On Saving the Warburg

On 6 November​ , after ten days of legal argument in the High Court, judgment was handed down in the dispute over the University of London’s obligations towards the Warburg Institute. The institute developed out of the private library of Aby Warburg (1866-1929), a wealthy art historian in Hamburg, who was supported by his four brothers, all of them bankers. Warburg had unusually wide interests, collecting books not merely in his immediate discipline, but also in the history of religion, magic, science, literature and related topics. His intention was to create an institution that would support research and he hoped that it would in time be absorbed by the new University of Hamburg. But this did not happen, and although after Aby’s death his brothers continued to support the library (the term ‘institute’ was already used frequently), the world financial crisis meant that the Hamburg branch of the family had to be subsidised by the two brothers in the United States.

Shortly after the Nazis came to power in early 1933, the library, as a Jewish institution, had to close its doors. Its director at the time, Fritz Saxl, lost his professorship at the University of Hamburg, as did Erwin Panofsky and Ernst Cassirer, both of whom were closely associated with the institute. Various proposals for moving the library abroad failed because the destinations suggested – Jerusalem, Rome and Leiden – would only accept it as a gift, and this would have left the family with a punitive tax bill for the export of assets. A group of British academics and public figures led by Lord Lee of Fareham, a former cabinet minister who had given Chequers to the nation, recognised that the only way forward was to offer to take the library on loan. This arrangement, valid for three years, was negotiated with the authorities in Hamburg by Max Warburg, and the whole collection of books, photographs and furniture, together with most of the staff, arrived in London in December 1933. At first the institute was accommodated in Thames House, now the headquarters of MI5. Funds for upkeep and staff salaries were provided by Samuel Courtauld, who, with Lee and Sir Robert Witt, had recently founded the Courtauld Institute of Art.

By 1935 it was clear that there was no realistic chance of a return to Hamburg in the foreseeable future. At this point Felix Warburg in New York began to press for the removal of the institute to the United States, but was persuaded that this would breach the loan agreement, with disastrous consequences for the members of the family still in Germany. Max extended the loan agreement for another seven years, and at the last moment Courtauld promised to provide increased funding. Thames House was no longer available, but after pressure from Lee, the University of London offered space in the Imperial Institute at South Kensington, in rooms recently vacated by the University Library, which was moving into the newly constructed Senate House. Conditions were less than ideal, as Saxl reported to Lee in 1938:

The two principal rooms cannot be used during November. They are being used as store rooms for the Needlework Guild, and are adorned with such inscriptions as ‘Babies’ and ‘Children’. Last week Queen Mary was here, and young Mr Blunt, who is on our staff, was presented to her, as his aunt happens to be the chief official of the Needlework Guild.

By this time the idea of incorporating the institute into the University of London had been raised, and in July 1939, after the last of the Warburg brothers had left Germany, Max’s son Eric wrote to Lee that the family was prepared to offer the library to the university, provided that ‘such handing over is accompanied by an assurance that London University will see to it that it is maintained in the same form as hitherto’. The outbreak of war prevented further consideration of the proposal, and in 1940 supporters in the US, notably the National Gallery of Art and the Library of Congress, suggested that they look after the library for the duration of hostilities, and promised generous funding. The proposal was rejected by Lee and Witt, who feared that once the library went it would never return. After the librarian was killed by a bomb, the books were put into storage and a small group of staff retreated to a rented house near Slough where they continued to research and publish.

In 1942, not long before Courtauld’s funding was due to end, supporters of the institute, including Kenneth Clark and Lee, decided to approach Rab Butler, the secretary of state for education, who happened to be Courtauld’s son-in-law. His first idea was to attach the Warburg to the Victoria and Albert Museum, financed by a grant in aid, a proposal rejected by a senior civil servant at the Treasury, who pointed out that ‘if a grant in aid to the Warburg Institute were to be proposed to Parliament, it would be asked who or what is the institute. The answer seems to be Dr Saxl and such persons as he may select to help him. We cannot but think that this answer would be found very unsatisfactory. I am not particularly referring to the fact that Dr Saxl is not of British origin, though that does not make matters any easier.’

