The cost of continuity

1
The cost of continuity

• Alan Howell
• State Library of Victoria

2
Overview

• Background
• Six key issues
• Shortage of relevant data
• Costs higher than for traditional resources
• Costs occur throughout the life-cycle
• Short-termism
• Long-term value is difficult to determine
• Uncertainties about who pays
• Summary of findings

3
Background

• Context
• Modes of digital death
• Definition
• Digital preservation cost paradox

4
1.1 Context

• Over 93 of new information is created in digital
  form, or born digital (Lyman, 2000)
• Digital resources are
• inherently fragile, easily corrupted or altered
• require access technologies that are changing at
  an ever increasing pace
• worth preserving (as much as traditional
  resources)
• Time-frame for preservation is reducing

5
1.2 Modes of digital death

• Hanna (2001) and Wiggins (2001) propose how to
  lose resources. They advise
• Reorganise the content (404 death)
• Encourage the sponsor to fear history
• Stick with outdated media, hardware, systems
• Lose your software (and the manuals)
• Lose critical metadata (indexes, dependencies)
• Fail to protect authenticity
• Ignore the cost of preservation

6
1.3 Definition / scope

• Costs of digital preservation costs
• Costs includes all costs over full life-cycle
  of resources
• For a large or popular collection, electricity
  begins to become a factor

7
1.4 Digital preservation cost paradox

• The overarching paradox of digital preservation
  is that people think of digital content as
  inexpensive and inexpensive things are not
  worth preserving
• (Brand, 1999)

8
2. Six key issues

• Shortage of relevant data
• Costs higher than for traditional resources
• Costs occur throughout life-cycle
• Pervasive short-termism
• Long-term value difficult to determine
• Uncertainties over who pays

9
2.1 Shortage of relevant data

• Difficult aspect on which to gather information,
  due to
• in many cases, respondents indicated that they
  just didnt know how much the archive was costing
  or would cost in the future (Hodge Carroll,
  2001 Russell Weinberger, 2000)
• Difficulty in extrapolating costs from pilot
  projects (of which there are many) to full-scale
  public services
• Cost of preservation enmeshed in cost of access

10

• Until several large archives have gone through
  at least one or two migrations or emulation
  developments, it will not be possible to separate
  the cost for the archives from the cost of doing
  business
• (Hodge Carroll, 2001)

11
(Feeney, Kelly Weinberger costing models)
12
2.2 Costs higher than for traditional information
resources

• Digital preservation will be more expensive and
  more intensive (library community)
• High cost in minimising data loss inherent in
  digital resources
• Strategies aiming for a loss rate no greater than
  1 in 100 million are being implemented (PRO 1999)
• No equivalent in paper-based preservation!

13

• If Captain
• Cooks Journal lost
• 1 part in 100
• million it would
• lose less than
• 1/1000th of a
• micron from
• each page!

14
2.3 Costs will occur throughout the life-cycle of
digital resources

• Preservation always an ongoing cost
• For traditional resoucess, active preservation
  only takes place when deterioration sets in
• Digital preservation represents a continuum.
  Re-creation (cost) is inherent
• Creation and preservation gap is shrinking
• Creation of a digital object is the true
  starting point for preservation (Kelly
  Weinberger, 2000)

15
2.4 Short-termism

• Exponential growth has emphasised the speed and
  ease of short-term access
• Little regard for the long-term preservation.
  no firm plans for migration, but will plan to
  stay up to date with current technologies by
  migrating the content to each new technology
  (Hodge Carroll, 2001)
• Mandates exist at legal repository level but
  without sufficient funding

16

• The big marketplace problem with preservation is
  that the natural beneficiaries do not seek
  service until many years after the stuff has to
  be saved
• (Gladney, 2001)

17
2.5 Long-term value of digital resources is
difficult to determine

• Selection for preservation is best based on
  significance
• Works for traditional resources (time)
• Cant wait for digital resources to demonstrate
  their worth over time
• Increases the importance of selection guidelines
  (PANDORA)

18
2.6 Uncertainties about who pays

• Creator
• Publisher
• Libraries of publishing organisations
• Data centres
• Third party repositories
• Legal depositories
• Users

19
Who pays? Creator

• Not very involved in the digital archiving
  process (but changing as metadata gains
  importance) (Hodge Carroll 2001)

20
Who pays? Publisher

• Minimum back-up
• As long as theres a business need
• For publishers and produces, the cost of
  archiving is still tied up in the cost of
  manufacturer (Hodge Carroll 2001)

