Introduction: definition and scope
In 1998 the Digital Library Federation offered the following definition of ‘digital libraries’:
"Digital libraries are organizations that provide the resources, including the specialized staff, to select, structure, offer intellectual access to, interpret, distribute, preserve the integrity of, and ensure the persistence over time of collections of digital works so that they are readily and economically available for use by a defined community or set of communities".(Council on Library and Information Resources,1998)
This definition emphasises that ‘a digital library, like any library, is more than a mere aggregation of information resources: it is a service which is based on principles of selection, acquisition, access, management and preservation, related to a specific client community’ (Cathro, 1999). Thus any analysis of the economic aspects of digital libraries will need to take account of the costs of all of the above activities, while also examining issues of value and benefit. In the definition above, ‘collections of digital works’ can encompass a wide variety of manifestations, including digitised content created by libraries, museums and archives; digitised newspaper collections; web archives and other ‘born digital’ content; university repositories; and collections of digital research data.
Australian work on economic models
Little research has been undertaken in Australia on economic models for digital libraries. Most Australian commentators on this topic have been content to cite the models and findings from research undertaken outside Australia. For example, in his ‘Sustainability issues paper’ prepared for the Australian Partnership on Sustainable Repositories, Bradley cites the work undertaken by Brian Lavoie and his framework of ‘responsibilities, incentives, and organization’ (Lavoie, 2004). Bradley notes that digital library content requires persistent attention in order to preserve it:
"Benign neglect, with its lower cost profile, will not ensure preservation....The requirements of ongoing sustainability demand at their base a source of reliable funding, necessary to ensure that the constant, albeit potentially low level, support ... can be maintained for as long as it is required’. Such funding, he notes, is ‘not at all typical of university based communities." (Bradley, 2005)
Other Australian commentators have cited economic models developed in the United Kingdom, including the Lifecycle Information for E-Literature (LIFE Project, 2006) and the Digital Preservation Coalition’s Preservation Management of Digital Material Handbook (2006). This Handbook also deals with the business model for the digital library, including the issue of benefit and value. It references approaches, such as the British Library’s use of ‘contingent valuation’, to ‘enumerate the value of collections and services which had previously appeared to be unquantifiable’.
Some Australian commentators have also explored this challenging area. For example, Missingham (2005) has examined the use of ‘contingent valuation’ by the British Library and by some public library services in the United States, and has reviewed some value studies of national bibliographic services in Canada and New Zealand.
An important business case for digital library development is represented in a report prepared for three of Australia’s major cultural institutions: the National Archives, the National Library and the National Film and Sound Archive. In 2006 those three institutions developed a joint proposal for additional funding to enable them to cope with the challenge of collecting, managing and preserving a rapidly increasing volume of digital content, while continuing to fulfil their responsibilities for print and other traditional content.
After this bid failed to attract policy support at its initial attempt, the three agencies engaged one of Australia’s leading economic consulting companies (Access Economics) to conduct a cost-benefit analysis of the funding proposal. Access Economics developed a model which assessed the cost estimates prepared by the agencies against estimated benefits. This benefit model analysed the value of the proposal from the perspective of society as a whole. The consultants considered both an ‘asset approach’ and a ‘user approach’. The asset approach looked at the stated value of traditional-format assets held by the institutions and developed a digital format equivalent value. The user approach looked at the costs and time incurred by users in accessing digital content compared to traditional content. Access Economics commented that both the asset and user approaches were likely to provide a relatively conservative benchmark for value, as neither takes account of any ‘consumer surplus’ (the willingness of some consumers to pay more than the market price).
The consultants commented that ‘in general, we favour measurement under the user approach where there is a discrepancy, as the asset approach requires a conversion factor to equate stated values of traditional format assets to digital format assets (which have not been formally valued), and does not reflect the value of future usage.’ The study examined costs and benefits over the long term, taking the period from 2008 to 2051 as the reference period. For this period, the bottom line was that benefits exceeded costs by a factor of six. The consultants commented: ‘The benefits are significant and in large part accrue to users of the agency’s material. The investments allow more material to be collected or preserved, and that material in digital form is more accessible to potential users than it otherwise would be’ (Access Economics, 2007).
Digital libraries in Australia
Digital libraries in Australia are still at a relatively early stage of development. An overview of digitisation activities in Australia and New Zealand was given by Cathro in 2007 (Cathro, 2007). To date, the major achievements have included:
- digitisation of cultural heritage collections by institutions such as the National Library, National Archives, state libraries, public libraries and the Australian Institute of Aboriginal and Torres Strait Islander Studies;
- harvesting and archiving of the Australian web domain by the National Library and partner institutions;
- establishment of institutional repositories by the university community, usually with the active participation of university libraries;
- a strong record of collaboration, especially in projects funded through the national higher education portfolio.
This overview will examine these four aspects from an economic perspective.
Digitisation, the conversion of analogue information into digital information, is an important mechanism for building digital collections. The ‘analogue information’ may be instantiated in library materials such as books, pictures, maps, music scores, manuscript collections and audio or video files.
The process of digitisation encompasses all of the processes that are necessary to create usable digital files, including scanning or digital photography, metadata capture, creation of master and derivative digital files, and the transfer of these files into a suitable storage environment.
