Return to Contents

CHAPTER SIX

BOOK AND SERIAL PRICING

It is clear, then, that the rising prices of serial subscriptions and their impact on acquisitions practices are major concerns. Indeed, anyone conversant with recent developments affecting academic libraries will be aware of the so-called serials crisis. The purposes of this chapter are to provide statistical documentation of the magnitude of the problem and then discuss conceptually the interlocking factors that seem responsible for the escalation in prices. While the main emphasis will be on serials, we shall also discuss book prices.

GENERAL TRENDS IN BOOK AND SERIAL PRICING

When we compare increases in the average price of hardcover books and the average price of periodicals subscriptions  (fig. 6.1)[small | large], we find that between 1963 and 1970 the respective price indexes increased at comparable rates.[4] Over these seven years the prices of these types of publications also increased roughly in line with the overall price index for all goods and services (the GNP price deflator).

Beginning in 1970, however, the pattern changed profoundly. While the price of books continued to increase at about the same rate as the GNP deflator until about 1978, the price index for periodicals began to increase much more rapidly and to diverge sharply from both the index for books and the overall index. This was precisely the decade when a great many new journals were founded, and there are reasons to believe that the proliferation of specialized journals had a marked effect on the prices of periodicals (see discussion later in this chapter).

Over the entire time period from 1963 through 1990, the average price of periodicals increased at an average annual rate more than one and one-half times that of hardbound books---11.3 percent per year versus 7.2 percent per year. Moreover, the prices of both hardbound books and serials increased more rapidly than the general price level (which increased at an average annual rate of 6.1 percent). In the single decade of the 1980s, the GNP deflator increased by a factor of 1.6 while the average prices of books and periodicals increased by factors of 1.9 and 2.8, respectively; the corresponding average annual rates of increase are 3.9 percent, 6.3 percent, and 10.4 percent.

It must be emphasized that the price data included in the periodicals index documents increases in the costs of periodicals published only in the United States, and much of the concern expressed over the rapidly increasing cost of materials has been directed specifically at foreign materials and the pricing practices of foreign publishers.[5] While we cannot provide a reliable estimate of magnitudes, there is no question but that a broader index, which included foreign as well as U.S. publications, would show an even steeper rate of increase.[6] Furthermore, as we noted in the last chapter, to the extent that foreign periodicals are purchased in the country of origin, the dramatic decline since 1985 in the value of the dollar vis-a-vis foreign currencies has exacerbated the effects of rising prices on the acquisitions budgets of libraries.

Book Prices By Field

Trends in book prices take on more meaning when we look at the actual dollar prices of the "average" book in specific fields. (See table 6.1 [small | large] where fields are ranked by the average price of books in 1990.) Comparisons spanning several decades fail to highlight what has happened most recently, as the data on this table illustrate so well. Between 1980 and 1986 book prices in only four fields increased at a rate greater than inflation, and one of these fields, education, was in the lower half of the range of prices. What is more significant is that between 1986 and 1990 book prices in all fields but one increased at a rate greater than inflation.[7] Books published in only six of the sixteen fields included in table 6.1 had average price increases of less than 30 percent over this four-year interval; five had increases of more than 40 percent, and two more had increases of 39 percent.

It should also be noted that the two fields with the greatest percentage increases in price (technology and medicine) were among the top three fields with respect to the most expensive books. It appears, then, that price increases have been more significant since the mid-1980s and that the most expensive books are also those that have been increasing most rapidly in price. This evidence suggests that book prices are now showing some of the tendencies characteristic of serials---not an encouraging sign for those who must be concerned about library budgets.

Serials Prices by Field

While price increases of science and technology journals clearly head the list, prices in all fields are seen to have increased since 1970 at a rate significantly greater than inflation.[8] We can illustrate the magnitude of what has transpired more clearly by taking a "snapshot" look at prices in actual dollars for subscriptions in specific fields in specific years (1963, 1970, 1982, and 1990).

