Cooperative Collection Management in the Humanities in the Netherlands

The coverage of the aggregate collection of Dutch academic libraries (the 'national collection') was assessed in 1996. At that time, the libraries had already been confronted with a strong decrease in purchasing power for several decades, due to high price rises and a large increase in the number of books and journals published. It was feared that, as a result of this, not only local collection development would suffer, but also that the diversity of the national collection could be endangered. It was expected that decreases in purchasing power would cause libraries to acquire the same titles, at the expense of more specialised and rare works.

Table 1 shows the most important results with respect to monographs.The first column lists the disciplines included in the study and the size of the random sample.The second column shows the coverage of the strongest Dutch library and the third column the coverage of the national collection.From this it emerges that not even 45% of the recent acquisitions from the German Sondersammelgebietsbibliotheke is held in the Netherlands.The results also revealed how the collections complement each other.For example, the coverage of the strongest Dutch library in the area of theology (in this case the University of Groningen) is just 22%, whereas the aggregate coverage of all Dutch academic libraries amounts to 49%.
An initial examination indicated that not all of the missing titles seemed to be relevant, at least not for Dutch researchers.Therefore, per subject area the titles not held in the Netherlands were presented to two subject librarians for assessment.A title was considered to be relevant if at least one of the two subjects librarians assessed this to be the case.The coverage percentages could now be recalculated.The fourth column of Table 1 shows that more than 60% of the relevant titles are held in the Netherlands.[ * These disciplines contain a more or less extensive package in the areas of the humanities and social sciences, but at the very least they include history, language and literature.
The coverage of monographs for the humanities, social sciences and sciences was now 63.3, 62.6 and 66.4% respectively.A similar study was carried out for journals.When only the relevant titles were taken into account, the coverage was 52.4,68.0 and 86.2% respectively.The combined coverage of monographs and journals was 58.9% for the humanities, 64.3% for the social sciences and 77.6% for the sciences.Within these disciplines there were considerable differences between subject areas.For a number of humanities subjects the scores for both the monographs and journals were fairly low.Table 2 summarises the most important results.A more fully report has been published in Alexandria (Voorbij, 1996).The question remains as to how these results should be interpreted.No single library and not even a national collection can be expected to be complete.Cost-benefit considerations alone prevent this.Therefore a national collection clearly has an upper limit.However, due to the lack of an absolute standard there is no consensus as to what an acceptable minimum level is.Accordingly, the level of the national collection could perhaps best be described as no longer sufficient, although for the sake of convenience it could reasonably be assumed that the most important titles are held somewhere in the Netherlands.
The collections in the sciences seemed to be the most satisfactory, certainly in view of the relatively high coverage of journals, the most important document type in that discipline.The situation was least satisfactory for the humanities.It was expected that without additional financial measures, the purchasing power, and with this the diversity of the national collection, would decrease even further.