The National Library of Australia has been undertaking a routine collection digitisation programme for several years. This programme encompasses material such as pictures, rare maps, early Australian sheet music and selected manuscript collections. The Library has also been progressively digitising its collection of more than 40,000 hours of sound recordings. The Library has undertaken digitisation in-house, and its digitisation workflow is supported by a locally developed system known as Digital Collections Manager, which forms part of the Library’s wider digital services architecture.
Internal National Library data indicates the following typical digitisation unit costs:
- about A$13 (£6) per picture;
- about A$37 (£17) for a music score containing five pages (as an average);
- about A$35 (£16) for a rare map;
- about A$17 (£8) for a manuscript item containing three pages (as an average).
These costs take account of the employee cost of scanning, quality control, creation and maintenance of metadata, creation of derivatives, uploading to the storage system and preservation treatment, as well as corporate overheads and the amortised cost of equipment.
These costs reflect the National Library’s internal digitisation workflow and its own
quality procedures. The Australian Institute of Aboriginal and Torres Strait Islander Studies (AIATSIS) is conducting a major digitisation project as a result of A$12 million of funding by the Australian government in 2005. The project will enable AIATSIS to digitise up to a third of its film, video, photographic and audio collections. AIATSIS has reported digitisation costs of about A$14 per photograph and about A$25 per map or poster.
Most of Australia’s state and territory libraries are undertaking digitisation programmes. The State Library of South Australia has reported digitisation costs ranging from about A$7 per image for single sheets of size A4 and smaller, ranging up to about A$33 for maps and other single sheets of size A0.
The National Library has also embarked on a project to digitise nine major Australian newspapers, covering out-of-copyright issues from 1803 to 1954. More than three million pages will be digitised. These pages will be made fully searchable and browsable through the use of optical character recognition (OCR) and other conversion services. This is a project with a strong collaborative aspect, with support by the state libraries in selection of newspaper titles and in the development of off-budget funding for digitisation of regional newspapers. The state libraries have supported a model involving the shared use of the National Library’s infrastructure, including the workflow and delivery software, storage infrastructure and database platform. From a national perspective, this collaborative approach will create efficiencies in the digitisation and delivery processes.
Newspaper digitisation typically involves an ‘industrial scale’ process involving millions of pages, and it is common practice for major components of the process to be outsourced to companies which employ staff based in countries with low to moderate employee costs. Depending on the workflow, quality standards and degree of preservation treatment required, the unit cost of a multi-million page newspaper digitisation project could fall in the range £0.75 to £1.50 per page.
Australia was a pioneer in the field of web archiving. Since 1996 the National Library of Australia has been developing and maintaining PANDORA,1 an archive of selected significant Australian websites and web-based online publications. The purpose of PANDORA is to ensure that Australians of the future will be able to access a significant component of today’s Australian web-based information resources (Cathro et al., 2001). PANDORA is now a routine part of the selection, acquisition and collection management processes of the Library. PANDORA is a collaborative activity, as the archive is being built by the state libraries and some other cultural institutions in addition to the National Library. This collaboration involves a shared software and database platform and agreement on non-overlapping collection responsibilities. This collaborative activity improves the productivity of the archiving and content delivery process.
Nevertheless, it is clear that the selective model, as exemplified by PANDORA, has relatively high unit costs. The model involves individual selection of suitable websites, a dialogue with the website publisher to obtain permission to archive and republish, determination of a gathering schedule, cataloguing of the website at the title level, quality assurance checking of the gathered content, and (if necessary) actions to deal with any quality problems.
Phillips has presented the results of a study of acquisition costs for the PANDORA Archive (Phillips, 2005). This study found that, for the National Library of Australia’s workflow, the cost of acquiring each instance of a selectively archived website was A$179 (at current exchange rates (April, 2008), this is equivalent to £81). It should be noted that the selective approach gives the Library a legal basis for opening the PANDORA Archive to public access, and that the creation of title-level metadata can assist in integrating the discovery of websites with the discovery of other collection resources.
Since 2005, the National Library has also been undertaking an annual large-scale harvest of the Australian web domain (Gatenby, 2006). This activity was the result of a long-standing aspiration to complement the selective PANDORA approach with a whole-domain approach. The Library contracted the Internet Archive2 to undertake the whole domain harvest.
In each of the most recent two annual harvests, the Library captured more than 500 million files, comprising over 19 terabytes of data. Each of these harvests represented more than seven times as much data as was gathered during the first nine years of harvesting for the PANDORA Archive.
The cost of the annual Australian Whole-Domain Harvest is about A$260,000: this cost includes the cost of the crawling process, the storage infrastructure and the indexes. As each harvest contains about 1.2 m website hosts, the cost of acquiring each host instance is about A$0.22.
Clearly, then, the cost of selective archiving, using the National Library’s workflow, is several hundred times that of the whole-domain approach. This is to be expected, given the labour intensive processes of selection, cataloguing and quality assurance associated with the selective approach.
In recent years in Australia, as in many other countries, we have seen the rapid development of repositories to assist universities to manage and expose the research outputs of the academic community. Before 2004, a few Australian universities had been using the Southampton E-prints software or DSpace as platforms to build these repositories.
From 2004 a number of projects funded by the National Higher Education Ministry (now called the Department of Innovation, Industry, Science and Research) have stimulated widespread development of repositories (Cathro, 2006). The key projects have been:
- the ARROW Project (Australian Research Repositories Online to the World), led by Monash University;
- the Australian Partnership for Sustainable Repositories (APSR), led by the Australian National University (ANU);
- the MAMS (Meta Access Management System) Project, led by Macquarie University; and