In almost every field, by far the largest relative price increases occurred during the twelve-year period between 1970 and 1982 (table 6.2[small | large] and fig. 6.4[small | large] and fig. 6.5[small | large]). [9] In the fields of chemistry-physics, the "typical" U.S. journal cost $33 in 1970 and $178 in 1982; in engineering, the corresponding dollar figures are $12 and $62; in mathematics and other sciences, $18 and $88. Medicine is the one professional field with a similar experience. In these groups of fields it was normal for serials prices to rise by factors of five-plus over this period (approximately 15 percent per year, on average). In the social sciences, absolute levels of prices are lower and rates of increase, while still rapid, were not quite as great as in the sciences---for example, the typical journal in political science increased in price from $7 in 1970 to $26 in 1982.[10] The experience in the arts and humanities was roughly comparable to that in the social sciences. In history, for example, the average journal cost $7 in 1970 and $20 in 1982; and in literature and languages, the corresponding costs were $8 and $19. In the humanities and social sciences, serials prices rose rather consistently at average annual rates of about 10 percent per year over this period, a rate of increase that looks modest only in comparison with the 15 percent-per-year rate of increase that occurred in the sciences, engineering, and medicine.[11]

During the eight years between 1982 and 1990, serials prices continued to rise very rapidly but not quite as rapidly as previously. Prices in chemistry-physics, mathematics and the other sciences grouped with it, and in engineering rose at average annual rates of roughly 10 to 11 percent (as compared with increases averaging about 15 percent per year during the 1970-82 period). In the humanities and social sciences, average rates of increase fell to 6 to 7 percent (as compared with increases of 10 percent per year in the period 1970-82). Of course, the absolute dollar increases were much greater in the more recent period as a result of higher base values. The most extreme case is, again, chemistry-physics, where the average price of a journal in 1990 was $413, as compared with $178 in 1982. When confronting such prices, it is small consolation to be told that the relative rate of increase has slowed.[12]

Three more general points can be made:

• First, in all three subperiods we have been examining serials prices rose faster than the GNP deflator. Only between 1963 and 1970 was the relationship at all similar, and even then prices of science and technology journals frequently rose twice as rapidly as the general price index; it was the journals in the humanities and social sciences that had price increases roughly comparable to the GNP deflator.
• Second, there is a striking consistency of rates of increase within similar groups of fields (in the humanities, history, literature-languages, and philosophy-religion; in the sciences, chemistry-physics, mathematics and other sciences, and engineering); clearly subject matter and associated variables count for a great deal in explaining both price levels and differential rates of increase in price.
• Third, the data reveal a persistent tendency for the most expensive serials to experience the largest relative price increases. One consequence is that the relative price differential between serials in, say, chemistry-physics and in literature-languages has steadily expanded: this ratio was 3.5 in 1963 (that is, the average serial in chemistry-physics was 3.5 times more expensive than the average serial in literature-languages), 5.4 in 1970, 9.2 in 1982, and 13.5 in 1990.

ORIGINS OF THE SERIALS CRISIS

Several studies have identified the principal distinguishing attributes of the serials that command the higher prices and have witnessed more rapid price increases.[14] Subject matter, as we have seen, is of particular importance: scientific and technical periodicals are generally more expensive. Titles published by commercial publishers tend to be more expensive than those published by learned societies or associations or other scholarly publishers (universities, university departments, university presses, museums, and research institutions). Other specific attributes that affect costs---a larger number of issues per year, a larger number of pages per issue, the presence of art work---naturally correlate positively with price. Journals that contain advertising were found to have lower prices.

Other studies have specifically considered the pricing practices of foreign commercial publishers, whose titles often command the highest prices.[15] Some price differential is, of course, to be expected, since increased distribution costs have to be recovered in some way. The question has been raised, however, whether the differential between what is charged local subscribers to European periodicals and what is charged North American subscribers is justified by the cost differentials. Prices of foreign periodicals often appear to correlate with various indicators of use and value, which suggests that in those instances pricing practices are largely value-based rather than cost-based.