The next idea was to attach the institute to a university, with the Treasury increasing funding to the University Grants Committee (UGC) to cover the running costs, thus avoiding the need for an annual vote in Parliament. It was decided that since the institute ought to be close to the national museums, the only possible destination was the University of London. The university agreed, on two conditions: that the UGC would cover the costs of the institute by a real and permanent increase in the annual grant to the university, and that at some future date it would provide a new building.

In November 1944, the university and the Warburgs signed a trust deed guaranteeing the institute’s future. It is just a page and a half long, consisting of a recital and four operative clauses. The university undertook to maintain and preserve the Warburg Library in perpetuity, to house it as soon as practicable ‘in a suitable building in close proximity to the university centre at Bloomsbury’ and to ‘keep it adequately equipped and staffed as an independent unit the whole to be known as “The Warburg Institute”’, and this ‘without regard to contributions from outside sources’. The university was to establish a committee of management, to include where appropriate two representatives of the Warburg family. Access would be guaranteed to members of the university and others engaged in research. The last clause specified that the director would be of professorial standing.

The Warburg was governed in the same way as the other Senate Institutes of the university. Unlike the colleges, which were largely autonomous, they were managed by committees answerable to the Senate, whose responsibility covered primarily academic matters. But whereas the other institutes, such as the Institute of Historical Research and, later, the Institute of Advanced Legal Studies, were managed by committees representing a single subject area and were meant, in the words of a former principal of the university, ‘to co-ordinate the facilities for postgraduate studies in a specific academic field’, the membership of the Warburg committee was always interdisciplinary, reflecting the distinctive character of the institute.

Over the following decades the library grew from about 80,000 volumes in 1944 to its present total of about 350,000; the photographic collection now contains about 400,000 images. The staff taught undergraduates and postgraduates, and many scholars from Britain and abroad came to depend on the library, which is open-access. Although resources were often scarce, the future seemed relatively secure in 1986, when the Warburg and four other Senate Institutes with important libraries were singled out by the UGC as centres of outstanding excellence with a clear national role, for which sufficient resources were to be provided to protect current levels of activity. This decision was taken during a period of major change in London University itself: the colleges were gradually acquiring greater autonomy and dealing directly with the UGC and its replacements, the Universities Funding Council and the Higher Education Funding Council for England (Hefce).

As the central university lost much of its power and many of its functions, it began to push the various Senate Institutes together. An umbrella organisation was created, the University of London Institutes of Advanced Study, which was replaced in 1994 by the School of Advanced Study (SAS). One of the rationales advanced for the move was that it would encourage interdisciplinary research. This was something of a mantra at the time, although the justification wasn’t obvious in this instance: the Warburg had always been interdisciplinary, and the other institutes were supposed to serve distinct academic communities. SAS received a block grant from Hefce for all the institutes. At about the same time the committees of management were turned into boards and then advisory councils, losing their management functions. As the colleges became increasingly independent, undergraduate teaching was no longer organised on an intercollegiate basis and the staff of the Warburg were left teaching postgraduates only.

In early 2003, about a year after I became director of the institute, a post I held until my retirement in 2010, a committee charged with considering the future of the University of London Computer Centre recommended that the University Library, the seven libraries within the School of Advanced Study and the computer centre be brought together in a new Division of Library and Information Services, with its own director. This proposal was accepted by the Collegiate Council, the committee of heads of colleges which now largely runs the university. The new body was called the University of London Research Library Services (ULRLS).