THE HUMANITIES PROJECT
The Netherlands Organisation for Scientific Research (Dutch acronym: NWO) shared this concern and provided additional funding to improve the collection development in the humanities.Initially, a sum of € 2,268,901 was awarded for a two-year period between 1998 and 2000.
[2] The plan Bibliotheekvoorziening Geesteswetenschappen (Humanities Project) was drawn up for the use of these funds.This included the following aspects: • Seven academic libraries with important humanities collections were involved in the project: University of Amsterdam, University of Groningen, Leiden University, Utrecht University, Free University Amsterdam, Radboud University Nijmegen, and the National Library in The Hague.
• A total of 38 subject areas were distinguished, varying from Romanian language and literature to history.
• A subject area was awarded to the institute, which had the greatest number of teaching commitments and the highest acquisition budget for that area.
• Wherever possible only one library was responsible for a subject area.For 10 subject areas, more than one institute was appointed.This mainly concerned the larger subject areas such as history, history of art, philosophy and theology.
• The extra funds were distributed over the 38 subject areas in proportion to the 1997 acquisition budgets of the libraries that had been assigned responsibility.Thus, subject areas with high acquisition budgets received high additional funds, subject areas with low acquisition budgets received low additional funds.
• The funds were to be spent solely on the acquisition of documents not already held in the Netherlands.
Within these boundaries the libraries could acquire whatever books, journals or electronic publications they wanted.
• The libraries were responsible for the processing and storage costs, for making the acquisitions freely available via interlibrary loan, and for maintaining the present level of collection development in the subject areas for which they had no national responsibility.
In the end, 37,660 publications were acquired.An amount of €3,176,000 was made available for a second round in the period July 2001-July 2004.Several changes were made with respect to the first round: • In order to stimulate structural financing by the universities, NWO imposed the condition that the libraries would contribute €680,000 (21%) towards the costs.
• In the first round it became apparent that the additional budget for some subject areas was too high.
Therefore a limited redistribution of the funding over the subject areas took place.
• In recent years, the National Library has more strongly profiled itself as the library for Dutch history, language and culture.As a result of this, the subject areas of French language and literature and English language and literature, which it had been awarded, no longer fitted naturally into the library's profile.A limited exchange of subject areas therefore took place between the libraries.
• Part of the funding was used to obtain licences for electronic resources on the condition that at least five participants were in favour of a licence.Efforts were made to establish licensing agreements for resources, which were important for as broad a target group as possible in the humanities and which would be difficult for individual institutes to finance.In particular Project Muse Arts & Humanities and PCI Full Text Collection 1, 2 and 3 satisfied these criteria.PCI could even be purchased; in forthcoming years only a small access fee has to be paid.In a situation where additional funds are only available on a temporary basis, this is an attractive option.
At the start of the project, NWO had indicated that it would finance a maximum of three rounds.The awarding of the third round was dependent on the results of an intermediate evaluation and the willingness of the libraries to increase their own contribution.

EVALUATION
The Humanities project was evaluated from four different viewpoints.Firstly, it was investigated whether the diversity of the aggregate Dutch humanities collection had increased.Secondly, the books bought with the project's funds, were analyzed by year of publication and language of publication.Thirdly, a questionnaire was sent to the subject librarians in the humanities at the seven institutes.Fourthly, interviews were held with humanities researchers at each of the seven libraries, as these ultimately form the target group of the project.

Diversity of the aggregate Dutch book collection
To gain an insight into the diversity of the titles held by Dutch libraries, counts were carried out in the Pica cataloguing system (Dutch acronym: GGC).
[3] The second column of Table 3 shows that at the end of 1997, the GGC libraries owned 19,449 different humanities books, published in 1996.For the social sciences this was 17,473 books and for the sciences 9,655 books.As the Humanities project started in 1998, these data refer to the situation before the project.
They can be taken as baseline measurements and were awarded the index figure of 100.After the introduction of the project the diversity of the collection in the humanities showed a marked increase: compared to the 19,449 unique titles published in 1996 and acquired no later than December 1997, there were 22,454 unique titles published in 1999 and acquired no later than December 2000, and unique 23,775 titles published in 2002 and acquired no later than December 2003.This strongly contrasts with the developments in the other disciplines.Over a period of six years, the number of unique acquisitions in the social sciences decreased by 16%, while the number of unique acquisitions in the sciences decreased by 22%.Although the favourable results for the humanities can easily be ascribed to the project, it should be noted that the counts relate to all GGC libraries, not solely to the libraries that were involved in the project.However, the seven libraries are by far the most important libraries for the humanities [4] and therefore account for most of this effect.An increase or decrease in the number of acquisitions does not necessarily mean an improvement or deterioration.One can draw such a conclusion only after comparing data on library acquisitions with data on book publishing.Good sources for data on book publishing are the reports from vendors such as Yankee Book Peddler (YBP), Harrossowitz, Blackwell and Baker & Taylor.These companies process titles from a large number of publishers, which may be relevant to academic libraries.The study made use of the New Title Reports from YBP.The fiscal years 1997/1998, 2000/2001 and 2002/2003 were chosen.These are reasonably comparable with the years chosen for the GGC counts.In these years YBP processed respectively 45,671, 51,146 and 58,766 new titles.These originated from more than 900 publishers, including about 600 American and 200 British publishers.Table 4 provides a brief overview of the number of titles that YBP processed in the three years stated, for each of the three disciplines.From this overview it is clear that book publishing has strongly increased over the past five years, not only in the humanities but also in the social sciences and the sciences.This is hardly due to a broader coverage of YBP.For example, the number of trade and professional presses covered increased from 745 in the period July 2001−June 2002 to no more than 785 in the period July 2002−June 2003, and the number of university presses covered increased only slightly from 175 to 178.There therefore seems to have been a real increase in book publishing across all disciplines.
It may be concluded that the increase in the number of humanities books held by Dutch libraries has kept pace with the increase in the number of books published.There is a decline in the other disciplines.However, this has been compensated to a fairly large extent by the strongly increased availability of electronic journals, which are of great importance to these disciplines in particular.