It has also been argued that some publishers are engaging in discriminatory pricing. The producer can charge different prices in different markets when (1) the various markets for a particular product are clearly distinguishable, (2) the demand for the product within each differs, and (3) there is little or no possibility for resale of the product from one market to another. Discriminatory pricing of journals apparently originated in the early 1980s, when foreign publishers were seeking to offset losses resulting from disadvantageous exchange rates; however, there were no compensatory decreases when the U.S. dollar subsequently weakened. The pricing practices of a small number of foreign commercial publishers---Elsevier Science Publishers, Gordon and Breach, Pergamon Press, Springer-Verlag, Taylor and Francis, and others---have been subject to particular scrutiny, and one institutional study revealed that subscriptions to titles published by Elsevier, Pergamon, and Springer alone accounted for 43 percent of the increase in serials expenditures at the university in question between 1986 and 1987.[16] Elsevier's acquisition of Pergamon in the spring of 1991 intensified concerns about possible oligopolistic control of the market.[17]

This synopsis of views represents some of the concerns expressed by members of the library profession. Publishers, for their part, have argued that any valid analysis has to take account of cost increases resulting from such factors as increases in the number of pages per issue, and some of the more elaborate studies (such as the one described here) are concerned with trends in subscription price per page.

The most fully articulated model of journal pricing has been developed by Roger Noll and W. Edward Steinmueller and their colleagues at the Center for Economic Policy Research at Stanford University.[18] The Stanford group proposes a model that explains the interaction of various attributes of domestic journals, especially their cost structure and the nature of the demand for them.

The principal elements of the underlying cost structure of scholarly publishing, which is shared with other media products, are relatively high first-copy costs (the costs incurred in producing the first copy of the title---editorial work, typesetting, and so on) and relatively low marginal or incremental costs (the costs of printing and distributing each subsequent copy of the title). The publisher has to recover first-copy costs by charging a sufficiently high unit price---that is, in the case of scholarly journals, the price of a subscription. The average cost of each copy produced decreases as the number of subscriptions increases, and a small subscription base compels the publisher to charge a relatively high price in order to recover first-copy costs.

Critical to the degree of market control is limited substitutability of products (the Musical Quarterly cannot be substituted for the Journal of the American Musicological Society; research libraries must subscribe to both) and relatively inelastic demand (a market in which sales are relatively unresponsive to price increases because of the inability of the purchaser to find a good substitute and the perceived need of the purchaser for the product). Demand for periodicals tends to be less elastic than demand for monographs, either because of an implicit assumption about the greater importance of the scholarly journal as a vehicle for scholarly communication or because, in Richard De Gennaro's words, "[l]ibrarians have a weakness for journals and numbered series of all kinds. Once they get volume 1, number 1 of a series, they are hooked until the end. They love neat and orderly serials records and complete runs of periodicals on their shelves. Journals, in short, are the sacred cows of libraries."[19]

The basic cost structure, moreover, has further important effects in a market where new entrants compete for limited resources. Subscriptions gained by new entrants almost inevitably serve to reduce the subscription base of existing journals, in turn resulting in upward pressure on unit prices as publishers are compelled to spread first-copy costs over a smaller number of subscriptions.[20] Preliminary results from the Stanford study suggest that subscription price indeed correlates strongly with circulation; other things being equal, titles with smaller subscription bases ordinarily command higher prices and vice versa.

