College and Research Libraries Resource Sharing or Cost Shifting?-The Unequal Burden of Cooperative Cataloging and ILL. in Network Charles B. Lowry The values and reward system that characterize library cooperation are drawn from an environ- ment quite different from that of the modern library network, which is founded on the shared cataloging/ILL utility. Reflecting as it does an earlier system of values about participation, the reward system for the technology-based network encourages use of but not contribution to the "national library database." This paper examines the dilemmas of participation in resource- sharing networks and proposes solutions based on shifting the inducements toward contribu- tion by remunerating the contributing library for the unit cost of original cataloging and the suggested restructuring. The restructuring suggested is applicable to library network/utilities in general. However, OCLC's recently announced "contribution pricing" may serve as a model for changing the reward system, which is a prerequisite to preserve the "national library database" that the OCLC represents. COOPERATION-WHAT IS IT? In the post-World War II period, the concept of library cooperation has tended to focus on two primary and critical activities-shared cataloging and interli- brary lending. But in recent years, a more comprehensive notion of the dimensions of library cooperation has become essen- tial to our lexicon. This extends right down to local cooperative activities and in- cludes shared collection development, preservation, and training, particularly for technological purposes .1 To understand the dimensions of library cooperation today, we must understand the modem development of library net- works. JoAn Segal notes that "there are implicit levels of networking . . . of sev- eral sorts: type-of-library networks, size- of-library networks, and geographical- area networks at various aggregations of 2 . area. Recently Edward Walters has applied the social sciences concept of institutions to provide an organizing model to group the variety of organizations that we call li- brary systems, networks, consortia, asso- ciations, and cooperatives. He has identi- fied five forms of networking: (1) the unit · cost networks such as OCLC or RUN, with the size and revenue sources to raise . substantial capital for development; (2) the multistate regional auxiliary enter- prise networks such as AMIGOS or 50- LINET, with their low capital investment and fee-based assessment with a more limited ability to amass capital; (3) the authority-sanctioned networks such as IL- LINET and FED LINK established by gov- ernments or government agencies that have jurisdiction and draw on those au- Charles B. Lowry is Director of Libraries at the University of Texas at Arlington, Arlington, Texas 76019. This article is based on a paper delivered as a keynote address for a miniconference on the Cost of Cooperation in Aca- demic and Research Libraries jointly sponsored by the University of North Texas, School of Library and Informa- tion Sciences, and AMIGOS Bibliographic Council held in Denton, Texas, May 11, 1988 . • 11 12 College & Research Libraries thorities for their power to coerce the ac- ceptance of standards and as a source of capital for new projects; (4) the discipline and type-of-library networks that are cre- ated for the purpose of sustaining either a discipline, (e.g. law, medicine, and agri- culture), or an institutional form (e.g. uni- versity or corporation); and (5) local con- sortia or proximity networks that are founded and sustained because geo- graphical proximity makes library cooper- ation more timely and often more effec- tive. These latter vary greatly in the kinds of organizations they include, size, gov- ernance, and programs. Each of the five types of library net- works is part of a national library "net- work picture" that is interlocking in its services and belief systems. One cannot talk about a type of network without de- fining its relationship with the others. 3 Moreover, most academic libraries are members of several of these types of net- works. The University of Texas at Arlington and the University of South Alabama-two libraries where the author has most recently been director-have been simultaneously members of four or five types. This degree of participation brings attendant cost and benefits. The question becomes "is the cost outweighed by the benefit?" Several problems arise in this networking environment because of the underlying assumptions we have about cooperation, and the need for man- agerial control based on a realistic cost model in our "not-for-profit industry." There are also numerous hidden or indi- rect costs to which no attention is paid as we work towards the abstract ideal of co- operation. DILEMMAS OF NETWORK RESOURCE SHARING It has been pointed out frequently that resource sharing in any type of network environment may be used by libraries as a prop to reduce their own obligations to build adequate local resources. 