Washington, D.C.
October 18-20, 1995

Building Partnerships that Shape the Future

JSTOR and the Economics of Scholarly Communication

William G. Bowen, President
The Andrew W. Mellon Foundation

I am delighted to have this opportunity to describe JSTOR (our acronym for “Journal STORage”). This denizen of the world of electronic databases began life as one of several demonstration projects funded by The Andrew W. Mellon Foundation, moved rapidly from infancy to adolescence, and now enjoys an independent existence, having been incorporated as a separate, nonprofit entity within the last few months.

Background: A Brief History and A Summary of Objectives

In its original incarnation, JSTOR was conceived to be an electronic database containing “faithful replications” of all pre-1990 issues of ten core scholarly journals in the fields of economics and history (including the American Economic Review and the American Historical Review). While the Foundation’s long-term interests in the field of scholarly communications are broad, in this instance we wanted very much to be practical and to test concepts and general approaches by working with a manageable set of well-defined materials within a circumscribed terrain. Hence, the narrow definition of the initial content of JSTOR.

In creating JSTOR, we set out to serve three objectives simultaneously: (1) to improve dramatically access to journal literature for faculty, students, and other scholars by linking bitmapped images of journal pages to a powerful search engine; (2) to mitigate some of the vexing economic problems of libraries by easing storage problems (thereby saving prospective capital costs involved in building more shelf space), and also by reducing operating costs associated with retrieving back issues and reshelving them; and (3) to address issues of conservation and preservation such as broken runs, mutilated pages, and long-term deterioration of paper copy.

I will not take time to explain why we chose to start with the fields of economics and history (or the ten specific journals with which we are working), but I will anticipate another frequently asked question: why did JSTOR initially include in its database only issues of journals published before 1990? Why were current issues excluded? In making the decision to focus on back issues, we knew that we were swimming against the proverbial tide and “challenging marketplace solutions.” Consistent with Willy Sutton’s explanation of why it was banks that he chose to rob, most publishers and other vendors have been interested primarily in current issues because “that’s where the money is.” Current issues generate revenue streams; back issues rarely do.

Since it is “contrarian,” our emphasis on back issues has surprised many people. Early on, we explained our plans to the head of one widely known commercial enterprise, who was quick to comment, “No sane person would do what you propose.” We were undeterred. We thought that we had an opportunity, and perhaps even an obligation, to make up-front investments that could have long-term social value for the scholarly community at large. Unlike commercial entities, the test of success for us is how well we facilitate teaching and scholarship by improving the mechanisms of scholarly communication.

At the same time, we recognize that such broad statements of good intentions often mean little — as one of my friends likes to put it, “good intentions randomize behavior.” Fiscal discipline is needed, and we have always believed that JSTOR would have to be self-sustaining eventually. Perpetual subsidy is both unrealistic and unwise; projects of this kind must make economic sense once they are up and running. If users and beneficiaries, broadly defined, are unwilling to cover the costs, one should wonder about the utility of the enterprise. In this important respect, we are strong believers in market-place solutions — provided that what the economist calls “externalities” can be captured.

Given our objectives, there were strong pragmatic reasons for focusing on back issues. After all, they comprise that part of the journal literature that is least readily accessible, most in need of preservation, and most avaricious in its consumption of stack space (our ten journals run to some 750,000 pages). Also, from our perspective, the fact that back files do not generate much revenue for publishers has been a plus, not a minus. To launch this project, we needed to obtain copyright permissions, and we knew that this would be much easier if the project offered no threat to basic revenue streams. We wanted to work in concert with publishers as well as with libraries, to advance the common interests of both. It is only by adopting what might be called a “system-wide” perspective, which recognizes the legitimate needs of both the providers of scholarly materials and their users, that socially optimal arrangements can be put in place.

Let me now summarize a great many intervening developments in relatively few words. First, following much discussion, we were able to obtain copyright permissions from all ten journals to bitmap their back issues. Second, we arranged for colleagues at the University of Michigan (working closely with Ira Fuchs, Chief Scientist of JSTOR) to oversee the technical aspects of the project. Third, we selected several liberal arts colleges to serve as test sites along with the University of Michigan. If all goes according to plan, the full back files of four of the ten journals will be accessible at these test sites very shortly, and the remaining journals will be available by the end of the year.

Characteristics of the Archival “Product” (or “Scholarly Tool”)

In getting to this point, we made several key decisions about the archival “product” that deserve to be highlighted.