The demand side of the equation (that is, the demand for a particular journal) is affected by both library budgets and the pressures exerted by readers, especially faculty members, to purchase journals. Noll and Steinmueller make the important if obvious point that "to understand the market for journals requires an understanding of faculty utilization of them."[21] After describing the role of publication (especially in the "best" journals) in determining promotions and salary increases, they observe that:

... [A]s more faculty seek publication outlets, the demand to be published in a fixed number of "best" journals grows, and a smaller proportion of scholars succeed in publishing at the top of hierarchy. Recognizing this, both publishers and scholars seek to create new publishing outlets that create a new hierarchy, rather than enter at the bottom of an established one. Thus, scholars and publishers seek to narrow the scope of journals, attempting to create an outlet that will be read by people in a subspecialty and that will attain the status of being the second best place to publish for that subspecialty, rather than the twentieth or thirtieth best place to publish in the entire discipline. The result is a special kind of journal proliferation. As more academics seek to publish, and as more universities try to promote faculty research and reward scholarly publication, faculty and publishers jointly seek to create not only more journals, but ever more specialized journals that become important to all scholars in a subspecialty, although irrelevant to most scholars and students in a discipline.[22]

Journal proliferation and specialization drive down the average circulation of journals, which drives up the average subscription price. Moreover, because faculty prefer to avoid a lengthy hierarchy of journals in a discipline, in some sense all new journals are "essential" in that they constitute a natural home for articles of value. In a sense, all journals become at least second best in the hierarchy for a small number of scholars. Hence, libraries face not only increasing average prices owing to declining average circulation, but also intense demand to subscribe to all journals because every one is in some sense important.... The resulting performance of the journals market is socially undesirable and economically inefficient....[23]

It is also easy to see why prices of titles published by commercial publishers are almost inevitably higher than prices charged by nonprofit publishers. Nonprofit publishers---scholarly societies, for example---have a variety of ways of reducing first-copy costs so that unit prices can be kept relatively low. There are a variety of hidden subsidies that are unlikely to appear in the calculus of commercial publishers: in some instances societies levy page charges, which provide revenues denied the commercial publisher;[24] and in many instances, perhaps most, the editor of a scholarly journal published by a nonprofit press is either not compensated at all or receives only modest compensation. Ironically, the potential for revenue from advertising would appear to be greater for commercial publishers. Such revenue could, in theory at least, be applied primarily against first-copy costs.

This set of relationships, which underlies production of the scholarly journal, is likely to produce exactly the pattern we have seen: the size of the periodicals universe increases; relatively fixed materials budgets at libraries result in a decrease in the number of subscriptions per title, as available resources are redistributed among a greater aggregate number of available titles in the periodicals universe; prices per title increase as publishers seek to recover first-copy costs from smaller subscription bases; libraries redeploy materials expenditures in response to periodical price increases, protecting serials subscriptions at the expense of other library materials but still cutting back on some subscriptions; and library budgets are encumbered by rapidly rising outlays for serials.

Although the existing literature principally concerns the economics of the scholarly journal, there is some evidence of a similar kind of market dynamic in the production and acquisition of scholarly monographs. At one academic press whose experience is thought to be representative, the average number of hardcover copies sold per title in the humanities and social sciences declined between 1976 and 1986, from between 1,250 and 1,500 to fewer than 1,000. The smaller press runs caused by this decline in copies sold per title unquestionably put upward pressure on unit prices, since here too first-copy costs must be spread over the relevant number of units sold. Concomitantly, there was an increase in the number of titles published.

One explanation offered by several university press directors was that there have been fewer sales of each title to academic libraries, one of the principal markets for university press books.[25] Here again, we see the interaction between acquisitions practices and an underlying cost structure.

More generally, this set of interrelationships---among cost structures, patterns of demand, market characteristics, and the forces leading to more scholarly output---are entirely consistent with the empirical realities. They explain why in recent years increases in dollars spent on library materials have yielded little or no increase in the overall rate of acquisitions, while at the same time the number of items available for purchase has continued to increase. They are the explanation for the widening gap between the number of volumes added gross and book titles published.

