1. Identifying and Securing the Requisite Resources Anne R. Kenney, Cornell University USA Business Models related to Digital Preservation, Amsterdam, 20-22 Sept 2004

2. How do we identify and secure the requisite resources to maintain a digital repository?

3. “If we had some horses, we’d have a cavalry – if we had some men." Victor Mature in Demetrius and The Gladiators, 1954

4. Step 1. Identify Cost Categories Cost categories –Startup costs –Ongoing costs –Contingency costs

5. Step 2. Identify Cost Centers Capital costs –Technical/procedural infrastructure Direct operating costs –By functions, e.g., identification, acquisition, ingest, metadata Overhead –Indirect operating expenses, e.g., indirect staff, facilities, general/administrative/support services

6. Step 3: Calculate Costs Case Study: ArXiv automated electronic archive and distribution server for research papers in physics, cs, math, etc. Begun in 1991 281,00 item database in FY2003/4 40,000 submissions that year Cost/submission: $1-$5

7. Calculate Capital Equipment Costs Annual equipment cost = –purchase price of hw/sw –amortization rate (3-5 year range) –annual maintenance,licenses, and development fees, estimate at 20%-50% purchase price

8. Capital Equipment: ArXiv Server Equipment cost: $10,600 Amortization: 5 years Annual fees: $4,713 Annual equipment cost = $6,833($10,600/5 + $4,713)

9. Equipment costs/ preserved digital object Calculated by dividing the annual equipment costs by the current and estimated growth rate of digital objects The greater the number of similar digital objects, the lower the per/object equipment cost.

10. ArXiv Annual Equipment Costs 281,000 submissions 40,000 in FY 2003/4 $6,833/40,000 = $.17/submission $6,833/281,000 = $.02/item/year

11. Calculate Personnel Costs Staff performing specific tasks, management overhead, ancillary expenses Calculate staff costs on “weighted hourly rate” not salary, which is over twice the hourly wage

12. Weighted Hourly Rate 1)Avg. number of workdays/ year 2)Number of hours/day 3)Number of productive hours 4)Weighted hourly rate = salary + fringe productive hours

13. Calculating The “Real” Costs at Cornell Staff: –222 day work year (excluding vacation, sick, personal and holidays) –7.3 hour work day (excluding breaks) –1,621 hours/year at work –Assume 75% “production time”: 5.5 hours/day and 1,216 hours/year –“Weighted hourly rate” = salary + fringe 1,216

14. ArXiv: Programmer/ Analyst Band F Annual hours worked: 1,216 Beginning salary: $35,000 Benefit rate: 31% Hourly wage: $18 “Weighted hourly rate”: $35,000 x 1.31 = $37.71 1,216 Weighted rate is 2.1 times the hourly wage.

15. Management and Ancillary Costs % of management overhead assigned/staff member Ancillary expenses (min: $2,00/year) –training, travel, workstation support, supplies, phone/data, reference materials Overhead

16. Example: Full Freight Programmer/Analyst $37.71 (WHR) + $ 6.85 (10% mgr overhead) + $ 1.16 (ancillary expenses) $45.72 X $ 1.57 (overhead) $71.78

17. Calculate Services Costs Annual or contractual fees paid for out-sourced functions, including backup, space in server farm, mirroring Related services, e.g., technology monitoring, subscription to software repository, consortial fees, royalties

18. Step 3. Calculate Costs: Contingency Unanticipated expenses associated with ramp-up, training, trouble shooting Will vary from institution to institution, depending on complexity, experience, size of effort

19. Step 3. Calculate Costs: Overhead Negotiated rate vs. itemized expenses How much comes back to the digital repository Recovered overheads vs. sunk costs vs. hidden costs

20. ArXiv Annual Costs for Maintenance/Preservation Equipment: $6,833 Staff (2 FTE): $91,700 Mgt overhead: $9,170 (10%) Ancillary staff costs: $4,200 Contingency: $11,190 Overhead (57%): $70,163 Total: $193,256/year

21. Total Costs 281,000 submissions 40,000 in FY 2003/4 $193,256 /40,000 = $ $4.83/submission $193,256 /281,000 = $.69/item/year

22. Maintaining/Preserving ArXiv Staff costs represent the greatest expense Present day costs do not equal future costs Total costs will decrease only with automation of human effort Per submission and per item costs will continue to decrease as the collection grows

23. Step 4. Secure Resources Additional funding Reducing expenses Recovering costs

24. Additional Funding One-time funding Endowments On-going institutional support Public or government support Assume responsibility by broad coalition of stakeholders

25. Preservation Pledge Drives

26. Reducing Expenses Streamlining operations Automating processes Eliminating redundancy of effort and content Spreading cost over consortium

27. Recovering Costs Partnership with content providers Partnership fees (harness the free riders) User fees Third party use/mining of materials

28. Funding the Effort: Lessons Learned Digital preservation as a common good is not economically viable Sustainable programs may not be possible at the institutional level On-going dedicated funds essential Content creators, users, and cultural repositories are not prepared to fund the full enterprise

29. DP Programs Will Be More Sustainable If… Digital materials are viewed as both institutional assets and public goods Institutions reach an advanced stage of digital preservation commitment The focus is on cost benefit rather than merely cost Preservation costs are directly tied to access

30. DP Programs Will Be More Sustainable If… Digital resources can substitute for traditional services, publishing efforts, and collections and/or… Stakeholders embrace digital content and are willing to pay for some added value, convenience, or service Institutions share rewards and responsibilities

31. “Stewardship is easy and inexpensive to claim; it is expensive and difficult to honor, and perhaps it will prove to be all too easy to later abdicate.” Cliff Lynch, 2003

32. “They say you can’t do it, but remember, that doesn’t always work.” Casey Stengel