• First, we elected to set a high bitmapping standard of 600 dots per inch. The objective was to produce images of exceptional clarity, with every shading, every subscript in every equation, and every figure or photograph clearly visible. In short, we wanted to create something of archival quality, with print-outs so good that readers would regard these copies as equivalent to the originals — and, in many instances, superior to yellowed or soiled pages. This objective is being achieved, and we have sample pages for those of you interested in seeing them.

• Second, we decided to create and include in the electronic file a searchable table-of-contents index. This feature permits readers to obtain bibliographic references to all articles, book reviews, and other components of the database by a specified author, containing one or more keywords in their title (such as “Council of Economic Advisors” or “human capital”), published in a certain journal, or in a particular set of years — and then to call up the articles themselves on the screen.

• Third, although the user will see only page images on the screen, this database is linked to a text file created using Optical Character Recognition (OCR) software that will enable users to search not only the table-of-contents index but also the actual text pages of all the journals. Specifications provide that the OCR-version of the text will be accurate at the 99.95 percent level; in other words, it will contain no more than one error in 2,000 characters.

• Fourth, our colleagues at Michigan have joined these elements to a sophisticated search engine that allows the entire database to be accessed online, via Internet connections, by authorized users. This online feature has obvious advantages, including both ease of access and the minimization of storage problems for libraries (since the database is stored centrally). The software also includes printer applications for MACs and PCs which permit authorized users to print exceptionally high resolution copies of pages or entire articles at their terminals.

In seeking to translate these concepts into actual electronic files, progress has been slower than envisioned in our original schedule — which, we admit, was aggressive, to say the least. The firm responsible for the bitmapping and OCR work and our friends at Michigan responsible for technical aspects of the project have encountered all (it seems like all) of the inevitable problems associated with creating any new tool of this complexity from scratch. Still, as I have said, we expect the back files of four of the ten journals to be up and running at the test sites very shortly. By now there has been enough preliminary experience with parts of the database, and enough “demos,” to convince us that JSTOR is real, and to give solid grounds for believing that it will perform as advertised.

Linking Current Issues to the Archive

In the last few months, as work on the archive of back issues has proceeded, we have given a great deal of thought to the place of current issues in this project. Here again we decided to start out with a highly specific application from which it might be possible to generalize. Fortunately, the Ecological Society of America (ESA) expressed strong interest in being our initial partner in this undertaking, and the Society and JSTOR have now signed an agreement that provides for linking electronic formats of Current Issues of the Society’s publications (including their main journal, Ecology) to complete electronic archives of these same journals.

In principle, such an approach has great appeal. From the standpoint of users, there are obvious advantages in being able to search simultaneously the entire run of a journal, from the most recent issues back to the earliest ones, using common software. For both users and libraries, this approach would solve the problem of how the historical archive is to be updated, since updating would occur annually and automatically if linking were achieved. Also, some libraries might conclude that it was unnecessary to retain hard copies of current issues at their main location (or even to receive them in the first place), thereby achieving considerable savings in binding and processing costs, as well as relieving them of the burdens of storing, preserving, and retrieving paper copies of these journals from what can be expensive stack space.

Attractive as the basic approach may be, seeking to achieve this kind of linkage raises issues that are both technical and economic. Putting the technical issues to one side (since, in any case, I am not competent to discuss them), the economic challenge is to forge a structure and a plan that will simultaneously: (a) provide faculty, students, and other interested readers with dramatically improved access to journal literature at fair prices; (b) protect the publisher’s revenue streams and membership rolls; (c) help control the operating and capital costs of libraries; and (d) generate sufficient revenue to allow JSTOR to create, maintain, and update the database and the controlling software over the long run. The pricing and other issues implicit in this formulation would have to be faced even if we continued to work only with back issues; it is better, we are now persuaded, to confront them directly in the context of linking current issues to back issues, since the advantages of this approach are so manifest.

After much discussion, we decided to focus on creating and distributing two distinct but complementary products: an electronic version of “Current Issues” (perhaps those published within the last calendar or fiscal year, but the cut-off could also be pushed back to encompass the last two or three years); and an electronic “Archive” of all other issues. Under the arrangements we have been considering, the ESA, as publisher, would continue to be responsible for editing and publishing current issues and for determining the subscription prices and other terms (including restrictions on access) associated with sales of subscriptions to libraries and individuals. The ESA would also provide JSTOR with electronic source materials for Current Issues. JSTOR, in turn, would: (a) cover many of the up-front costs of launching this venture; (b) create an historical archive according to the specifications stated earlier, which JSTOR would make available at appropriate fees to libraries and individuals; and (c) develop an electronic mechanism for linking Current Issues to the Archive and updating the Archive each year.