Over the short term, libraries have responded to these circumstances primarily by redistributing their resources.[26] This mode of response cannot be sustained indefinitely, however, and it is already under challenge as both temporizing and inadequate. Increasingly, there is the realization that no institution, no matter how amply endowed with resources, can hope to maintain a self-sufficient collection into the indefinite future. The decisions of some of the wealthier institutions to increase acquisitions expenditures may be seen as only an interim expedient, although a critically important one, since maintaining reasonable continuity of coverage of the world's scholarly output must be considered an important objective, in order that the even more fundamental objective of maintaining access to the capital of scholarship can be met.

A more viable long-term solution will almost certainly entail fundamental reconfiguration of the dynamics of scholarly communication: perhaps some modifications to a reward system that in part explains the proliferation of scholarly journals and monographs; certainly some application of developing technologies to the problem of first-copy costs; surely much fuller use of new technologies to facilitate greater sharing of resources; and, conceivably, even alterations in the law that governs the rights to "published" material. Part 2 of this study discusses a range of options, most of which depend on greater use of new technologies, that could lead to quite different patterns of scholarly communication.

Endnotes

[1] Also, see Ann Okerson, "ARL Libraries React to Projected Serials Price Increases," ARL: A Bimonthly Newsletter of Research Library Issues and Actions 153 (November 7, 1990):2; and Ann Okerson, "Monographic and Serial Purchasing in 1992 Projected to Decline Again," ARL: A Bimonthly Newsletter of Research Library Issues and Actions 159 (November 12, 1991):8.

[2] Report of the ARL Serials Prices Project (Washington, D.C.: Association of Research Libraries, 1989). See therein the report by Ann Okerson, "Of Making Many Books" for a brief overview of the history of concern over serials prices.

[3] A number of members of the library profession have noted the limitations of national indexes for their purposes. See Mary E. Clack and Sally F. Williams, "Using Locally and Nationally Produced Periodical Price Indexes In Budget Preparation," Library Resources & Technical Services 27 (1983):345-356, especially p. 345; and Higher Education Price Indexes, 1990 Update (Washington, D.C.: Research Associates of Washington, n.d.), 19, where it is observed that "a price index reflects a pattern of consumption for a group of consumers, not for the individual. A single national index only approximates the price changes for a single represented consumer." In "German Book Prices," Book Research Quarterly 2 (Spring 1986):82-84, Steven E. Thompson published data that revealed that prices listed for books in various subject categories by the vendor Otto Harrassowitz are almost uniformly higher (and often considerably so) than those published in Buch und Buchhandel in Zahlen, a standard national index. The prices listed by Harrassowitz are more relevant to the acquisitions practices of American academic libraries because of the nature of the materials surveyed. Similarly, in Average Prices of British Academic Books, 1974-1984, Centre for Library and Information Management, Report No. 41 (Loughborough, Leicestershire: Centre for Library and Information Management, Department of Library and Information Studies, Loughborough University, 1985), Lawraine Wood demonstrated that the average price of academic books was higher than that of all books.

[4] Data on prices for hardcover books are from various issues of Publishers Weekly (see references cited). For purposes of this analysis the same categories that were excluded from consideration earlier (fiction, general works, home economics, juveniles, language, sports and recreation, and travel) are again excluded here, so as to define a data set that is most relevant to research libraries. Of course, as noted earlier, these data are drawn from national indexes; were one able to define a data set limited to academic materials, the rate of increase almost certainly would be seen to have been even greater.

Price data for periodicals are taken from various issues of the Library Journal (again, see references cited). Here too various categories were excluded from consideration. The data set is limited to the following fields: agriculture; business and economics; chemistry and physics; education; engineering; fine and applied arts; history; journalism and communications; labor and industrial relations; law; library science; literature and language; mathematics, botany, geology, and general science; medicine; philosophy and religion; political science; psychology; sociology and anthropology; and zoology. These categories are not identical to those used in Publishers Weekly, but they do seem appropriate for our purposes.