Analysis of acquisitions
The project partly aimed to fill gaps that arose in the past two or three decades.However, the first part of the evaluation already showed that a large number of recent publications have been acquired.An analysis of the acquisitions in nine subject areas at three libraries, purchases with funds from the Humanities project, confirms this impression.Table 5 reveals that the number of acquisitions published before 1991, was only 9.5% in the first round and just 3.2% in the second round.This may either mean that older publications were hard to get or that priority has been given to more recent titles.On the other hand, although this is not clear from the table, various libraries have acquired older materials in microform.

63.7%
Table 6 shows that, both in the first and second round, more than half of the titles acquired were written in English.With 22% in both rounds German is still reasonably represented.Only a slight minority of the acquisitions are in languages such as French or Italian.The surveys among the subject librarians (part 3 of the evaluation) and the end users (part 4 of the evaluation) showed that the collection needs to be improved further in these respects in particular.

Experiences of subject librarians
To gain an understanding of the experiences with the project to date, a questionnaire was sent to the subject librarians in the humanities, at the seven participating libraries.The questionnaire was completed by 66 respondents.Among them were 26 persons in charge of extra funds from the Humanities project.The most important results are:

Awareness and suggestions for acquisitions
All subject librarians in charge of extra funds brought the project to the attention of the faculty at their own university.
In both the first and second rounds they received an average of 11−25 suggestions for acquisitions from their own faculty; five subject librarians received more than 100 suggestions for acquisitions in both rounds.However, the faculties from the other participating universities without extra funds for that subject area were not always informed.Also, the subject librarians without extra funds submitted relatively few suggestions for acquisitions to the libraries with extra funds.Clearly there is a need for more promotion and awareness of the Humanities project.

Adequacy of the national collection
On a scale of 1 to 10 the aggregate book collection was given a score of 7+ by the subject librarians and the aggregate journal collection received an average score of almost 7.Although this is a subjective and extremely inaccurate measurement, it nevertheless indicates that the Dutch collections are still of a reasonable standard.Gaps have only arisen in recent decades due to a decrease in the purchasing power of libraries.The subject librarians believed that the project has contributed to a recovery.According to 43 subject librarians, the project has brought about some improvement in the collection of humanities books and 21 persons even spoke of a significant improvement.The same applies to journals, although to a lesser extent.
Nevertheless, about two-thirds of the subject librarians pointed to gaps within their own subject area.These concern specific sub-areas, electronic text files, and documents in languages other than English and Dutch.Moreover, there is considerable concern that, unless extra funds are continued to be made available, new gaps will arise immediately.It is significant that, bar a few exceptions, the subject librarians had no difficulty in spending the extra funds.Again with a few exceptions they were of the opinion that a possible third round required at least the same level of funding as the second round.

Project guidelines
Some subject librarians indicated that in some cases it should be allowed to purchase publications that are already held by a Dutch library.This applies in particular to works that would fit well into the collection and expensive works such as bibliographies and reference volumes, which cannot be borrowed.

Experiences of humanities researchers
Individual or group interviews took place with the end users at each of the seven libraries.A total of 20 interviews were held with 31 interviewees.All of the interviewees were known as active library users.Therefore, although the findings cannot simply be generalised to the entire population, they are highly informative.Moreover, where applicable, they concur with the results from the other parts of the study.