No decisions have been made concerning pricing and controls over access, but I can outline our provisional thinking:

• The ESA might offer library and individual subscribers the electronic version of Current Issues at roughly the same price it now charges for the paper version (assuming that subscribers chose to take one or the other — there could be a substantial discount for subscribing to both). The ESA will have to decide whether it wishes to offer “site licenses” (which would permit the institution to make the database available on a campus network) or only “library licenses” (which would restrict access to locations within the library system and perhaps in departmental offices, as well).

• With regard to the Archive, we anticipate that JSTOR would ask libraries to make two types of payment for access: (a) a one-time capital charge for acquisition of permanent rights to the base Archive; and (b) an annual fee to cover the recurring costs of maintaining and updating the database and associated software. The one-time charge would be related to the amount of material in the Archive and would be significantly lower than the costs associated with storing and maintaining the equivalent set of materials in paper. Hypothetically, this charge might be set at, say, a penny or two a page. The annual charge would probably also be related to the size of the base Archive and to whether the library wanted new material added each year. It might be in the $25-$35 range for a journal such as Ecology and would certainly be significantly less than the recurring costs of processing, binding, storing, and retrieving new as well as old issues. Alternatively, JSTOR might offer libraries the option of combining the capital and annual charges into a single stream of “lease payments.”

• We anticipate that JSTOR would also offer individual subscribers direct access to the Archive (so that they would not have to depend on their home institution making access available on a campus network). Charges to individuals could be very modest (for example, perhaps $10/year in the case of Ecology). Every dollar contributed by individuals would represent additional revenues for “the system” and would lessen the financial burdens on publishers and libraries, thereby reducing the risk that electronic publication might lead to substantial cost-shifting from individual subscribers to institutions.

I want to emphasize again that these are initial thoughts and purely illustrative numbers. No commitments or decisions have been made by anyone. But I thought it would be easier for you to think with us about the prospects for JSTOR if we gave you some hypothetical reference points that had dollar signs attached to them.

Could this kind of model work economically? What levels of charges would be required for JSTOR to break even? Since there are tremendous economies of scale involved in creating and distributing a new electronic tool of this kind, much depends on the interest demonstrated by libraries and individuals. Our guess (it is nothing more than that) is that at least half of the current subscribers — libraries and individuals — would elect to participate. Under that assumption, and making another set of guesses concerning costs, we think that the project is viable. But no one can know for sure. The proposed collaboration between JSTOR and ESA is avowedly an experiment — the first of its kind, we believe — and much fine-tuning will surely be required, along with a willingness by all parties to admit mistakes and make adjustments on the basis of experience.

As at least partial protection against the risks associated with ignorance, we have developed a mechanism that should provide reasonable safeguards for all concerned. Our starting point is the concept of a partnership. Since the proposed experiment is a novel one, and no one can predict with confidence either costs or revenue streams, we are contemplating an “upside” sharing of any “Net Funds Available” that remain after costs have been covered. Also, while experience is so limited (nonexistent, really), we plan to provide a “downside” protection for the ESA and perhaps a small number of other publishers that choose to cast their lot with JSTOR in its earliest days. JSTOR might guarantee a publisher such as ESA that, for a period of two years, its gross subscription revenue would not fall below an agreed base level because of any declines in individual subscriptions attributable to the experiment.

The “upside” sharing might involve a commitment by JSTOR to rebate half of any “Net Funds Available” to the publisher and another one-quarter to participating libraries. JSTOR might itself retain the remaining one-quarter for development purposes. By committing itself in advance to give rebates to both publishers and libraries out of any “Net Funds Available,” JSTOR protects everyone against the risk that any one party would benefit disproportionately. At the same time, since it is understood that JSTOR must be self-sustaining, charges would have to be renegotiated if JSTOR’s revenues were insufficient to cover its costs. We believe the partnership concept to be the best — and perhaps only — way of handling the uncertainties inherent in a new initiative of this kind.

As a nonprofit entity, JSTOR is not permitted by law to accumulate funds beyond its legitimate needs, and, in any case, would have no interest in doing so. Its sole purpose is to serve the scholarly community by increasing the availability of scholarly journals and enhancing their usefulness, while concurrently reducing library costs. For this reason, the posture of JSTOR differs fundamentally from that of commercial vendors. In creating the founding Board of Trustees of JSTOR, care was taken to assemble a group of individuals who would be known and trusted by the scholarly community — in part to assure credibility from the outset.

