not immediately available or not (yet) digitized.

The relevance of ILL today

ILL has never been more than about 4% of the average academic library's budget and is now much less, hence it is often dismissed as a peripheral service. This ignores the fact that much vital material is not accessible from the library directly – however large it may be. Indeed this problem was seen to be of great importance in a recent report: ‘The report's key finding is that access is still a major concern for researchers. Although researchers report having no problem finding content, gaining access is another matter due to the complexity of licensing arrangements, and restrictions placed on researchers accessing content outside of their own institution and the laws protecting public and private sector information’. It has ever been thus, except that today, access to records has increased vastly whilst access to material is more constrained than in the pre-electronic period.

“There exist 300,000 serial titles, current and non-current, of which only about 10,000 (3%) have been digitized back to vol.1, no.1.”

The scale of publication and digitization

Even librarians can underestimate the amount of ‘stuff’ that has been published. For example, users have immediate access electronically to a vast amount of literature – at least 90% of current issues of current titles – at a price. This obscures how much remains undigitized and inaccessible, unless the item is in the user's library. There exist 300,000 serial titles, current and non-current, of which only about 10,000 (3%) have been digitized back to vol.1, no.1. Google estimates that about 130 million books have ever been published, of which they have digitized about 10% (and are aiming for 25%); others add another 2–3%. Thus, even on an optimistic estimate, about 70% of books will not be digitized in the foreseeable future, and, in terms of accessibility, most in-copyright books will be behind pay-walls. There are millions of theses, conference proceedings, newspapers, etc., of which only a tiny proportion has been digitized, with a smaller number available freely to users. The expensive electronic access to all the current literature – affordable only by the richest libraries – combined with the vast amount of non-digitized literature would argue that ILL will remain essential indefinitely.

Big deals and ILL

Clearly, big deals have had a huge impact on ILL. Back in 1999, the British Library predicted this impact when looking at the future of the Document Supply Centre (BLDSC). However, the impact has been even greater than predicted and the deals have been ‘stickier’. That is to say, librarians have felt unable to revert to title-by-title selection when usage patterns become clear, partly because usage (or downloads) are very high and partly because the pricing policy of publishers makes it uneconomic to withdraw from deals. In the UK, in October 2010, Research Libraries UK (RLUK) mounted a campaign to wrest price reductions from Elsevier and Wiley when their long-term contracts were due for renewal. An alternative plan – known rather unoriginally as Plan B – was developed and a model was created that showed the consequences of purchasing serials title by title, the expected impact on ILL, and the costs involved. As a result, some fascinating data on downloads was uncovered. Fairly unsurprisingly, it showed that very large numbers of titles were not used at all and a further number were used so little that ILL would be a cost-effective substitute to subscription. But more surprising was the experience of other libraries that had already partially or wholly cancelled big deals. The impact of losing access to many thousands of downloads had very little effect on ILL demand; so much so that the ratio (admittedly from a relatively small number of universities in the UK and the US) was around 0.3%, i.e. an average of three ILL requests for every 1,000 downloads lost! This is an extraordinary result, even bearing in mind that many users must pay for ILL. If confirmed by larger studies, it suggests that the very high download figures overestimate the real importance of access to all these titles. Perhaps we should not be too surprised – after all, a user can only absorb so much information, the brain is not significantly better at processing information than when we lived in caves and there are still only 24 hours in a day.

“The expensive electronic access to all the current literature – affordable only by the richest libraries – combined with the vast amount of non-digitized literature would argue that ILL will remain essential indefinitely.”

At present we are in a period of uneasy truce. Big deals live on, but there is increasing hostility from librarians at the high prices (often more than £1 million annually) and many commercial publishers appear anxious to retain a business model that has served them and their shareholders well over the last ten years. Society publishers too have frequently charged high prices, often using the surplus to fund member services such as conferences. Whether ibraries can continue to afford these prices remains uncertain.