Opinion about the local and national collection
In general, users were reasonably satisfied about their own local collection; the most important journals and reference works and a significant proportion of the monographs can be found there.This mainly applies to researchers at the older institutes and researchers in the areas where the library of the researcher had extra funds from the Humanities project.For various reasons, a lot of value can be attached to the opinions of the interviewees.Not only are they specialists in their fields, but they often track down titles via citations, references in bibliographies and suggestions from colleagues.Only then do they go and search for the relevant titles in the catalogue of their own library.This means that they can also come across titles, which are not held there, so that they form a good opinion about the quality of the collection.Furthermore they do not pose unrealistic requirements.Most researchers realise that not everything can be held in the Netherlands and are happy with the fact that expensive research literature is in principle only held at one location.
Several reservations were expressed about the national collection.Some researchers noticed a tendency towards a stronger focus on Dutch and Anglo-Saxon subjects.For those who want to carry out research into an Anglo-Saxon philosopher sufficient material is available; however there is no literature available for someone who wishes to study a Polish philosopher.And whoever wishes to study modern French literature must go abroad because even the seminal works are not acquired in the Netherlands.Another criticism is that publications that are neither fundamental nor specialised are often only held at one library in the country.
The interviewees also included enthusiastic users of electronic journals.They pointed out that mainly Anglo-Saxon journals are available, whereas for research purposes more German, French or Italian journals are needed.Moreover, little is available on specific Dutch subjects.This is more of an encouragement for publishers to issue electronic versions of these journals, rather than a criticism of the acquisition policy of libraries.
The expectations with respect to the national collection were also realistic.It was realised that regional and local journals or source publications, and difficult to access material such as foreign dissertations cannot always be found in the Netherlands, and it was stated that ideally Ph.D. students should be allowed to conduct part of their research in foreign libraries.

Use of other libraries
An inherent part of the project's philosophy is that one or two libraries per subject area receive the resources for the acquisition of specialised literature.With this, researchers are sometimes forced to use libraries elsewhere in the country.
Experiences with interlibrary loan are variable.Some consider it to be a 'fantastic service', whereas colleagues at the same institute consider it to be inconvenient and time-consuming.The facilities for researchers also differ per library.
At some institutes there are no costs for staff to use the interlibrary loan system, at another institute the old, lower fees apply, at a third institute a certain amount of funding is available to researchers, and at a fourth the researchers must cover all of the costs themselves.However, in general it would seem that the majority of researchers only use this facility to borrow a book if they are certain that it is useful, and not just to browse through it.
Another option is to visit libraries elsewhere in the country.Almost nobody from the Randstad area (west part of the Netherlands) considers it inconvenient to visit another library in the Randstad, but they are not as willing to visit Nijmegen and Groningen, which are more remote.
Experiences with using the interlibrary loan system to obtain books from abroad are predominantly negative: it does not function well, is incredibly complicated and is far too expensive.
It may be concluded that a researcher who knows exactly what he needs, can obtain it via the interlibrary loan system, although the costs can sometimes form a barrier.For researchers who want to study collections on site, the distance to the library concerned can sometimes form a barrier.As a result of this, some faculty advocated, just like the librarians, in some cases the acquisition of a second copy of a book, in particular if the first copy is located outside of the Randstad area.

Experiences with the project
In some libraries the gaps have been filled in close consultation with the researchers.For example, several source collections and journal files (in digital form or on microfiches), monograph series, titles on specific aspects which were underrepresented and digital reference works have been acquired.In certain subject areas expensive electronic resources have been acquired which would have been too expensive to purchase without the extra funds.
The effect of the project was not so explicitly noticeable in all cases.This would also seem to be obvious: As a researcher you usually do not know what would not have been purchased had there been no extra funds.It was, however, found that the publications sought were generally held and that many suggestions for acquisitions are now honoured, which was not the case five years ago.
Other comments confirmed the findings of the other parts of the study.Firstly, disappointment was expressed that gaps could not be filled up with the project's funds if the title was already held elsewhere in the Netherlands, even if this could not be borrowed or was owned by a library far away.This applies in particular to expensive reference volumes.
Secondly, it has become clear that improvements in the communication are needed.Many researchers were not even aware of the extra funding at other libraries, and certainly did not realise that suggestions for acquisitions could be submitted.Thirdly, there was strong support for the continuation of the project.Just like the subject librarians, the researchers pointed out that an increasing number of books are being published.The fear was expressed that if the extra funds were to cease, large gaps would once again arise within a short space of time.

