table of contents previous chapter next chapter
Regardless of the quality and robustness of the digital resources created, they will not last long, let alone grow and survive in the long term, if the projects that produce them cannot be sustained. Sustainability thus refers to all of the considerations that go into maintaining the institutional context for the creation and maintenance of digital objects and resources, and supporting their long-term viability. Preservation—the maintenance of the digital objects themselves—is a distinctly different but related topic, dealt with in Section XIV. Digitization of cultural objects is still relatively new; the majority of projects and programs interviewed are less than ten years old. We are not yet able to take a long-term view of the life cycle of a digitization project, hardly even a medium-term view. In most cases, energies are naturally channeled towards start-up and building a critical mass. Although there was some recognition among projects and programs that a strategy for sustainability was desirable, it is probably true to say that sustainability in the long term is still generally pushed down the list of priorities by more immediate and pressing concerns. Readers should also consult Section II on project planning, which puts these issues into the context of the larger planning process.
Few projects have identified a point when, with their work complete, they will close down and pass the content they have created in digital form over to others for long-term maintenance and delivery. And yet for many projects, whether they plan for it or not, this is the likely outcome, either because of loss of funding or simply because the project has finished its work and has no further reason to continue. It is worth considering at the outset whether the digitization work you are undertaking has a finite scope, or is likely to continue indefinitely. Projects that are defined by the scope of a particular collection or author should plan for an appropriate exit strategy when their work is complete, including plans for ongoing delivery of the digital resource created.
The Survivors of the Shoah Visual History Foundation, for instance, advises that project managers plan out the whole system before any digitization takes place, envision the entire process and life-cycle of the project, and plan for it up to and including preservation of the materials. This plan should ideally include contingencies for re-locating the digital objects if, for example, funding declines to a point where it is no longer viable to maintain the materials.
Projects whose mandate is broader will be limited (or not) by other factors, such as institutional priorities and funding. For some institutions, such as national libraries, it is entirely appropriate that they should not have an exit strategy: the mission and funding of such institutions is consistent with long-term support and maintenance of the digital materials they create. But it is not only the national libraries that have such far-ranging plans: some large academic institutions are also making a commitment to keep digital materials over the very long term. The Robertson Media Center at the Clemons Library, and the Special Collections of the Alderman Memorial Library, both at the University of Virginia, intend to keep their digital deliverables viable and available indefinitely. The University of Virginia Library as a whole is launching a Central Digital Repository that will, in theory, become the sustainable home for projects currently delivered via one-off systems at the Robertson Media Center and Special Collections. The Harvard University Library Digital Initiative (LDI) adds another aspect to the strategies mentioned above. Repository policy strongly encourages deposit of at least one version of the digital object in an approved preservation format to ensure that new deliverables can be generated affordably in the future.
Most projects reported that loss of the digital materials would matter greatly—institutionally and to their user communities—but few could be confident of the continued stream of funding and other support that would be necessary to maintain their digital assets. There are high costs associated with ensuring long-term access, for example: regularly updating the interface, maintaining backup copies, migrating the data to new platforms and formats, and supporting the use of the resource, let alone providing for its continued incremental development over time.
Programs and projects need to think more strategically, and need to adopt a longer-term view by planning out the development of sustainability plans at the outset of their initiative. Funding agencies have a role to play in insisting that provision for long-term sustainability be factored into grants They also have a wider responsibility in recognizing the cultural value of digital objects and contributing towards their sustainability. However, long-term sustainability is difficult to achieve on grant funding alone, and a reliance on “soft” money puts a project on a very insecure footing: not only because of the uncertainty of funding from year to year, but also because so much of the project’s resources must go towards the ongoing search for support. If funders value the digital cultural objects they can help maintain their longevity by encouraging the establishment of longer-term, strategic funding plans. Needless to say, the prevailing economic and political climate is relevant and not necessarily conducive to a strategic, long-term view.
In this section we will look first at the funding issues in more detail, and then discuss planning strategies and methods of sustaining resources through use.