21
Who pays? Libraries / archives of publishing
organisations

• Publishers, data producers, societies, cultural
  organisations, government agencies, academic
  institutions, universities, industries
• In organisations such as pharmaceutical,
  chemical and petroleum companies, critical to
  the perpetuation of research and development
  (Hodge Carroll 2001)
• Part of mission, ethos, urgent need to create
  and preserve their own archival histories
  (Lyman)

22
Who pays? Data centres

• National Digital Archive for Datasets (UK), NASA
  Distributed Active Archive Centres (USA)
• Large volumes of relatively simple data
• Most mature archival model
• Simplest organisational model (clear IP rights
  owner funder)
• Model allows access by funding agency

23
Who pays? Third party repositories

• Publication Service Providers
• Ingenta, HighWire Press (journal publishing
  industry)
• Primary role vendors secondary archiving
• Economic model not clear (bundled services)
• Repository Management Agents
• JSTOR, OCLC Electronic Journals Project, LOCKSS,
  Internet Archive
• Trusted third-party repositories
• Safety net

24
Who pays? Legal depositories

• National depositories (archives)
• Document the business of government
• Government initiatives in UK, USA and Aus.
• National, state and university libraries with
  legal deposit status
• Will preserve digital resources while mandate to
  maintain the cultural, history and intellectual
  output of the country or state remains
• Instrumental in raising issue
• National Libraries active

25
Who pays? User

• Not very involved in the archiving process
• Endowment model would set aside a portion of the
  payment for the use (whether storage or access)
  of the archive for its perpetual care (Hodge
  Carroll, 2001)

26
3. Summary of findings

• Projects dont scale to services
• Preservation costs enmeshed in access costs
• Costs of preserving complex objects unknown
• Costs start with creation
• Costs are ongoing (throughout life-cycle)
• No equivalent to benign neglect
• Who pays? variety of economic models likely
• Build something in !

27

• The preservation of digital information cannot
  wait for exact information because it may never
  appear
• (Dale Bellinger, 2001)

28
References

• Barry, T. (2001). 'Re DLs in UK (renewed
  objection)'. Email to DIGLIB discussion list
  22/10/01. cited 22 October 2001.
• Beagrie, N. and M. Jones (2000). 'Preservation
  management of digital materials workbook
  (pre-publication draft)'. Online. Available
  from Internet URL http//www.jisc.ac.uk/dner/pre
  servation/workbook/. cited 17 November 2001.
• Brand, S. (1999). "Escaping the digital dark
  age." Library Journal 124(2) 46-48.
• Feeney, M., Ed. (1999). 'The digital culture
  maximising the nation's investment a synthesis
  of JISC/NPO studies on the preservation of
  electronic materials'. London, British Library
  Board, National Preservation Office.

29

• Garrett, J. and D. Walters (1996). 'Preserving
  digital information report of the Task Force on
  Archiving of Digital Information'. Online.
  Available from Internet URL http//www.rlg.org/A
  rchTF/. cited 16 November 2001.
• Gladney, H. (2001). 'Writings on preservation'.
  Email to DIGLIB discussion list 22/10/01.
  cited 22 October 2001.
• Hanna, M. (2001). 'Born digital - live digital?
  Maybe'. Online. Available from Internet URL
  http//www.icsti/forum/38/. cited 4 November
  2001.
• Hodge, G. and B. C. Carroll (2001). 'Digital
  electronic archiving the state of the art and
  the state of the practice'. Online. Available
  from Internet URL http//www.icsti.org/icsti/wha
  ts_new.html. cited 10 November 2001.
• Kahle, B. (2001?). Quoted in Wiggins, Richard
  2001. cited 18 November 2001.

30

• National Library of Australia (2001). 'Preserving
  Access to Digital Information (PADI)'. Online.
  Available from Internet URL http//www.nla.gov.a
  u/padi/. cited 18 November 2001.
• Russell, K. and E. Weinberger (2000). 'Cost
  elements of digital preservation'. Online.
  Available from Internet URL http//www.leeds.ac.
  uk/cedars/documents/CIW01r.html. cited 15
  November 2001.
• Wiggins, R. (2001). 'Digital preservation
  paradox and promise'. Online. Available from
  Internet URL http//www.libraryjournal.com/digit
  al_preservation.asp. cited 18 November 2001.

31
Overview

• Background
• Six key issues
• Shortage of relevant data
• Costs higher than for traditional resources
• Costs occur throughout the life-cycle
• Short-termism
• Long-term value is difficult to determine
• Uncertainties about who pays
• Summary of findings

32
The cost of continuity

• Alan Howell
• State Library of Victoria