The underlying data on which the figures are based are shown in tables 6.1 and 6.2, Appendix B. The overall average prices for books and periodicals (shown in fig. 6.1) are unweighted averages of the values for the separate fields. Book prices do not include philosophy-psychology because data for this category do not begin until 1970.

[5] See, for example, Deana Astle and Charles Hamaker, "Pricing by Geography: British Journal Pricing 1986, Including Developments in Other Countries" in Library Acquisitions: Practice & Theory 10 (1986):165-181; Charles Hamaker, "Library Serials Budgets: Publishers and the Twenty Percent Effect," Library Acquisitions: Practice & Theory 12 (1988):211-219; Robert L. Houbeck, Jr., "British Journal Pricing: Enigma Variations, or What Will the U.S. Market Bear?" Library Acquisitions: Practice & Theory 10 (1986):183-197; Frederick C. Lynden, "Prices of Foreign Library Materials: A Report," College & Research Libraries 49 (May 1988):217-231; Kenneth Marks, Steven P. Nielsen, H. Craig Petersen, and Peter E. Wagner, "Longitudinal Study of Scientific Journal Prices in a Research Library," in College & Research Libraries 49 (March 1991):125-138; and H. Craig Petersen, "University Libraries and Pricing Practices by Publishers of Scholarly Journals," Research in Higher Education 31 (August 1990):307-314. Also see Economic Consulting Services, "A Study of Trends in Average Prices." The ECS report concentrates on four large publishers: Elsevier (Netherlands), Pergamon (U.K.), Plenum (U.S.), and Springer-Verlag (West Germany). The report reviews price data against publishers' estimated costs from 1973 through 1987.

[6] See Kenneth E. Marks, Steven P. Nielson, H. Craig Petersen, and Peter E. Wagner, "Longitudinal Study of Scientific Journal Prices," for a longitudinal study that attempts to distinguish between foreign and U.S. journals as well as between commercial and noncommercial publishers. The rate of increase in journal prices published by foreign commercial publishers was found to have been greater in the period between 1967 and 1987 than the rate for any other journal type studied (i.e., U.S. commercial, U.S. noncommercial, and foreign noncommercial).

[7] Law is the exception. Book prices in this field increased 15.5 percent, as compared with 16.4 percent for the GNP deflator.

[8] Although the price of periodical subscriptions in the field of business-economics is shown in figure 6.3 as increasing at essentially the same rate as medicine throughout this period (nearly as rapidly as science-technology), it should be noted that business-economics began from a much smaller base---$6.31 per subscription in 1970 compared with $23.44 per subscription for medicine. The business-economics composite in figures 6.3 and 6.5 is an average of the business-economics and labor-industrial relations categories displayed in Appendix Table 6.2.

[9] To facilitate comparisons, the GNP deflator in figures 6.4 and 6.5 has been indexed to the average price of a periodical subscription in each broad field. In the arts and humanities, for example, the GNP deflator was set equal to $5 in 1963, which is a simple average of the average price of periodicals in 1963 of fine and applied arts, history, literature-language, and philosophy-religion.

[10] Psychology is something of a middle case, standing between the natural sciences and the social sciences in price behavior, as it tends to stand between them substantively as well.

[11] Since the GNP deflator rose at an unusually rapid rate during this period, the real rates of increase obtained by dividing by the deflator appear more modest. However, many of the factors driving up the GNP price deflator (for example, the extraordinary increases in the price of energy related to the Arab oil embargo) had little to do with increases in serials prices. This is a good example of a situation in which nominal price increases for a particular commodity or service are more revealing than real price increases. Working only with deflated values would obscure the power of some of the forces pushing up serials prices, such as the proliferation in the number of journals.

[12] A variation of the pattern is well documented in Warren T. Seibert and Marjorie A. Kuenz, Growth and Change in 67 Medical School Libraries, 1975-1989 (Bethesda, Md.: National Library of Medicine, 1992).

