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

Towards Electronic Journals:Bytes Out of Myths and

Bits of Reality

0

1000

2000

3000

4000

5000

6000

7000

0

10,000,000

20,000,000

30,000,000

40,000,000

50,000,000

60,000,000

70,000,000

80,000,000

Growth of Scholarly Journals

0

1000

2000

3000

4000

5000

6000

7000

Growth of Internet Domains

0

10,000,000

20,000,000

30,000,000

40,000,000

50,000,000

60,000,000

70,000,000

80,000,000

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Source: Internet Software Consortium Domain Survey available at <http://www.isc.org/ds/hosts.html>

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

Questions

1 Are scholarly journals worth saving?

2 What are the price and demand relationships?

3 Why have journal prices spiraled upward?

4 Where do we go from here?

Trends in the Use, Usefulness,

and Value of Scholarly Journals

Average Number of Scholarly Article Readings Per Year

0

20

40

60

80

100

120

1977 1978-1983

1984 1985-1989

1990-1993

1994-1998

Years of Observation

Time Spent Reading

0

50

100

150

200

1977 1984 1993/1994

University Non-University

Facts Behind the Myths

• Growth of journal literature is correlated with the number of scientists

• 70% of all readings are done by non-academicians

Why these myths?

1 Citation counts do not measure all readings

2 The data from some studies done in the 1960s and 1970s was misinterpreted

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

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

Growth of...

0

1000

2000

3000

4000

5000

6000

7000

1839

1849

1859

1869

1879

1889

1899

1909

1919

1929

1939

1949

1959

1969

1979

1989

1999

0

10,000,000

20,000,000

30,000,000

40,000,000

50,000,000

60,000,000

70,000,000

80,000,000

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Scholarly Journals

Internet Domains

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)

University Scientists’ Use

• Electronic journal use depends on the field of science

• Studies show about 50% of faculty prefer electronic journals

Usefulness & Value ofScholarly 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

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

Total Cost(excluding $’s exchanged)

• 1977 $16 billion (1998 $)

• 1998 $45 billion

Average System Costs

• $5900• $7200

• $65• $60

Per Scientist Per Reading1977

1998

Trends on System Costs

• Scientists’ costs are up

• Library resources costs are down

• Publishing resource costs are down

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?

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)

Average Cost per Subscription

$0

$100

$200

$300

$400

$500

$600

$700

$800

500 1,000 2,500 5,000 10,000

Number of Subscriptions

What do average prices mean?

• Price per journal

• Price per subscription

• Price per article

• Price per page

Cost per Subscriber

QuartileCirculation

RangeAve Circ

w/in RangeCost per

Subscriber

0-25% 150-900 520 $747

26-50% 901-1900 1,310 $316

51-75% 1,901-5,700 3,290 $145

76-100% 5,701+ 18,100 $53

Average Annual Price Increase (%) in Scientific Journals

0.00%1.00%2.00%3.00%4.00%5.00%6.00%7.00%8.00%9.00%

1960-1975

1967-1986

1972-1988

1975-1995

1991-1995

1995-1998

Time Periods Examined

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

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

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

Drop Required Cost

Average Number of Personal Subscriptions to Scholarly Journals

0

1

2

3

4

5

6

1977 1978-1983

1984 1985-1989

1990-1993

1994-1998

Years of Observation

Proportion of Readings of Scholarly Scientific Articles

0%

10%

20%

30%

40%

50%

60%

70%

Personal Library-Provided Other

1977 1993-1998

What factors affect demand?

• Price

• Journal attributes

• Availability & relative cost of alternatives

• Combinations of distribution means and media are finding a niche

What are we really buying?

Two components of costs/price

• Article processing

• Distribution/access

Article processing

• Wide range of quoted costs

• Costs similar for paper & electronic versions

• Range of cost a moot point

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

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

Number of Separate Copies of Articles Received by Scientists

1977 1993-1998

ILL/Document Delivery

4 million >40 million

Other 39 million >60 million

43 million >100 million

Some alternative pricing policies

• Site licenses

• Differential pricing

• Unit pricing

• No magic bullet

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