4 More- over, there is a growing concern among larger research libraries that the primary objectives they have for cooperating through utilities-cataloging and interli-: brary loan-are heavily subsidizing ser- vices they neither need nor want. 5 There is January 1990 considerable anecdotal and some empiri- cal evidence to illustrate that libraries, in- cluding members of ARL, take unfair ad- vantage in relation to their contributions to network participation. The assumption is often made that research libraries with the largest budgets and collections are the largest ILL net-lenders. However, among those twenty-five ARL libraries with the largest budgets, only ten are among the top twenty-five net-lending libraries in ARL. Of the other fifteen, eleven are not among the ARL top forty largest lenders. Curiously, of the ARL libraries with the twenty-five largest collections, fifteen are not among the twenty-five largest lenders and eleven are not among the ARL top forty-exactly the same numeric result. By contrast, nine of the top twenty-five ILL net-lenders in ARL have budgets that are not among the ARL top forty, and eight of the top twenty-five net-lenders have col- lections which are not in the top forty. 6 These striking numbers indicate clearly that the ethos of ILL resource sharing is not necessarily associated with either the largest budgets or collections. An exami- nation of contributed cataloging would likely produce similar results. There cer- tainly has been no lack of conjecture at OCLC Users Council meetings that li- braries with extremely high search levels and low holdings and FTU use are ''rip- ping off" the database. "Networking becomes more an eco- nomic decision than one of 'apple pie and motherhood.' '' Daily, library managers must refuse to acquire new journal titles, cut book ex- penditures, and in numerous ways de- cline to acquire materials for patrons be- cause they do not have the money to do so. Only in the area of "free" ILL do we assume we can get something for nothing. Indeed, free ILL can become a straw man used in defense of inadequate budgeting for access to materials primary patrons need. Paul Gherman argues that ''networking becomes more an economic decision than one of 'apple pie and motherhood.' Ac- cordingly, the goal of 'the greatest good for the greatest number' is tempered by the recognition that nothing is free, by the desire to avoid exploitative imbalance of services between libraries, and by prefer- ence for contractual or cost-based transac- tions over open-ended moral commit- ment."7 Criticism of the value of network coop- eratives comes not just from the academic research libraries but also from the public library sector. Notable among these is Thomas Ballard. He states that, ''an en- tirely unsubstantiated belief in coopera- tive resource sharing as the best way to improve library services has been consid- ered sufficient reason to spend millions of dollars. This is dogma-purely and sim- ply." Ballard offers criticism at another level, calling into question the basic as- sumptions of cooperative network activi- ties with a belief that ''for the patron, bib- liographic networking is merely a return to closed stack concept," and that the ''levels of inter-library loan indicate that people really don't want everything. " 8 Such opinions from colleagues no doubt raise hackles among the majority in our profession. This occurs because of an intri- cate belief system that (1) no library has the resources to satisfy all its patrons' needs, (2) library cooperative efforts will fill the gap between patron needs and re- sources, (3) a national library network built from the bottom up is a self evident good, and (4) libraries have a democractic responsibility to minimize the gap be- tween the information rich and the infor- mation poor. 9 Library networking even re- ceived pontifical support when in 1982 "Pope John Paul II ... expressed deep in- terest in the Theological Library Network founded . . . in Rome by 60 representa- tives of pontifical universities and other Catholic as well as non-Catholic institu- tions. " 10 SQUARING "BELIEF SYSTEMS" AND BEHAVIORS How do we reconcile our belief systems with the reality of behavior among institu- tional participants in library cooperation? Resource Sharing 13 With the emergence in the 1970s of major bibliographic utilities containing large ex- traordinarily rich electronic databases, the foundation was laid for the development of a complex array of networking arrange- ments. Today, the local library coopera- tive is possible only as part of this larger scheme. These utilities came into exist- ence and libraries became cooperating members for two primary reasons-tore- duce the cost of cataloging library materi- als and to use the databases built in this process as a new vehicle for interlibrary lending. No one wants to go back to doing things pre-1970, because the utilities have been extraordinarily effective in helping li- braries reduce the cost of labor-intensive manual systems. 