Institutional Subscribers: Libraries and Campus Networks

Let us now consider in more detail how the economics of JSTOR might work for institutional subscribers. Conversations with a number of presidents, provosts, and librarians have persuaded us that there is considerable interest in the historical Archive (setting aside, for the moment, the question of linkages to Current Issues). We anticipate that, as one person put it, faculty and students will find JSTOR “irresistible” and will urge their institutions to acquire this resource. In addition to creating new search capacities, JSTOR solves one of the most vexing problems faced by readers and librarians alike: the unavailability of a particular volume because someone else is using it, because it has been misfiled, because the library is closed, and so on. JSTOR provides predictable access to multiple users at every hour of every day (assuming the continuing cooperation of the gods of cyberspace).

The positive case for acquiring this valuable new scholarly resource will be convincing across a broad spectrum of institutions, including:

(1) large research universities with faculty and graduate students who make considerable use of back issues and expect to have at their disposal the most up-to-date mode of access, even though their library may already have all of the back issues in paper format available in open stacks;

(2) libraries using a closed-stack system that requires the retrieval and reshelving of large numbers of journals by staff members; and

(3) the many colleges and universities, in this country and abroad, that lack complete files of back issues in paper format.

For this last set of institutions, JSTOR offers a unique opportunity to acquire complete runs of journals, in pristine condition, for a small fraction (perhaps 10 percent) of the cost of acquiring paper copies — and without having to incur all the associated costs of building and maintaining library space. To illustrate, if JSTOR were able to offer the Archive at a one-time capital charge of as little as one or two cents per page, an institution with limited library resources could obtain the entire back file of Ecology, from 1921 through 1994, for something like $750 to $1,500 (there are 77,000 pages). What a bargain!

In other situations, where the back issues of Ecology are already available in paper format, JSTOR allows the library to save valuable shelf space on the campus by moving the back issues off campus or, in some instances, by discarding the paper issues altogether. The natural fear of “losing something” will be minimized if, as we expect, arrangements are made for regional collections of paper copies. As another “fail-safe” mechanism, JSTOR would be able to provide CD-ROMs.

What are these savings in storage costs worth? Much depends, of course, on local circumstances. Utilizing a methodology devised by Malcolm Getz and cost data assembled by Michael Cooper, we estimate that the one-time capital costs for storing a single volume of a journal in open stacks, excluding the cost of land (hardly negligible for libraries in major metropolitan areas, such as New York), to be anywhere from $24 to $41. It follows that the capital costs associated with storing the complete 76-volume run of Ecology could range from $1,800 to $3,100. This cost is roughly twice the hypothetical capital charge for both acquisition and storage of the JSTOR electronic Archive. One key reason for this large disparity is that JSTOR centralizes the storage function at off-site locations while retaining all the advantages of browsing and providing readers with more or less instant access.

There also appears to be large savings in operating costs associated with the “circulation” function — retrieving and then reshelving paper copies of journals. The New York Public Library (NYPL), which is a non-circulating closed-stack library, reports that it spends an average of $1.94 to retrieve and reshelve a journal volume. We recognize that costs in New York will be higher than average and that closed-stack libraries are more expensive to operate from this standpoint than open-stack libraries. Michael Cooper has estimated the staff-only costs of circulation from open stacks in the University of California-San Diego library to be roughly $.60 per volume. Of course, as Cooper, Getz, and others who have written on this subject recognize, open-stack systems save operating costs by, in effect, relying on users to do much of the “leg work” involved in retrieving volumes. The time-costs for users are of course considerable, and dwarf staff-only costs.

We cannot hope today to be at all precise in suggesting the right figures, and we certainly do not want to exaggerate the potential savings; but it seems reasonable (actually very conservative) to use $1.00/retrieval as a reference point — half the actual cost incurred by NYPL. The implications of using even this modest figure are staggering. As part of the research associated with the introduction of JSTOR, staff at the University of Michigan have tallied usage on their campus of the ten economics and history journals in their paper formats. Also, we have surveyed experiences at the five college test sites. Weighting each institution equally (and thus giving the much heavier usage at Michigan only one-sixth the overall weight), we find that library users request volumes from each of these core journals an average of 45 times per year (the comparable average at the University of Michigan considered separately is 180 times per year). Using the cost figure of $1 per volume, the annual cost of circulation is estimated to average $45 per journal for this set of six schools and $180 per journal at the University of Michigan. And these estimates should probably be doubled for libraries which operate like The New York Public Library.