CONCLUSION
Together, the different parts of the evaluation point to the same conclusions.The most important of these are: • The diversity of the national collection in the humanities has increased considerably, and both the subject librarians and the researchers are of the opinion that the Humanities project has made a significant contribution to this.
• As an ever-increasing number of books are being published, it is vital to continue the project.
• Improvements are needed in several areas: in special cases the acquisition of a second copy using the project funds should be possible; the acquisition of material in languages other than English and Dutch must, wherever possible, receive more attention • The faculty / the end users should be made more aware of the project in general and the possibility to submit suggestions for acquisitions in particular.

FOLLOW-UP
Based on the evaluation study, NWO has decided to award funds for a third and final round.A necessary condition for this was that the libraries realise a structural solution by increasing their own contribution and by guaranteeing the continuation of additional funding for the first two years after the conclusion of the project.Table 7 reveals that the contribution from the libraries has clearly increased over the three rounds.On a yearly basis this was €226,890 in the second round and is €340.335 in the third round (+50%).A number of additional agreements have been reached with respect to the use of funds in the third round: • A limited redistribution of the funds will take place; the National Library will transfer responsibility for several subject areas that no longer fit into its profile to other libraries.
• Project funds may be used to acquire a second copy of expensive works that cannot be borrowed.Funds may also be used to acquire titles already held elsewhere if these are particularly missed and the library concerned has a more or less complete collection in that area.
• Part of the funds can once again be used to obtain shared licences for electronic publications.Again purchase is preferred.It is expected that about 30% of the funds will be used for this purpose.
• Libraries are free to use part of the remaining funds to acquire individual licences.Where possible and desirable the priority is to acquire non Anglo-Saxon electronic files.However, at least 50% of the total budget must be spent on subject-specific printed materials.
• The existence of the project and the possibility to submit requests for acquisitions must once again be brought to the attention of researchers.
The Humanities project would not have been a success without the extra financial resources made available by NWO.
Obviously, without the additional funds there would not have been a project at all.However, some other factors also contributed to the success: • The empirical research: the evaluation of the national collection in 1996 and the evaluation of the effects of the project.
• The tradition of cooperation between libraries.In this project: the distribution of the funds and the responsibilities for the various subject areas between the libraries.
• The willingness of libraries to provide part of the additional funds and to finance the processing and storage costs.
• A good insight into the holdings and orders of other libraries, in particular via the Pica shared cataloguing system or the national union catalogue.

NOTES
1.For theology an initial coverage of 49% was established: 206 of the 421 titles were held in the Netherlands.Of the 215 titles not held, 73 were considered to be relevant by at least one subject specialist.The coverage of relevant titles was therefore 206/(206 + 73)=73.8%.For the sake of convenience it is assumed that all of the 206 titles held are relevant.
2. The plan required additional financing of €1,361,000 per year for a period of at least five years after which the situation, which might be fundamentally different due to developments in ICT, would need to be reviewed.This amount was determined as follows: The average coverage of the aggregate Dutch collections with respect to the corresponding German library was about 60% for the humanities and about 80% for the sciences.The coverage for the humanities would have to increase to the level of the sciences, and therefore an increase of 80/60 or 4/3 would be required.The joint acquisitions budget of the seven libraries for the humanities at that time was € 4,084,000.This amount was increased by 4/3: from €4,084,000 to €5.445.000, a difference of €1.361.000.
3. The program Anaconda developed by the University of Groningen was used for this.
4. Anaconda also calculated the joint coverage of the four strongest libraries.For the humanities these are, almost without exception, the project participants.Together these four libraries account for about 95% of the collection.This applies to all of the three report years studied.

Table 1 :
1] Coverage of monographs per subject area

Table 2 :
Coverage of relevant publications

Table 4 :
Number of academic books processed byYBP, 1998YBP,  −2003 * Not included are general, military and maritime studies, and library studies

Table 5 :
Acquisitions per year of publication

Table 6 :
Acquisitions per language area

Table 7 :
Funds available for the project