1. History and Hysteria Carol Tenopir ctenopir@utk.edu University of Tennessee & Donald W. King dwking@umich.edu Society of Scholarly Publishing Annual Meeting June 1, 2000

2. Towards Electronic Journals: Bytes Out of Myths and Bits of Reality

5. Growth of Scholarly Journals

6. Growth of Internet Domains Source: Internet Software Consortium Domain Survey available at

7. Carol Tenopir and Donald W. King. Towards Electronic Journal: Realities for Scientists, Librarians, and Publishers. Washington, D.C.: Special Libraries Association, 2000.

8. Questions 1Are scholarly journals worth saving? 2What are the price and demand relationships? 3Why have journal prices spiraled upward? 4Where do we go from here?

9. Trends in the Use, Usefulness, and Value of Scholarly Journals

10. Average Number of Scholarly Article Readings Per Year

11. Time Spent Reading

12. Facts Behind the Myths Growth of journal literature is correlated with the number of scientists 70% of all readings are done by non- academicians

13. Why these myths? 1Citation counts do not measure all readings 2The data from some studies done in the 1960s and 1970s was misinterpreted

14. Estimated Number of Readings The extrapolated estimate is about 520 readings per article In reality the number is closer to 860 A current estimate is about 900 readings per article

15. Amount of Journal Readings Scientists read from an average of 18 journals each year Only one of 18 have over 25 readings Half are read less than five times Increasingly users are relying on a variety of sources for information

16. Growth of... Scholarly Journals Internet Domains

17. WWW Impact PubMed searches reached up to 400,000 per day in 1998 A month worth of searches in PubMed equaled a year of MEDLINE searches (about 7.6 million)

18. University Scientists’ Use Electronic journal use depends on the field of science Studies show about 50% of faculty prefer electronic journals

19. Usefulness & Value of Scholarly Articles Information serves many purposes Highly important to these purposes Readers are willing to pay a high price for the information in their time The information results in improved performance

20. Scholarly Journals Examined from a Systems Perspective Several 1970s studies for NSF Identified/characterized functions, participants, input resources & outputs of hundreds of activities Assessed current & future effects of technologies & other resources

21. Total Cost (excluding $’s exchanged) 1977 $16 billion (1998 $) 1998 $45 billion

22. Average System Costs $5900 $7200 $65 $60 Per ScientistPer Reading 1977 1998

23. Trends on System Costs Scientists’ costs are up Library resources costs are down Publishing resource costs are down

24. The Question!!! Why have average prices risen by a factor of nearly 10 times over a period of time in which the relative cost of publishing has actually decreased?

25. To understand price one must understand publishing costs Five publishing functions: –Article processing (= $190,000) –Non-article processing (= $19,500) –Reproduction (= $101,000) –Distribution (= $80,500) –Support (= $168,500) –Total (= $559,500)

26. Average Cost per Subscription

27. What do average prices mean? Price per journal Price per subscription Price per article Price per page

28. Cost per Subscriber Quartile Circulation Range Ave Circ w/in Range Cost per Subscriber 0-25%150-900520$747 26-50%901-19001,310$316 51-75%1,901-5,7003,290$145 76-100%5,701+18,100$53

29. Average Annual Price Increase (%) in Scientific Journals

30. Causes & Consequences of Spiraling Prices Inflation & increased size Triggers in the 1970s Personal & library subscription elasticities High fixed costs Readers, libraries & publishers all lose Yet, journal system costs have not changed

31. Why have journal prices spiraled upward? Size and Inflation—56% Drop in personal subscriptions Addition of new, low-circulation journals— 17% McCabe thesis High profit/net revenue

32. Costs of Low-Circulation Journals 2,500 – 2,400 2,000 – 1,900 1,500 – 1,400 1,000 – 900 500 – 400 $6 8 18 41 186 DropRequired Cost

33. Average Number of Personal Subscriptions to Scholarly Journals

34. Proportion of Readings of Scholarly Scientific Articles

35. What factors affect demand? Price Journal attributes Availability & relative cost of alternatives Combinations of distribution means and media are finding a niche

36. What are we really buying?

37. Two components of costs/price Article processing Distribution/access

38. Article processing Wide range of quoted costs Costs similar for paper & electronic versions Range of cost a moot point

39. Distribution/Access Electronic distribution.access costs negligible Paper subscription ($25-$35 per subscription) Paper separate copy ($15-$30 per item) Paper subscription costs per reading is low for frequently read journals Paper versions may cost less for some journals when scientist costs are included

40. Some merit in considering alternative sources of revenue to recover article processing costs

41. Number of Separate Copies of Articles Received by Scientists 19771993-1998 ILL/Document Delivery 4 million>40 million Other39 million>60 million 43 million>100 million

42. Some alternative pricing policies Site licenses Differential pricing Unit pricing No magic bullet

43. Where Do We Go From Here? New and specialized journals will be electronic Journal availability in print and electronic Impact of full-text databases Emphasis on accessibility of information Time is valuable