1. Digital Preservation and Library Periodicals Expenses: Eileen Fenton and Roger SchonfeldCNI December 9, 2003* Variance between Non-Subscription Costs for Print and Electronic Formats on a Life-Cycle Basis * Data Revised as of December 31, 2003

2. Outline Archiving – Our Definition The Need Components of a Trusted Archive Study: Non-subscription costs for print vs. electronic formats on a life-cycle basis –Methodology –Data –Findings

3. Definition of Archiving Preservation and Access over Time –Time horizon is extremely long –Beyond the period that the materials are economically productive

4. The Need Preservation and Access Must Be Ensured: –The community is increasingly reliant upon e- resources. –With growing frequency, the e-version is considered the copy of record. –If e-resources offer opportunities to generate savings, these will not be fully realized without a trusted archiving solution.

5. Elements of a Trusted Archive 1.Mission –Mission is critical because it drives resource allocation and routine organizational priorities and activities. 2.Business Model –Sustainability is key. –The archive must generate funds adequate to cover the work of the archive from sufficiently diversified sources.

6. Elements of a Trusted Archive 3.Technical Infrastructure –An infrastructure must be developed which supports the key functions of the archive: Ingest Verification Storage Delivery Migration –It must allow for sufficient replication with appropriate geographic spread.

7. Elements of a Trusted Archive 4.Relations with Libraries –The archive must meet the needs of the library community and the scholars they serve. 5.Relations with Content Producers –The archive must secure the rights necessary to the archival task and must arrange for timely, secure deposit of content.

8. Open Issues Progress must be made in several areas simultaneously. –Content –Technical –Business Technical issues have received the greatest level of attention to date.

9. Open Issues Study launched by JSTOR’s Electronic- Archiving Initiative sought to: –Understand the impact of growing reliance on electronic journals on library budgets. –Our question: How do the non-subscription expenditures for print vs. electronic periodicals differ? –Focused exclusively on non-subscription periodicals costs.

10. Non-Subscription Costs These are the periodicals-related costs other than the actual cost of the subscription. The largest of these include: The cost of space; The cost of computer workstations and infrastructure; The cost of binding; and, most importantly, The cost of library staff time.

11. Group Members Eileen Gifford Fenton, Executive Director, JSTOR Electronic-Archiving Initiative. Kevin Guthrie, Chairman, JSTOR. Donald King, Research Professor, University of Pittsburgh. Ann Okerson, Associate University Librarian, Yale University. Roger Schonfeld, Coordinator of Research, Andrew W. Mellon Foundation.

12. Outline Methodology for Data Collection Overview of Data Life-Cycle Approach Life-Cycle Findings Implications for the Future Funded through a grant from The Andrew W. Mellon Foundation

13. Data Collection Overview Staff logged their time spent on different activities Each institution completed a survey 11 academic libraries participated

14. Data Collection: Staff Activities 1. Collections development8. Physical Processing 2. Negotiations and licensing9. Stacks maintenance 3. Subscription processing, routine renewal, and termination 10. Circulation 4. Receipt and check-in11. Reference and Research 5. Routing of Issues and/or Tables of Contents 12. User Instruction 6. Cataloging13. Preservation 7. Linking Services14. Other

15. Data Collection: Excluded Costs Electronic infrastructure. Microform. ILL. We have some evidence on these costs, which suggests that they would not have meaningfully affected the results.

16. Data Collection: Participating Libraries SmallMediumLarge Bryn Mawr College George Mason University Cornell University Franklin & Marshall College Drexel University New York University Suffolk University Western Carolina University University of Pittsburgh Williams College Yale University

17. Data Collection: Collections Covered Bryn Mawr (all libraries) Drexel (all libraries) Franklin & Marshall (all libraries) Western Carolina (all libraries) University of Pittsburgh (all libraries)

18. Data Collection: Collections Covered Cornell (excludes several, including law and medicine) George Mason (excludes law library) NYU (excludes law and medicine) Suffolk (excludes law library) Williams (excludes department libraries) Yale (Sterling collections only)

19. Overview: Annual Non-Subscription Spending on Periodicals ($ millions)

20. Overview: Share of Annual Non- Subscription Budget by Format

21. Life-Cycle Approach: Overview We have used a life-cycle approach for analysis Examined a 25-year period The long-term financial commitment that a library makes when accessioning one year of a typical periodical

22. Life-Cycle Approach: Overview Grouped costs into one-time and recurring components Calculated costs per title Space occupied by print backfiles is charged at high-efficiency off-campus rates. Net present value analysis, 5% discount rate

23. Life-Cycle Findings: Print

24. Life-Cycle Findings: Electronic

25. Life-Cycle Findings: Format Comparison

26. Implications: Cost Reduction for Electronic-Only Collection ($ thousands)

27. Implications: Cost Reduction as % of Total Annual Non-Subscription Costs

28. Implications: Caveats Underestimated print: –STM omitted at several schools –Duplicate subscriptions omitted –Microform omitted –ILL omitted Underestimated electronic: –Electronic infrastructure Net, we believe this has led us to underestimate print costs

29. Implications: What Else Is Missing? Electronic archiving. There is no contribution to archiving of electronic periodicals at any of the participating libraries. These are missing only from the electronic and not from the print format.

30. Conclusions 1.Non-subscription unit costs are lower in the electronic format compared with print. Because: As has been widely foreseen, some manual processes can be eliminated. More electronic costs occur earlier in the life-cycle, while more print costs occur later in the life-cycle.

31. Conclusions 2.Although there are potential total cost reductions, these are less significant than some have expected. Because: At the larger institutions that have the most print subscriptions, long-standing print operations have already achieved economies of scale. Institutions with smaller print holdings tend to have much larger electronic collections.

32. Conclusions 3.The potential cost reductions assume no contributions towards electronic archiving. It is important to note that if we value the long-term accessibility of electronic resources, archiving them will somehow have to be paid for. It remains to be seen to what extent any cost reductions might be re-directed towards other purposes, such as electronic archiving.

33. Conclusions 4.This study has only just been concluded, and we are still considering the implications of our findings. Other conclusions?

34. Eileen Fenton and Roger Schonfeld www.jstor.org