11 Nonetheless, there are inherent temptations in the way we struc- ture member participation, which encour- age libraries to ignore the belief system and to take advantage of their peers. Two examples of this structural problem-ILL and shared cataloging-are the focus of the balance of this paper. These two activi- ties alone are fundamental illustrations of the great success of automated/electronic networks, and probably point the way to the most cost-effective cooperative activi- ties of the future. 12 However, they are not without flaws. In the discussion which fol- lows, OCLC is the prominent example, but the other major bibliographic utilities are affected by similar conditions with the same results. INTERLIBRA Y LOAN Interlibrary loan should properly be characterized as interlibrary sacrifice, be- cause we have tended to focus narrowly when determining the cost on a tightly de- fined range of issues, including local staff time, equipment costs, lending charges by the utilities, charges from the lending li- brary, success or fill rate, and turnaround time. 13 These are not inconsequential is- sues, but they are not sufficient to help us in determining the real cost benefit of ILL cooperation or in defining new ways toes- tablish an adequate added value to this service. It has been frequently argued that the tradition of "free" interlibrary loan is no longer adequate in the electronic environ- ment of the modern bibliographic utility . 14 College & Research Libraries and that small libraries will lose their ini- tial euphoria over the good of participa- tion as they approach net lending versus net borrowing. 14 Herbert White recently stated that ''our attitude comes from ac- ceptance of interlibrary loan not as a prag- matic commodity, but as a moral precept .... This premise now conflicts directly with our evolving responsibilities as re- sourse managers accountable to those who fund us and those we serve di- rectly. ''15 The problem, as he sees it, is that the lowest priority in ILL becomes lending our materials and the highest priority be- comes borrowing materials for our pa- trons. ''The pricing of OCLC/ILL charges from the perspective of actual costs or as an inducement to lend makes little sense.'' The ILL system as presently constituted tempts participants to make cooperation a one-way street, though many avoid the temptation. "The net lenders know who they are, and they also know that they provide far more than they receive.'' 16 An example of the dilemma is UT Arlington Libraries' ILL Department which has long prided itself on handling an OCLC/ILL re- quest within twenty-four hours and has a well-earned reputation for responding quickly. The net result is that lending is on the increase. If it were not for the fact that requests for borrowing have skyrocketed due to a major surge in doctoral education and externally funded research, UTA would have long since been a large net- lender. In this situation, a library manager is faced with the dilemma of either adding staff and equipment to the ILL Depart- ment in order to meet demands to lend materials, or setting the principle that bor- rowing is the first priority and responses to requests to lend will be allowed only when all of the borrowing is done. If we were to adopt this stance, requests to bor- row would bypass UTA in the OCLC queue and, before long, requests would begin to decline. January 1990 The pricing of OCLC/ILL charges from the perspective of actual costs or as an in- ducement to lend makes little sense. The borrowing charge is 83 cents, and the lending credit is 20 cents. There is no pal- pable difference between these two costs from OCLC' s perspective, and certainly no reason to encourage lending at so insig- nificant a level. The 20 cents credit for lending does not even qualify as a sop, al- though OCLC' s primary objective is to support system costs, not encourage ILL lending. Increasingly we hear calls for the actual ''determination of cost honestly, in terms that meet the accounting criteria of the supplier whose perception is ulti- mately all that matters. Cost is not the same thing as price."17 A recent experi- ment by the Missouri State Library in which "lending libraries were reimbursed - $5.00 for each item loaned, above the number of items borrowed by that li- brary'' seems to point in a reasonable di- rection.18 Likewise, medical libraries par- ticipating in TALON pay a $4.00 fee per item borrowed. But the determination of standard costs or the call for standard fees19 will take more than guesswork, and must include new cost elements which will be discussed below. SHARED CATALOGING It is curious that the cost structure of shared cataloging has not had the same scrutiny nor produced as much literature as ILL. There are two obvious omissions here. The first is that studies of pricing tend to focus on the surcharges added by regional networks to OCLC costs. 20 These are of some importance, but they are not paramount since the surcharges are a small percentage of the total cost of cata- loging and contribute to direct services of the regional networks. The credit on OCLC for original contributed cataloging is $1.00 a record. Libraries participating in the "enhance project" receive $1.29 for records which they upgrade. One wonders how many libraries could demonstrate that $1.00, or even a $1.29 compensated them adequately for the ex- pense of original cataloging. On the other hand, the OCLC charge for prime time use of a record in the database is $1.29. If there ever was a disincentive to contribute origi- nal cataloging, this price structure is it. The truth is that utilities, particularly OCLC, have reaped the benefit of added value of cataloging supplied by libraries to their union databases. At present OCLC estimates that member input (including GPO and NAL) is about 76 percent of cur- rent cataloging in the OLUC (OCLC on- line union catalog). Although libraries have viewed utilities as entities they have created to do work for hire, the reality is that libraries are supply- ing piece-work cataloging to the utilities at a sweat labor price which is resold to other libraries for a tidy sum. How do we correct these shortcomings and remove the temp- tation to reduce our costs at the expense of other libraries by waiting for some other li- brary to assume the cost of original cata- loging? COST -BENEFIT CRITERION AND UNIT COST METHOD The cost benefit criterion theme stated simply is that as a system changes its ex- pected additional benefits usually must exceed its expected additional costs. 21 There have been numerous calls for cost benefit or cost effectiveness analysis. This always sounds complex. Perhaps it can be simplified. 22 Kaye Gapen says it is time to begin treating the ''access cost'' as equal to the "collection cost." Put another way, ''There will be new choices to be made in smaller libraries between low fixed costs and high incremental costs per search strategy. Larger libraries are more likely to choose the high fixed cost and low incre- mental cost strategy at least for very com- monly used datafiles. " 23 . Cost benefit would work something like this: a library borrowing on ILL pays to the lending institution a cost based on the high fixed cost experienced by the lending institution for maintaining materials in its collection and recovering them for lending purposes. Similarly, a library contributing original cataloging benefits from this labor in a manner equivalent to the effort and expense of its work. In either case, there is a positive inducement to cooperate and share collection resources or cataloging la- bor and no inducement to take advantage of the system. The actual unit cost for ILL activities Resource Sharing 15 must be established in order to judge the cost benefit. "Techniques have been de- veloped permitting uniform cost analysis for varying libraries. These techniques take into account the fact that nearly all the direct budget contributes to more than one service (book stock and technical ser- vices), are a kind of overhead, and that de- tails of cost allocation depend on the ac- tual levels of service rendered. " 24 Obviously, unit cost will vary from library to library, but a sample can provide a rela- tively standard method of determining unit costs. A good example of the unit cost method is the work done recently at Vir- ginia Tech. 25 How would this work in our electronic network environment? Put an- other way, how will unit cost be estab- lished and who will pay? Examples of ILL and cataloging help explain the use of the cost benefit and unit cost approach. To use White's phrase, ILL is temporary acquisition. ''The real decision is between permanent and temporary acquisition. Permanent acquisition costs more ini- tially, but that cost is not repeated for each future use. Temporary acquisition costs less one time, but if it is repeated the costs reach an equal level. " 26 We should also recognize