In addition, use of paper copies involves recurring costs of preservation and conservation. The way library patrons use this literature, often spreading bound volumes on copy machines, subjects the paper to substantial wear and tear. Older journals must be mended, deacidified, and sometimes rebound. While electronic archives are not subject to these “handling” costs, they do entail costs of other kinds — especially maintaining computer equipment and answering questions posed by users. We expect to estimate these expenditures at our test sites this fall, but we find it hard to imagine that they can be close to comparable to the operating costs associated with paper formats.

Libraries participating in JSTOR might be able to lower their on-going costs even more if the historical Archive were linked to an electronic version of Current Issues. Acquisition and processing costs, estimated by one university librarian to be more than $20 per journal, would be substantially reduced. Also, it would no longer be necessary to bind individual issues into volumes, a process that can cost anywhere from $15 to $30 depending upon how one allocates the staff costs associated with preparing the journals for the bindery.

We recognize that we are enthusiasts, and no doubt our excitement about our “child” should be discounted somewhat. Still, the purely economic case for JSTOR, seen from the perspective of the institution, seems overwhelming. The highly favorable economics are driven by powerful scale effects, since JSTOR permits tasks (especially storage and retrieval) that are now done repetitively at thousands of libraries to be done once — centrally — and to be done far more effectively, as well as at much lower cost. Combining capital and operating costs, and expressing both on an annual basis, we estimate that the continuing costs of storing, retrieving, conserving, processing, and binding one journal, such as Ecology, in paper format are at least twice (and perhaps three times) greater than what they would be if the library contracted with JSTOR to provide what we are convinced would be far superior access.

Individual Subscribers

How would an individual subscriber gain access to either the electronic version of Current Issues or to the electronic Archive? Our presumption is that, in both instances, individual subscribers would be given PIN numbers which would entitle them to access the database from any location. “Personal” access would be restricted to those individuals who had personal identification numbers. Of course, those individuals who work on campuses that have purchased site licenses for the Archive and/or for Current Issues might have a reduced incentive to take out individual subscriptions. They would no longer have to trek to the library to browse; they could use the campus network.

This “fact of life” in an electronic world is a legitimate source of concern to the publishers of journals with significant numbers of individual subscribers and to professional associations that link membership to the provision of journals. At the same time, it is easy to exaggerate the risks. Apart from the power of inertia (not to be underestimated), publishers need to recognize that by no means will all individual subscribers have ready access to JSTOR files by means of a campus network. Many will not. In addition, publishers could consider giving individual subscribers “something extra,” and a number of creative ideas have already been proposed.

At present, of course, individual subscribers have no personal access to any historical Archive except in those situations in which they happen to have accumulated an entire back run of a journal in their office or home. (And, as one colleague observed, it should be easy to demonstrate that a full run of even a few core journals in a field such as economics or history could “bury someone alive in his/her office — perhaps even before the person got tenure.”) Since, in almost all instances, individuals must go to the library to peruse back issues, we expect the availability of an electronic Archive to be exceedingly attractive to a number of individual faculty members, who might well be willing to pay a modest annual charge (say, $10) for personal access.

We also suspect that there will be strong synergies between purchase of electronic access to Current Issues and purchase of electronic access to the Archive — which would be regularly updated if the individual subscribed to both. For this reason, demand for individual subscriptions might actually increase, rather than decrease, as a result of JSTOR. If it proves possible to include several core journals from a field (or sub-field) in JSTOR, the appeal of electronic access to a “collection” might be greater yet. We just do not know anything, at this stage, about the elasticity of demand for what would be a totally new scholarly tool. This is an important subject for future research, as experience begins to accumulate. Also to repeat the presence of such large unknowns is strong justification for the “partnership” concept, whereby the interests of publishers, libraries, and JSTOR are aligned.

Broader Considerations

The behavioral consequences of JSTOR (and other initiatives of a similar kind) also deserve careful thought. One question of great interest is how JSTOR will affect patterns of research and teaching. Our assumption is that dramatic improvements in ease of access, combined with the advantages of a powerful search engine, will cause students and faculty alike to make more use of back issues than they do today. We are told, and anecdotes abound, that students sometimes fill papers with what they perceive to be their quota of citations by relying heavily on recent literature, since Current Issues of journals are easier to find and use than back issues. Yet it would surely be better if all of us — students, faculty and research staff alike — were able to mine the full range of scholarly literature with less difficulty than present arrangements impose. For instance, one would hope that students writing about the end of World War II would consult articles written by scholars in the immediate postwar years to gain a sense of how the world looked then, rather than relying on more recent compilations of opinion. Similarly, current discussions of welfare reform might be helped by more study of writings in the 1930s about the New Deal.