Copyright versus contract

It could be argued that the balance of power between libraries and publishers has swung in favour of the publishers in the last ten years. Copyright law in most EU countries (which have a certain amount of discretion in implementing EU legislation) has become more restrictive. Whilst there have been improvements in the speed and ease of use of the ILL service, it is constrained by publisher-imposed obstacles, the most significant of which is the high cost of article access by the end user, typically £20–30 when the marginal cost of processing is approaching zero! A less obvious constraint is that ILL librarians are compelled through contract law, not copyright law, to print, scan and transmit for one time use articles from journals to which they subscribe only electronically. A statement from the International Association of Scientific, Technical & Medical Publishers (STM) in 2010 described their aim to impose even further restrictions and effectively turn the clock back to pre-electronic days for ILL by insisting that users physically attend the library in order to receive a paper copy of a requested item.

However, there are signs that the wider world is recognizing this imbalance of power. Most recently in the UK, the Hargreaves report explicitly stated that ‘there is a clear need’ for the law to be adapted ‘so that the right balance is maintained’; perhaps surprisingly, the current Conservative government has accepted all the recommendations. This contrasts with the previous Labour government who did not act on the recommendations of the Select Committee on scientific publishing. An easing of the obstacles to electronic transmission would improve the quality of the ILL service immensely, but would threaten the high prices charged by many publishers.

Low use of individual journal titles

The rationale of ILL has always been that no library can hold sufficient material to satisfy all the needs of all its users. But a related factor is that some material is not used frequently enough to justify purchase, meaning that ILL would be a more cost-efficient alternative. The advent of journal usage measurement has onfirmed that most material is read not at all or very little. A number of studies over the years have confirmed this, for example:

  • CIBER's monumental study of serial usage showed that the top 50 per cent of journals viewed in life sciences accounted for between 92% and 97% of usage; other subjects showed similar skewing.
  • A Spanish study found that ‘during these two years’[of the study] ‘there are still 50% of the titles (those from which have fewer than 49 articles have been downloaded) which we could say have little or no interest for CISC researchers’.
  • An analysis by CIBER of OhioLINK's 6,000 big deal titles showed that ‘half of all journals accounted for about 93% of usage’. In other words, 3,000 titles account for only 7% of usage.

The new measure being introduced for article (as distinct from title) measurement will highlight even further the heavily skewed pattern of article readership and hence the relevance of ILL for the very large proportion of material that is read very little or not at all.

“The new measure being introduced for article (as distinct from title) measurement will highlight even further the heavily skewed pattern of article readership and hence the relevance of ILL…”

High price of serials

The high price of serials is too well known to readers of this journal to detain us here. But clearly the higher the price, the more attractive it is to use ILL for expensive and low-use material. This was highlighted by the Plan B modelling referred to above which quantified the consequences of withdrawing from the big deal offerings of Elsevier and Wiley and reverting to title-by-title selection and fast ILL.

American ‘exceptionalism’

There is a long tradition going back to the 17th century, when it was an English colony, of the US seeing itself as exceptional. ILL is certainly an example of such exceptionalism. Overall, the US bucks the global trend of the big decline in ILL; the decline in non-returnables (article copies, etc.) has not been as great as elsewhere and the growth in returnables (books, etc.) has more than compensated, resulting in a modest overall growth which is only now tapering off. Why is this? Collette Mak explains that ‘the drive to develop efficient, user-centered services provides an environment where interlibrary loan undergoes continual process improvement. Our ability to meet the research material needs of our faculty, graduate and undergraduate students allows them to see a direct connection between the service and their ability to research and publish. As a result, interlibrary loan is both a highly-visible and a highly-valued service’, and she identifies a number of common factors including: a de-facto national bibliography with 1.6 billion associated holdings, a request routing system, resource sharing code and an article transmission system. The relentless attention to the needs of the end user in the US is evidenced repeatedly by innovative services such as the recent explosion in purchase on demand where ILL requests are fast-tracked for acquisition. These innovations are covered in Interlending & Document Supply and also in the US-orientated Journal of Interlibrary Loan, Document Delivery and Electronic Reserve.

“The relentless attention to the needs of the end user in the US is evidenced repeatedly by innovative services such as the recent explosion in purchase on demand where ILL requests are fast-tracked for acquisition.”