The Library of Congress’s National Digital Library Program (NDLP) intends to keep its digital deliverables available for as long as possible. The adoption of an exit strategy, in case the Library found itself unable to meet the costs of sustaining the NDLP’s digital assets was considered unthinkable. On the contrary, the Library recognizes that it is more likely to provide an exit strategy for other institutions that find themselves unable to sustain the digital resources they have created.
For the Library of Congress, the sustainability of its digital resources would not depend upon their generating income streams to make them self-sufficient. In our interviews, the Library reported that it had already obtained some funding towards providing a level of sustainability for its digital assets and is working to make these resources visible to national audiences, thereby building a large group of invested stakeholders.
Similarly, the National Library of Norway’s overall statement of intent, or mission statement, refers to the need to establish a framework for digital libraries and collections that make information available to everybody, wherever they are, in a democratic way. The emphasis is on access and inclusion. They intend to establish a long-term digital library with core functionality, and then provide different “windows” onto this information. As with the Library of Congress, there is no exit strategy — on the contrary, the digital activities at the Library aim at building a sustainable digital library, with sustained and secure funding.
Museums face the same challenges in finding opportunities to fund digital collections project–by project and collection–by collection. While the Berkeley Art Museum/Pacific Film Archives builds its digital projects with sustainability and preservation in mind, it must fund the creation of new digital collections in exactly this way. Early digitization efforts at the Peabody Museum of Archaeology and Ethnology at Harvard University also followed this methodology.
Many projects, like the Cornell Institute for Digital Collections (CIDC) at Cornell University have benefited from very generous lead-gift funding and continue to rely on soft money. However, the use of grants and gifts to fund digitization initiatives means that the prioritization and selection of material for digitization may be driven by an assessment of those areas most likely to receive funding at any given time (through their timeliness or appeal to a particular funder), rather than by an overall digitization strategy. These projects recognize the need for a stable funding environment.
Similarly, the Virginia Center for Digital History (VCDH) at the Alderman Memorial Library at the University of Virginia (UVA) has benefited from funding from a wide variety of sources including UVA, National Endowment for the Humanities (NEH) and private donors. The main problems with the funding have been uneven distribution and inconsistency. With the benefit of hindsight, the Center felt it would have been better placed if it had had solid base funding with project funding added on top.
There are three primary ways to secure long-term funding for digital assets: commitment by the originating institution to meeting the costs associated with the digital asset, endowment funding, and making the resource self-sustaining. Of these, endowment funding is the most difficult to obtain and none of the projects interviewed for this Guide described a successful approach of this sort.
Several projects, however, have set in place a funding model that relies on generating ongoing support from the resource itself. The Brown University Women Writers project is dependent upon self-generating funds; it has looked for endowment funding but now hopes that the resource itself will generate sufficient income to meet its long-term costs. The sustainability of the Colorado Digitization Project also depends on self-generating funds and it is considering establishing a membership-based organization to generate revenue in a coherent and consistent way. In making its resource self-sustaining, JSTOR has established a basic fee structure for participating institutions: a one-time Archive Capital Fee (underwriting digitization costs and an archives reserve fund) and an Annual Access Fee (supporting ongoing access to the database and the addition of new volumes to the database). JSTOR director, Kevin Guthrie, comments that nonprofits should develop revenue sources that “match the nature of the mission-based uses of their funds.” As JSTOR’s work of archiving journals continues, so stable and recurring funding is essential. JSTOR is thus building a reserve fund, the proceeds from which will help fund the ongoing cost of archiving.[1]
Some plans for self-maintenance face greater challenges. The Thesaurus Musicarum Latinarum (TML) project is dependent for long-term sustainability on self-generating funds, but the TML has not yet secured these resources. Should resources prove insufficient, the exit strategy would be to break up the TML and the larger consortium group, and move them elsewhere. Another possibility would be to publish the TML from a private server. The TML experience suggests that projects need to recognize that even institutional support is tenuous, that digital data are easily destroyed, and that some projects will not survive.