[13] For an excellent general treatment of the question of serials prices, see Ann Okerson, "Periodical Prices: A History and Discussion," Advances in Serials Management 1 (1986):101-134.

[14] See, for example, Henry H. Barschall, "The Cost-Effectiveness of Physics Journals," Physics Today 41 (July 1988):56-59; Sandra R. Moline, "The Influence of Subject, Publisher Type, and Quantity Published on Journal Prices," The Journal of Academic Librarianship 15 (1989):12-18; and H. Craig Petersen, "University Libraries."

[15] See references given in n. 5.

[16] This finding was reported by Hamaker of Louisiana State University and quoted in Richard Dougherty, "Periodical Price Escalation: A Library Response," Library Journal 113 (May 15, 1988):27-29.

[17] For a sampling of general reactions, see "Librarians Fear Elsevier Purchase of Pergamon," The Chronicle of Higher Education 37 (April 10, 1991):A5; and "Journal Prices Increase 52%," The Chronicle of Higher Education 37 (June 26, 1991):A5.

[18]

Some of the preliminary results of this study, still in progress, are summarized in papers submitted to The Andrew W. Mellon Foundation, which is funding the research: Lisa Lieberman, Roger Noll, and W. Edward Steinmueller, "Proposal to the Mellon Foundation: Economic Analysis of Scholarly Periodical Costs," "An Economic Analysis of Scientific Journal Prices: Preliminary Results," and "Economic Analysis and Empirical Protocol for Examining Scholarly Periodicals Pricing." Their preliminary analysis appears in published form in Roger Noll and W. Edward Steinmueller, "An Economic Analysis of Scientific Journal Prices: Preliminary Results," Serials Review 18 (1992):32-37. This research concentrates on domestic journals only because patterns of circulation are critical to the analysis; accurate time-series data on this statistic are available only for domestic journals, since U.S. publishers must publish circulation information in order to receive second-class mail privileges.

Some of the propositions reflected in the Stanford model are anticipated in two excellent studies: David W. Lewis, "Economics of the Scholarly Journal," College and Research Libraries 50 (November 1989):674-688, and Paul M. Gherman and Paul Metz, "Serials Pricing and the Role of the Electronic Journal," College and Research Libraries 52 (July 1991):315-327.

For another conceptualization of journal pricing and how it may be affected by electronic technologies, see: Malcolm Getz, "Electronic Publishing: An Economic View," Serials Review 18 (1992):25-31.

[19] "Escalating Journal Prices: Time to Fight Back," in De Gennaro, Libraries, Technology, and the Information Marketplace, Selected Papers (Boston, Mass.: G. K. Hall & Co., 1987), 103-113, especially p. 104. One study suggested that faculty expectations are also involved. Richard M. Dougherty reported that in conversations with faculty members at the University of Michigan who served as editors of scholarly journals, he discovered that "the economic stress libraries were feeling was not generally understood or appreciated" and that "serials cancellations can ... be a very volatile issue to some faculty and that they will rise to defend the library's materials budget."

[20] A number of members of the library profession have suggested in private conversations that when a new journal title is added to their collection, an existing subscription has to be cancelled, which suggests that successful new entrants do indeed affect the subscription bases of existing titles.

[21] Noll and Steinmueller, "Economic Analysis," 33.

[22] Ibid., 33-34.

[23] Ibid., 35.

[24] In "Combating High Journal Costs," Science 244 (June 9, 1989):1125, Philip H. Abelson sketched a brief history of the emergence of commercial publishers as important participants in the process of scientific communication. Abelson suggests that the levying of page charges by the societies is among the factors that may have created the opportunity for commercial publishers to enter the market.

[25] Bailey, The Rate of Publication of Scholarly Monographs, 14-17. Explanations by the university press directors were offered in personal conversations.

[26] Some institutions have also attempted various collaborative efforts as discussed in Chapter 4, n. 14. However, even with these efforts institutions have found it necessary to redistribute resources toward serial subscriptions and away from monographs.