that each time a library lends through ILL it may lose an opportunity for its own patrons to use an item in which it has invested at high fixed cost. This hap- pens in two ways. First, the item may not be on the shelf when the patron goes to look for it. Second, there is a loss due to wear-and-tear on the item which now has one less circulation in its expected ''life span." A recent in-house study entitled ''Cost Data for the Virginia Tech Library" estab- lished the total cost of purchase and shelv- ing monograph volumes at $106 each and the total cost for purchase and shelving of serials volumes at $181 each. This per item investment represents a finite number of uses, and each time a library lends materi- als to non primary patrons the net result is a potential loss on that investment. As- suming 25 circulations per item for a typi- cal publishers book binding, Virginia Tech is dissipating $4.24 per monograph for each ILL loan that it makes. 17 Likewise, if a "class A" serial binding lasts 50 circula- tions, the loss per photocopy-loan 16 College & Research Libraries amounts to $3.62. Conversely, the bor- rowing library is saving an equal amount. An ARL library lending 70,000 items per year (half books and half serial copies) has lost $275,100 of its capital investment in collections. Using the unit cost method, a formula could be developed that factors in the lending libraries staff expenditures, periodical expenditures and monograph expenditures of over a five-year period. The staff expenditure would have to be in- dexed on the average salary for staff. Peri- odical expenditures would be indexed on the average cost per title and book ex- penditures on the average cost per vol- ume. This would take into account re- gional differences in wages and the variable character of collecting costs de- pending on the relative emphasis in each institution on different types of materials. It might also be helpful to develop an in- dex of collection size, because there is an implicit cost for maintenance, although this may be captured in the staff cost in- dex. By gathering these statistics, annu- ally libraries would provide the basis for establishing the unit cost for lending both photocopy and book stock. This unit cost would be an added value to the interli- brary loan transaction, which would be paid directly to the lending institution by the borrowing institution. A palpable billing nightmare could be caused by such a system, but there is a simple way out. It has been suggested that ''since OCLC already bills requestors for its own transaction fee, it could possibly expand that accounting capability to in- clude collecting and disbursing ILL fees. " 28 At its March 16, 1989, meeting the AMIGOS Board voted to request formally that such a scheme of payment be incorpo- rated into the OCLC "New System" de- sign. That action was based on a white pa- per submitted by the AMIGOS ILL Policy Review Committee. The white paper pro- poses that each year participating libraries supply OCLC with their ILL charges based on type of materials. ~hese ILL charges, or the actual unit cost, would be posted on the OCLC system in the name and address directory. Obviously, li- braries would make borrowing choices, at least in part, based on these borrowing January 1990 charges. When a library borrowed mate- rial, the unit cost (ILL charge) of the lend- ing library would be added to the OCLC borrowing charge. On a monthly basis the difference between lending and borrow- ing costs would be reconciled for OCLC billing. This sort of system would not cause a billing and paying nightmare, but would eliminate one that currently exists for some libraries. For instance, a State of Texas payment voucher for ILL costs UTA Libraries, the University and the State more than $40.00 in paperwork expenses just to pay the lending library as little as a dollar. Moreover, this sort of unit-cost sys- tem would not prevent libraries from con- tinuing existing arrangements of recipro- cal borrowing agreement,s if they chose to do so. ''The real cost of original cataloging is significantly higher than the OCLC payment for original contributed cat- aloging or the 'enhance project.' " Many libraries are moving toward unit cost recovery, whether they call it that or not. Furthermore, each library could choose to fund this real ILL acquisition cost or pass on the expense to their pa- trons. In any event, a library choosing in favor of a ''high cost per transaction but a small fixed cost'' would bear the respon- siblity for that choice and Interlibrary Loan would cease to be Interlibrary Sacri- fice.29 The cost of cooper:ative cataloging presents a slightly different set of issues, but deserves much more discussion than it has received. The