Thinking about JSTOR also causes this observer to consider a number of aspects of budgeting and decision-making in higher education:

• First, JSTOR emphasizes the importance of considering various budget categories together when making allocation decisions. Specifically, capital costs, which are often neglected by colleges and universities, must be taken into account in calculating the net effects of subscribing to JSTOR. For this reason, it is important that provosts and others with campus-wide budgetary responsibilities be involved in deciding whether an institution should (or should not) start down the JSTOR path. Compartmentalized decision-making, focused solely on acquisition budgets, for example, would not permit a proper assessment of the costs and benefits of JSTOR.

• Second, JSTOR reminds us that too narrow and compartmentalized a view of economic self-interest can be harmful in other ways. For example, if libraries refuse to worry about the effects of their subscription practices (such as purchasing site licenses through consortia) on the number of individual and library subscribers, they may inadvertently inflict considerable economic harm on publishers — who may in turn respond by raising the prices of library subscriptions.

• Third, JSTOR also warns us that attempting to charge for journal literature “by the drink” can lead to results that are far from optimal socially. As the just-cited article by Lieberman, Noll, and Steinmueller emphasizes, the marginal costs of including another user are exceedingly modest, and this is one reason why it is so important that the product be offered on a subscription basis. It would be undesirable, from the standpoint of resource allocation, to discourage an impecunious student from using JSTOR because of a per-use pricing model.

• Fourth and last, JSTOR is unusual in that it violates the usual axiom about the importance of forcing choices between “more” and “cheaper.” Let me explain. When new technologies evolve, they offer benefits that can be enjoyed either in the form of more output (including opportunities for scholars to do new things or to do them better) or in the form of cost savings. It is my experience that in universities electronic technologies have almost always led to greater output, and rarely to reduced costs. Yet, it is imperative, if tight resource constraints are to be observed, as they must be, that technological gains lead to at least some cost savings. I make this speech regularly to my colleagues at the Foundation, as well as to my one-time colleagues in academia. In the case of JSTOR, however, it is hard to press for this version of the “discipline of choice” because JSTOR offers both great advantages to potential users and cost savings.

My puritanical tendencies cause me to worry about the potentially debilitating effects of such an unlikely product on the will to choose; but I will acknowledge that it is also nice, if only occasionally, to be confronted with an opportunity to argue on behalf of an innovation that is both better and cheaper. At least that is what I hope that our experimentation with JSTOR will demonstrate to be the case.

As a wise friend of mine said on another occasion (when contemplating the impeachment of Richard Nixon, to give the exact situation), “We’ll know more later.” So we will, and the JSTOR contingent will undertake to keep all interested parties posted on what we learn (including what mistakes we make). At the same time, we encourage you to give us your suggestions and best thinking as all of us contemplate the murky but exciting future of scholarly communication in an electronic age.

APPENDIX

TABLE A

The Costs of Storing, Retrieving, Conserving, Processing, and Binding Paper Journals

Assumptions

Typical core journal: Ecology

First year of publication: 1921; Number of Volumes: 77

Cost Description for Ecology Annual Cost

Costs associated with Back Issues:

Storage (@$3.07 per volume) [1] $236

Circulation (average usage of 45 requests per year) [2] $ 45

Conservation and preservation [3] $ 5

Annual costs for the archived journal literature $286

Costs associated with Current Issues:

Processing and check in [3] $ 20

Binding [3] $ 20

Circulation [4] $ 45

Annual cost for Current Issues $ 85

Total annual costs $371

[1] Combining Getz’s methodology and Cooper’s construction costs from four different locations in California gives a range of storage costs of $21 to $41, an average of approximately $30. Using Getz’ amortization assumptions of a 25-year life and a 7% discount rate yields an annual cost of $2.57. Adding Getz’ annual maintenance and utilities costs of $5 per square foot adds $0.50 per volume, for a total of $3.07.

[2] For pre-1990 issues, from a survey of six libraries participating in JSTOR. The $1 per retrieval assumption is explained in the text.

[3] Figures for conservation, processing and binding are estimates based on informal discussions with librarians.

[4] This figure for the circulation costs for Current Issues is nothing more than a guess. Our assumption is that there are just as many requests for issues from the most recent five years as for the many years of issues prior to 1990.