Cooperation and joint initiatives with other types of institutions or agencies may also provide a route towards longer-term funding. Where several institutions combine to support the creation of a resource, the burden for each may be quite manageable. Similarly, funders may look more favorably on joint efforts that show a broad institutional commitment to the resource. The CIDC at Cornell, for example, has cooperated with libraries, museums, and archives, at all levels from local to international. While these are useful processes, cooperation is not without problems and firm rules have to be agreed to produce a successful project. However, cooperative agreements of this sort can provide a route to funds that are otherwise inaccessible. A more detailed discussion of inter-project collaboration and its benefits and risks can be found in Section IX.
The Digital Imaging and Media Technology Initiative (DIMTI) at the University of Illinois is an excellent example of “joined up thinking” in relation to digitization. The benefits of establishing a body such as DIMTI at the outset of digitization are demonstrated in the strategic overview that it enjoys. The fact that it is a single body with responsibility for dealing with generic issues from the outset has advantages in creating a unified institutional vision. Furthermore, DIMTI demonstrates that such a body does not need to be large when supported by other appropriate administrative or managerial bodies. The fact that the University of Illinois Library has a relatively flat structure seems to have been beneficial in this case.
Although various forms of base and soft money remain the predominant forms of sustainability, the projects surveyed for this Guide tended to be large and relatively well established. Therefore, they may be generally less concerned about their ability to secure future funding than smaller and more recent projects. However, some projects have developed alternative models for sustainability.
The licensing model developed by the JSTOR project secures long-term funding by creating durable relationships with other stake-holders: most significantly, the collection’s current and future users. In this model, access to JSTOR collections for UK higher education institutions is licensed via the UK Higher Education Funding Council’s Joint Information Systems Committee (JISC). The JISC succinctly summarizes the benefits to both parties as follows: “the agreement between JSTOR and the UK Funding Councils (acting through the JISC) is a twenty-five year Agreement commencing on 1st August 2001 and terminating on 31st July 2026. The length of this agreement is to reflect the on-going partnership between JSTOR and the JISC, to allow institutions to feel confident in long term access to JSTOR collections, and to underscore the value this resource brings to the UK academic community.” Such sub-licensing agreements are effective methods for projects that have a large and identifiable institutional audience to secure funding and sustain the project through long-term use.
The Colorado Digitization Project’s collaborative, not-for-profit model is a second alternative. The project has a broad range of funders (nine in total) and participation is open to libraries, library systems, archival organizations, selected pilot schools, historical societies, and museums. The involvement of more than 50 collaborative projects and four regional scanning centers, and the emphasis on sharing and transferring expertise means that sustainability does not depend on a single funder or institution but is shared by the collaborative partners.
Neither of these strategies is guaranteed to ensure sustainability. Licenses may not be renewed or obtained in sufficient numbers and contracts for large consortia, particularly regarding ownership of material, are difficult to achieve. Nevertheless, the demand for grant-based funding will always outstrip its supply and such models indicate alternatives that can go some way to achieving sustainability.
Focusing resources on a clearly identified target audience has been shown to be helpful in sustaining projects, as that audience has an important stake in the future sustainability of the program. The UVA Electronic Text Center reported that the most effective dissemination strategy for building a sustainable resource was predicated on the Center’s local focus on research, teaching and learning, and the grounding effect this produces. Similarly, the William Blake Archive (also at UVA) has produced an explicit statement of intent that covers its rationale, scope, significance, primary audience, long-term sustainability, level of faithfulness to the originals and suitability for different target audiences. This level of focus has enabled it to sustain a high degree of momentum and hence of attractiveness to its audience, which in turn strengthens its funding position.
Ensuring internal and external use of a resource can contribute to its sustainability. The Walker Art Center (WAC) and Minneapolis Institute of Art (MIA) both intend to ensure that the digital process becomes part of their institutional culture, ensuring that all types of user use the resources from day to day. They will educate staff to identify and use the objects in everyday work, and if they are successful, this integration of digital materials into the mainstream of users’ processes will help ensure the sustainability of the digital objects. Similarly, DIMTI is an example of a program that has set itself the goal of integrating digitization into existing Library activities and structures; this is a far-sighted and effective mechanism to garner institutional support and maintain digitization activities in the long term. The sustainability of digital resources will be increasingly assured if academics are successfully encouraged to use them.