point was made earlier that the cost of original cataloging is signif- icantly higher than the OCLC payment for original contributed cataloging or the ''en- hance project.'' Establishing the unit cost for cataloging in an individual library is probably not necessary. Instead, a sample of cost elements in cataloging for a num- ber of libraries could be used to establish an average unit cost for original cataloging of different formats perhaps at a minimum of $50 per title. Any library would be enti- tled to payment for its contributions to the OLUC based on these unit costs which would certainly be far higher than the present payment for original input. When would a library be entitled to pay- ment? There are several possibilities. Pay- ment could be made at the time of the ini- tial input of the record if the title had not already been cataloged or if the re- cataloging was being done by an "en- hance" library. A better system would pay a library when the cataloging copy was used by another library, either by paying in a lump sum (e.g., $50.00) or pay- ing incrementally for each use of the rec- ord until the lump sum was reached (e.g., $5.00 a use up to the maximum of $50.00). The latter method has the advantage of letting re-use of the record establish its fair market value. This brings us to one final question. Who will pay the unit cost for this original cataloging? The point was made earlier that the added value for origi- nal cataloging is contributed by the mem- ber libraries, and at least in some measure it is represented in OCLC' s corporate eq- uity $49 million or net worth. 30 Thus, OCLC should pay the lion's share of this new unit cost. CONCLUSIONS OCLC recognizes that the structure of member participation-whether ILL or contributed cataloging-induces use of, but not contribution to the OLUC. During the last year, OCLC has been exploring ways to resolve this dilemma. Recently it announced a highly innovative restruc- turing of pricing arrangements called ''Contribution Pricing.'' If implemented as planned in 1990, the structure of "Con- tribution Pricing'' may well provide a foundation for resolving the problems of the present network arrangement. Briefly, the new scheme makes ''least cost participation occur at full database contribution levels" by charging for ac- cess (bibliographic searches and holdings displays) and giving credits for contribu- tion (original cataloging, add/ delete hold- ings, create/update LDR' s, and ILL lend- ing and requesting). Removal of economic temptation to contribute less and reward Resource Sharing 17 for continued full cataloging and holdings contribution are among the benefits the OCLC says should arise from ''Contribu- tion Pricing." Whether the reward for contributed cataloging will reflect the real unit cost depends on how realistic OCLC is about paying libraries for the added value they provide to OLUC. Whether the temptation to borrow first and lend later is diminished will depend on whether an ac- counting mechanism for recovery of the unit cost of lending is provided in the New System for those libraries that wish to use it. OCLC deserves credit for understand- ing a fundamental flaw inherent in the structure of network participation and praise for a creative new approach to rem- edying it. 31 Moreover, the preservation of the national library database, which has figured largely in recent debates over ownership of bibliographic records (e.g., OCLC copyright) and the encouragement of broad participation by libraries in build- ing that database is likely to be dependent on just such a change. So long as the inducements in our net- work arrangements reward borrowing, using original cataloging contributed by other libraries, or downloading records without setting the holdings byte-we may expect the present state of affairs to continue. Unequal contribution is a natu- ral and expected outcome. What should surprise us is that many, perhaps most, li- braries make an honest effort to contribute their share. The point argued here is that the inducements should have long since been arranged to reward participation. This paper touches on the two most suc- cessful examples of cooperation in the electronic environment-shared catalog- ing and interlibrary lending. There are po- tential areas of cooperation that we are just beginning to explore including collec- tion development, preservation, and training. These incipient activities will present many of the same issues as those of cataloging and ILL. Thus, the library community faces a significant challenge in resolving the current dilemmas explored in this paper. We should be reminded in this undertaking of the old adage "not to decide, is to decide." 