None of the institutes was consulted before this decision was taken, and no evidence was made public concerning the savings it was expected to achieve. The plan predictably caused several problems. For example, the university was unaware that most of the books in the library of the Institute of Classical Studies actually belonged to the Society for the Promotion of Roman Studies and the Society for the Promotion of Hellenic Studies, neither of which took kindly to the prospect of losing control of their property. There followed years of tense negotiation, with the societies at times threatening to take their books away. The Institute of Advanced Legal Studies was also unhappy about being included in ULRLS, because law libraries, the management of which requires special expertise, are not normally run in conjunction with other libraries. One immediate consequence of the new policy was that the budgets of the institutes were split in two: in each case, the part covering the library was transferred to ULRLS, and the other to the institute itself. As with the Institute of Advanced Legal Studies, the library budget for the Warburg Institute was then transferred back to the director, a process justified on the grounds that both institutes were remaining in purpose-built accommodation, whereas the others were already in or scheduled to move into Senate House.

The decision to combine the libraries was taken in ignorance of the Warburg trust deed. I repeatedly argued that the specification about preserving the Warburg ‘as an independent unit’ prevented the library from being part of the new organisation, though it could co-operate with ULRLS, for example in avoiding duplication in the purchase of books. I also saw obvious practical problems with the scheme. More than a third of the Warburg library’s acquisitions come from gift or exchange, and most of the rest are bought with our own resources rather than public funds. And the institute has always made a practice of employing scholar-librarians, actively engaged in research and often in teaching. The head of ULRLS made no secret of the fact that he thought our librarians over-qualified, and did nothing to reassure me that Warburg funds would be used exclusively for the benefit of the institute. The Warburg Institute’s advisory council, through its chairman, Keith Thomas, took a view similar to mine. Our protests did not have much effect, because it wasn’t until late in 2004 that the university even asked to see a copy of the deed, not being able to locate its own. By then the vice-chancellor, Graeme Davies, had already assured me and the directors of the other institutes that he would respect its terms. ‘There is no plan by the university to break any trusts,’ Davies and the dean of SAS, Nicholas Mann, stated in 2007. This sounded reassuring, but it emerged in the recent court proceedings that no one in the university bothered to show the trust deed even to their in-house legal adviser until the summer of 2008.

Up to that time the university had been careful not to concede that the trust deed obliged it to treat the Warburg differently from the other institutes, and in particular to exclude it from the process of library convergence. Early in 2007 a SAS review panel’s report on the institute was rejected by Mann for two reasons: first, because it concluded that the proposed inclusion of the library within the ULRLS was ‘contrary to the wishes of the Warburg Institute, the Warburg family and the scholarly community as a whole, both in the UK and abroad’, and second because of the suggestion that convergence might be contrary to the trust deed. Soon afterwards I was required to provide a self-evaluation document about the institute for Ivor Crewe, who was about to carry out a major review of the funding of SAS for Hefce. ‘How far the implementation of the convergence policy is compatible with undertakings made by the university to the Warburg family has yet to be tested,’ I wrote. This muted, but accurate, reference to the trust deed was eliminated by the vice-chancellor, who substituted this passage: ‘Within the convergence policy the university has allowed significant autonomy to the institute recognising both its geographical location and its individualised cataloguing position.’ He then forwarded the document to Crewe as if it was entirely my work.

Disturbed by these developments, early in the summer of 2007 I consulted a solicitor about the deed, and she advised me that the university’s policy on library convergence seemed to be inconsistent with it. Keith Thomas wrote to the vice-chancellor, reporting the solicitor’s views and proposing that the university should also seek legal advice; he also suggested that if the advice was different from ours, we should jointly seek counsel’s opinion. When after 11 months the vice-chancellor finally replied, he did not refer to our suggestions at all. We then sought advice from Christopher McCall QC, who confirmed that our concerns were justified.

Only then did the university show the deed to its in-house legal adviser, after which both parties agreed to approach the Charity Commission for guidance. At the end of 2008, the commission urged the university to seek legal advice about the meaning of the deed and to ensure that it avoided conflicts of interest in the administration of the institute. By this time a new body had been established to run the university, the Board of Trustees. At first we had high hopes that the board would be more sympathetic than the vice-chancellor, but we were soon disappointed. During the only visit made by the board to the institute, its chairman, Jenny Abramsky, refused to allow me to raise any concerns about the deed. She also made no effort to have a substantive meeting with Keith Thomas to discuss the situation.