Key elements in establishing sustainability:
The Genealogical Society of Utah, Family History Department reported that it is not only technology that ensures the longevity of materials, but also migration of hardware, software, media and, importantly, people skills. The GSU is constantly reviewing its management procedures to ensure that no information can become obsolete due to hardware, software, media or skills being lost. (See the discussion of skills in Section II, Project Planning.) Nearly all projects will eventually need to migrate their digital resources to new environments and will require funding to do so, but these costs are rarely built into the project’s base funding.
Cultural heritage institutions face a number of challenges when placing digital collections online. Not least of these is the problem of how collections can be built and sustained in harmony with the sector's non-profit culture. In February 2001 the Council on Library and Information Resources (CLIR) and the National Initiative for a Networked Cultural Heritage (NINCH), supported by the Institute of Museum and Library Services, convened a meeting to discuss what sustainability models were available to the sector. The meeting included representatives from museums and libraries, legal experts, technologists, funders and web enterprises.
In discussing business models the group concluded that whether the digital collection is for profit or not, large amounts of capital, new skills and new organizational cultures are required. Furthermore, non-profit organizations must adopt the same rigor for accountability, efficiency and value for money as commercial enterprises. The latter may have a competitive advantage in their ability to identify, test and target audiences, but any successful project must have clearly identified its purpose, mission and audience. Nevertheless, non-profit organizations generally have a longer period of time to realize their goals, and the expectations of trustees and governors can differ considerably from that of shareholders or investors. For the report, Building and Sustaining Digital Collections: Models for Libraries and Museums, see http://www.clir.org/pubs/reports/pub100/pub100.pdf
It is becoming apparent that web access to digital collections is not just broadening access but providing access in new ways. This presents a challenging dichotomy for non-profit organizations. How can the convenient and unrestricted access to digital collections that users expect be reconciled with the need to recover costs? Moreover, the resources, outlook and skills required to build a digital collection are somewhat different from those required to distribute it. The non-profit organization may find that the latter are far harder to integrate into their larger mission.
Cultural heritage organizations could license their digital content to a commercial distributor, but it is by no means clear that the analog/digital divide is the point where non-profit organizations cease to owe users free access. Even though access to analog collections has never been entirely free, it is still uncertain at what point—and to what extent—cultural organizations should start using web delivery to load costs onto the user.
An alternative model is strategic collaboration between non-profit organizations. (One such model examined at the CLIR/NINCH meeting was between the George Eastman House, Rochester, New York, and the International Center of Photography in Manhattan. See page 7 of the report cited above). The costs are not necessarily going to be lessened through this approach but it does provide a means of spreading and managing the risk involved, building economies of scale and meeting the need for standards and practices for interoperability and access. However, as noted above, even between large organizations such collaboration is no guarantee of a successful return. Considering the infrastructure and skills required for creating and delivering digital content, small and medium sized libraries and museums are likely to find that collaboration with a larger partner is the only method available to them.
Despite the risks and costs involved, participants at the meeting voiced the opinion that not to enter the digital arena at all was itself a greater risk. At least the risk could be shared by working collaboratively. Furthermore, the burden of expectation might be diminished if funders viewed as experimental some ventures where the return might be non-existent or elusive.
The elements of sustainability identified at the meeting require further work by the sector. Participants emphasized the need for standards and best practices, coherent and common digital architectures, and a means of creating and sharing knowledge. In order to deliver these desiderata, the group suggested a number of ways forward, including guides to best practice, a clearinghouse for technical information, a registry of digital conversion projects, digital service bureaus, regional centers for digital content production, and “tool kits” for data and communication design and development.
[1] “Developing a Digital Preservation Strategy for JSTOR, an interview with Kevin Guthrie,” RLG DigiNews (August 15, 2000) http://www.rlg.org/preserv/diginews/diginews4-4.html#feature1. See also the section “Who pays what,” in Dale Flecker, “Preserving E-Journals,” D-Lib Magazine (September 2001) http://www.dlib.org/dlib/september01/flecker/09flecker.html