Other innovations include print-on-demand facilities that enable book requests to be satisfied speedily and cost effectively. Very sophisticated resource-sharing consortia exist across the country – outstanding amongst which is OhioLINK. This is a consortium of 88 Ohio college and university libraries and the State Library of Ohio serving 600,000 students, faculty and staff. It provides, for example, access to 50 million books and 17,000 e-journals. In 2010, users in affiliated institutions downloaded 26 million full-text articles, borrowed 606,000 books and downloaded 4.7 million e-books; an article describes its ILL activity in some detail. It would be interesting to compare and contrast the US decentralized system (with OCLC as its chief facilitator) with the UK centralized system (with BLDSC as its facilitator). There are practical lessons to be learnt on both sides of the pond, in particular regarding request management systems where the US is clearly more advanced with the widespread use of Clio, ILLiad and Relais.

ILL in public libraries

The cost of processing returnable ILLs is high but is dropping as technology advances. The problem of costs is particularly acute in public libraries, which traditionally have been less well funded than the academic sector. Scandinavia and The Netherlands have recently introduced very effective ILL systems for the citizen. DanBib allows the Danish citizen to request any item in the Danish national catalogue from anywhere in Denmark to be delivered anywhere in Denmark – without charge. As a result, ILL requests have been about two million for the past five years for a population of about 5.5 million! However, the scheme is reliant on substantial central government financing. A similar system has been introduced in The Netherlands. The UK is crippled by the inability of any government to see the benefit to the citizen of being able to access the national collection; hence lack of funding and the consequent high price to the citizen makes ILL a tiny part of the public library system. However, OCLC and The Combined Regions (TCR), which provides services for public libraries in most of the UK, have just launched a new service that brings the vision of access to the national collections for the ordinary citizen a bit nearer.

Open access

For the past ten years, I have written a literature review for Interlending & Document Supply and referenced an average of about 40 articles and reports in each review. The number of references that are freely and immediately accessible has risen to about 35%. The percentage varies of course from discipline to discipline, for example over 90% of preprints are freely available in high energy physics. Most academic libraries now check all ILL requests for free accessibility before passing a request through for a (paid-for) ILL. Clearly, this has had an impact on ILL, but how much more will depend on how open access (OA) develops over the next few years.

“… the impact of OA on ILL will increase significantly as we transition to a new period of free access to the published literature – at least in STEM subjects.”

At the time of writing (December 2011), the UK government has published a White Paper on innovation and research strategy which states that all publicly funded research should be open access. This should accelerate the trend towards OA and shift the burden of payment from libraries to authors and funders; good news for libraries, but bad news for funders and non-funded authors. And especially good news for publishers who may continue to be able to charge high prices, though via author fees rather than subscriptions. Obviously, the impact of OA on ILL will increase significantly as we transition to a new period of free access to the published literature – at least in STEM subjects.

Advances in ILL

ILL has changed as a service along with all the contextual changes noted above. Until the late 1990s, a user would scour catalogues, then complete and submit a handwritten request. This would be checked by an ILL librarian and submitted by post or fax and, if to the BLDSC, by automated request transmission using codes obeying strict format constraints. Today the user can cut and paste a reference into a form which will automatically format the request and transmit it to the library from their desktop at the press of a button. Often this can be sent electronically to a supplier and returned directly to the user. A process that took many days, if not weeks, a mere ten years ago now often only takes hours and rarely more than two days. Books and other returnables inevitably take longer and the ILL of e-books continues to be an issue – especially in public libraries. But it is a sad irony that with all the improvements in speed and ease of use, these measures are eclipsed for the end user by the immediacy of e-databases. Only the relaxing of publisher constraints will enable ILL to match this immediacy.

Conclusion – the continued importance of document supply

I have now hung up my clogs after ten years of editing Interlending & Document Supply and before that, 26 years working at the BLDSC. My experience over those years suggests that, with all the ups and downs, ILL is in rude health. It seems to me self-evident that libraries, academic, public and commercial, should have as part of their mission: ‘Can't find it? We'll get it for you’. The continued importance of ILL follows from the vast amount of material that is not easily and immediately accessible. Access to e-catalogues and e-serials has exposed countless more references to the attention of searchers that are not immediately available. In addition, the high price and low use of many academic journals means that libraries will find ILL a cost-effective alternative to subscription.

And finally, open access will not replace paid-for journal access in the foreseeable future, especially in the humanities and social sciences where funding for author fees is never going to be forthcoming to any significant extent.

“It seems … self-evident that libraries, academic, public and commercial, should have as part of their mission: ‘Can't find it? We'll get it for you’.”