The Board of Trustees did approach a lawyer, Robert Pearce QC, who didn’t commit himself on how far the university’s policy was consistent with the deed, but instead produced a draft for a new deed. The purpose, we were told, was to avoid the problems of interpretation of the original document. But it went much further than that. The trust would now explicitly cover only those books and photographs handed over in 1944; the university undertook to keep these together, but not necessarily in London; and the roles of the director and committee of management were omitted. The changes, we were informed, were needed in order to address ‘the current circumstances of the Warburg Institute’.

Those circumstances​ , which were of the university’s own making, were very threatening. In 2008, the university vastly increased the charges it levied across its entire estate for the services it provides. These charges were calculated on the basis of space occupied; and, perhaps in order to subsidise Senate House Library, the charge for open-access book stacks was fixed at about eight times the rate for closed-access stacks. One rationale for the charge was that it provided an inducement to institutes to occupy less space. Of course, while academics might be able to reduce their office space, there was no possibility of significantly reducing the space occupied by large open-access libraries like the Warburg’s. More than 98 per cent of our books are on open access, as against 90 per cent for the Institute of Classical Studies (which was able to negotiate a more favourable rate for its space charge because most of its books did not belong to the university) and 53 per cent or less for the other SAS libraries.

The increase in the space charge made it much more expensive to maintain the libraries, yet SAS decided not to alter the share of its funds going to them. As a result the institutes with libraries were disadvantaged in comparison with those without, and the Warburg most of all. Its problem was compounded by the fact that it was and is housed in a large building designed for it according to specifications agreed with the university and the funders. The new acting dean of SAS, Roderick Floud, pointed out in August 2009 that ‘no solution has yet been found to the circa £500k deficit incurred by the Warburg Institute library, which cannot be absorbed within the Hefce special funding to the School.’ This deficit, which appeared in the accounts of ULRLS rather than SAS, was an inevitable consequence of the decision to retain the previous division of library and non-library funding despite the great increase in space charges. According to figures supplied by the university, the cost of providing services for the Warburg’s premises is about £250,000 per year, yet the space charge amounted to almost three times that figure.

When I asked Floud how I was supposed to eliminate our deficit, he suggested that I make the library stacks closed-access, which would have reduced the space charge by more than 80 per cent, and concentrate all the books currently arranged over four floors on just two, in rolling stacks. The floor loadings in our building would not have permitted this solution, which would in any case have destroyed most of the value of the library to scholars. Crewe had recognised that access to the library must remain open. In court, the university agreed, but added that it had never contemplated changing this arrangement.

In a witness statement submitted to the court the present dean of SAS, Roger Kain, did not suggest how the Warburg might eliminate its deficit, beyond pointing to the need for additional activity. This seems to mean that the institute’s staff should obtain more research grants, some of which, such as those from the European Research Council, include overheads of as much as 20 per cent, or perhaps do more fundraising. Given the size of the staff and the scale of the deficit, it would be entirely unrealistic to suppose that such a policy could have a significant impact on the deficit, and in any case overheads on grants are not meant to cover the fixed costs of research libraries. As it is, the Warburg’s performance in obtaining research grants is already among the best of any institute in SAS, and it is the only one to have obtained a major ERC grant; and the trust deed specifies that funding obtained from outside sources, which would cover both research grants and fundraising, cannot be used to offset the university’s obligation to fund the institute. Both the dean and the present vice-chancellor have ruled out making significant changes to the division of SAS funds between library and non-library, yet it would be very easy to produce a division that would better reflect the current level of space charges on the libraries, by using the figures for space occupied by the libraries when the original division was made in 2003. Whereas staff and salaries and acquisition costs have increased roughly in line with the Hefce grant, in 2007-08 space charges for the Warburg increased from £283,000 to £711,000, of which £526,000 was attributable to the library.