18 College & Research Libraries January 1990 REFERENCES AND NOTES 1. Richard M. Dougherty, "Resource Sharing among Research Libraries: How It Ought to Work/' Collection Management 9:79-80 (2/3, Summer/Fall 1987); Edward M. Walters, "The Issues and Needs of Local Library Consortium," Journal of Library Administration 8:25-28 (3/4, Fall/Winter 1987); and Irene B. Hoadley, "The Future of Networks and OCLC/' Journal of Library Administra- tion 8:85-87 (3/4, Fall/Winter, 1987). 2. JoAn S. Segal, "Networks," in Financing Information Services: Problems, Changing Approaches, and New Opportunities for Academic and Research Libraries, ed. by Peter Spyers-Duran and Thomas W. Mann, Jr. New Directions in Librarianship, no.6, p.141 (Westport, Conn.: Greenwood Press, 1985). 3. Walters, p.15-24. 4. Richard DeGennaro, "Resource Sharing in a Network Environment/' Library Journal104:355 (Feb. 1, 1980); Dougherty, p.80; and Donald E. Riggs, "Networking and Institutional Planning/' Journal of Library Administration 8:64 (3/4, Fall/Winter 1987). 5. Thomas W. Shaughnessy, "Management Perspectives on Network Membership," Journal of Li- brary Administration 8:10 (3/4, Fall/Winter 1987). 6. Association of Research Libraries, ARL Statistics, 1987-88 (Washington, D.C.: Association of Re- search Libraries, 1989), p.42, 49, 54, 55. 7. Paul M. Gherman, "Vision and Reality: The Research Libraries and Networking," Journal of Li- brary Administration 8:52 (3/4, Fall/Winter 1987). 8. Thomas Ballard, "Dogma Clouds the Facts/' American Libraries 257 (April, 1985); see also Richard T. Sweeney, "Financial Impacts of Networking and Resource Sharing/' Public Library Quarterly 2:91-94 (3/4, Fall/Winter 1981). 9. Walters, p.23. 10. "Pope Encourages Library Network/' Catholic Library World 53:418 (May/June, 1982). 11. See, for instance, Kazuko M. Dailey, Grazia Jaroff, and Diana Gray. "RLIN and OCLC-Side by Side: Two Comparisons Studies,'' in Advances in Library Administration and Organization, ed. by Gerard McCabe, Bernard Kreissman, and W. Carl Jackson, v.1 (Greenwich, Conn.: JAI Press, 1982); Elaine Kemp and others, "A Comparison of OCLC, RLG/RLIN, and WLN," Journal of Li- brary Automation 14:215-30 (3, September, 1981); Danuta A. Nitecki, "Online Interlibrary Ser- vices: An Informal Comparison of Five Systems/' RQ 20:7-14 (Fall1981). 12. Sweeney, p.95. 13. Gloria S. Cline, "The High Price of Interlibrary Loan Service" RQ 27:80-86 (Fall1987); Izabella Taler, "Automated and Manual ILL: Time Effectiveness and Success Rate" Information Technolo- gies and Libraries 1:277-80 (September 1982). 14. De Gennaro, p.353. 15. Herbert S. White, "Interlibrary Loan: An Old Idea in a New Setting" Library Journal112:53 (July 1987). 16. Ibid. 17. Ibid. 18. RichardT. Miller, "Net Lending: A Follow-Up Report" Show-Me Libraries 38:3-4 (August 1987). 19. See, for instance, Cline, p.80-86. 20. Leslie R. Morris, "Network Prices: Let the Buyer Beware" Technical Services Quarterly 4:57-66 (Summer 1987). 21. Charles T. Horngren, Introduction to Management Accounting (Englewood Cliffs, N.J.: Prentice- Hall, Inc., 6th ed., 1984), p.13-16, 685. · 22. Martin M. Cummings, "Cost Analysis: Methods and Realities/' Library Administration and Man- agement, 3:181-83 (Fall1989); and Paul Metz, "Special Pleading versus Self-Discipline in the Fi- nancing of Public Services" Journal of Academic Librarianship, 14:208-13 (Sept. 1988). 23. D. Kaye Gapen, "Strategies for Networking in the Next Ten Years" Journal of Library Administra- tion 8:118-27 (Fall/Winter 1987). 24. Paul B. Kantor, "The Relation Between Costs and Services in Academic Libraries/' p.74. In Fi- nancing Information Services: Problems, Changing Approaches, and New Opportunities for Academic and Research Libraries. Ed. by Peter Spyers-Duran and Thomas W. Mann, Jr. New Directions in Librari- anship, no.6 (Westport, Conn.: Greenwood Press, 1985). Resource Sharing 19 25. Paul M. Gherman and Lynn Scott Cochran, "Developing and Using Unit Costs: The Virginia Tech Experience" Library Administration and Management 3:93-96 (2, Spring 1989). 26. White, p.54. 27. For an average number of circulations, see Maurice F. Tauber. ed., Library Binding Manual: A Hand- book of Useful Procedures for the Maintenance of Library Volumes (Boston, Mass.: Library Binding Insti- tute, 1972), p.22-24, 36-37, 140. 28. De Gennaro, p.354. 29. Gapen, p.127. 30. Online Computerized Library Center, Inc., OCLC Annual Report, 1987/88 (Dublin, Ohio: OCLC, Inc., 1988), p.l. 31. The description of "Contribution Pricing" provided here is based on handouts and presentations by OCLC staff given during the AMIGOS Bibliographic Council, Inc. Membership Meeting in Dal- las, Texas, May 17, 1988, and the OCLC Users Council Meeting in Dublin, Ohio, May 25-26, 1989, both attended by the author.