Because of the institute’s chronic deficit, due entirely to the level of the space charge, the university took away all of its reserves, including even gifts, and profits or temporary surpluses on our publishing activities. Although the situation was clearly unsustainable the university consistently refused to enter into meaningful discussion about it. Accordingly, the institute’s advisory council finally decided to seek guidance from the attorney general, who represents the public interest in matters relating to trusts. He told the university to ask the High Court to provide an authoritative interpretation of the trust deed, and, most unusually, instructed a QC to represent him at the proceedings. On every issue on which the court was asked to decide, the attorney took the same position as the advisory council, opposing the views of the university.

In court​ the university first argued that the trust deed did not give rise to any obligations that were still enforceable, a view the judge rejected. She also rejected the claim that the trust only covered the collection of books and photographs handed over in 1944, deciding instead that it covered not only all additions made since that time by purchase, gift or exchange, but also the institute as a whole, which comprised ‘an independent unit’. This ruling was very welcome, both because of the university’s efforts to separate the management of the library from that of the institute, but also because in 2013 the university had wanted to sell a set of the four earliest folios of Shakespeare which had been given to the University Library in the 1950s on the condition that they could never be sold. The judge also found that the university was obliged to fund the institute, and could not levy a service charge on it (the space charge). She did not accept the advisory council’s contention that the trust included the building in which the institute is housed, but underlined the obligation in the trust deed that the university must house the institute in a suitable building in Bloomsbury. Nor did she accept the contention of the advisory council that the institute should be controlled by a committee of management. In our view, such a committee would have provided a good way of reducing conflicts of interest in the administration of the trust. The judge was evidently conscious of such potential conflicts, and in the last part of her judgment found that the university was not permitted to seek funding for the Warburg Institute in conjunction with other institutes and then distribute the funds between them as it saw fit, as is currently the case with the Hefce funding to SAS.

The effect of this judgment has been to establish that the university has been in serious breach of the trust deed for many years. The Warburg Institute must now be adequately funded by the university, which will need to rethink its practice of space-charging, at least with respect to the institute. Most serious of all, the university can no longer fund the institute through the block grant of special funding paid by Hefce to SAS, as at present.

The university’s reaction has been curiously mixed. It immediately issued a press release under the heading ‘University of London Pleased with Court’s Decision Regarding the Warburg Institute’, in which it declared that the judge had found in its favour ‘on almost every point that was of importance to us’. Yet at the same time its lawyer sought and obtained permission to appeal. An appeal can only mean more expense (again, in the university’s case, using funds which could otherwise be spent on education and research), more uncertainty, which is harmful for everyone involved, and more reputational damage for the university itself. The Warburg’s advisory council stands ready to start positive discussions with the university about how to put into effect the judge’s rulings; the university’s willingness to enter into such a constructive process remains, at the time of writing, unclear.

The University of London itself is now a very strange institution, consisting of what was left after the colleges became largely autonomous. It is gradually divesting itself of its various components, such as the Marine Biological Station at Millport. Starting from scratch, no one would choose to create an organisation of this kind, and there is no realistic prospect that it will grow by acquiring significant additional numbers of students or academics. Despite being called a university, the role of disinterested (or even interested) academics in its decision-making process, which was crucial when the decision was taken to incorporate the Warburg in 1944, is now very minor, and continues to decline. The main rationale for the creation of SAS was as a mechanism for obtaining a block grant from Hefce, which was apparently unwilling to fund the institutes separately. But the institutes were very different from one another. Some had major research collections, some had substantial numbers of students, and some, notably two established after 2003 without encouragement from Hefce, had neither. There was no compelling academic reason to combine the institutes, and their responsiveness to their original constituents, guaranteed by their committees of management and by the control of their own libraries, has been gradually reduced. Now the role of the advisory councils even in the appointment of institute directors, which was previously decisive, has been virtually eliminated.

On a more positive note, the four most important SAS libraries, including the Warburg, have recently been restored to essentially the position they enjoyed before the adoption of the policy of library convergence. Much time and goodwill has been lost by the university in upholding for years a policy that was ill conceived from the start. It is to be hoped that its belated recognition of this will lead to a mutually acceptable understanding with the Warburg, especially as those responsible for creating the problems in the first place have now all left the scene.