One of the bedrock principles of JSTOR's mission is to serve as a trusted archive of electronic content. When we use the term "archive," we refer to the task of preserving and providing access to a collection of scholarly literature over time, without regard for how frequently these materials are being read or used. In that sense, our definition assumes that there is some portion, perhaps even a significant portion, of material that is not economically self-supporting in a commercial sense. While JSTOR's initial activity has centered around converting previously published paper content to electronic formats and then archiving them, our mission has always also included the commitment to archive future issues of journals that are published electronically. Because of that commitment, JSTOR represents one of the first tests of how the scholarly community may support electronic archives. The purpose of this article is to share early lessons and some of the questions that are surfacing from that "test."

As the scholarly community makes a transition to disseminating information electronically, are the responsibilities for archiving also making the transition? There is not yet a system in place to protect the electronic literature being published today. How can we be sure that such a system will evolve? Where will the resources come from to support it on an ongoing basis? This is just the beginning of a long list of questions that point to challenges that lie ahead if archiving is going to make the transition to the digital age.

Existing Archiving System

To understand this transition, we must familiarize ourselves with the existing archiving system and how it evolved in order to assess possible future archiving solutions. One important characteristic of the system we rely on today for paper is that it was not developed through a proactive effort to provide an archive; rather, it has been primarily a by-product of the need to provide access. If you want to read a paper document, you must have it in your hands, and for that to happen the place where it is stored matters a great deal. So the existing system evolved through countless local acquisition decisions combined with librarians' commitment to preserve materials once acquired. In addition, the infrastructure required to preserve those materials - the buildings, the shelves, and the associated maintenance costs - have been paid for and justified for a variety of reasons that have nothing to do with archiving per se. For example, the library was and is an important "place," and the construction of large structures to house the accumulated knowledge lends credibility and prestige to the institutions that maintain them. Further, funding streams can be tapped, including fund-raising that can associate donors with parts or all of the physical structure, that can help support these edifices of information. Once the huge fixed investments are made, the very small marginal cost of adding a volume to the physical collections supports the continuation of this pattern until, of course, a library runs out of space, at which time the effort to raise capital funds begins anew.

A Time of Transition

What happens when access is unbundled from geography and ownership? Information can reside on one server and reach any authorized user, anywhere in the world, who has a connection to the Internet. In such an environment, it does not make sense to duplicate the existing model, with its emphasis on local ownership. Individual colleges and universities are not demanding that content providers deliver to them the electronic materials so that they can build huge server and disk storage infrastructure on campus to store, maintain, and deliver the information. So, as documents make the digital transition, we are leaving the old model behind, without any replacement approach in operation.

Lessons from JSTOR

JSTOR's experience over the last six years indicates that many of the financial systems and structures in place at our academic institutions present challenges to the implementation of a more centralized archival approach that is required by the technological transition described above. We have studied the capital, operating and maintenance costs associated with storing paper journals and compared them with the costs of JSTOR participation. There is no doubt that the costs for an individual library to participate in JSTOR are substantially less - for smaller teaching institutions, more than ten times less - than the costs associated with storing and providing access to the paper journals. However, based on the feedback received from these libraries, the primary reason for JSTOR participation seems to be not the potential savings associated with central archiving but rather the benefits associated with providing better and more convenient access to the literature for faculty and students.

Justifying JSTOR on the basis of the benefits of enhanced access as opposed to the potential savings associated with archiving is not in itself a bad thing, nor is it an "either/or" situation. Both benefits are valuable. But for many libraries, the focus on access seems to overwhelm the interest in archiving. This is understandable. For one thing, JSTOR is an extremely young organization, and libraries are appropriately conservative. They are not going to move quickly to realize the shelf space benefits soon after they sign up. They will want JSTOR to prove itself first. Further, the focus on access is consistent with the point made earlier: the primary motivation for an acquisition is its current usefulness, not its archival or future value to scholarship. Finally, and perhaps most important, librarians focus on the access benefits of JSTOR because, from a practical standpoint, this is all that the existing organizational and financial structures allow them to do. The benefits of newly available shelf space do not appear in any budget under their control. That budget resides somewhere else, such as under the control of the provost or president. Even for the provost, if the space available to the library is sufficient at present and for the near future, the economic savings is not a foremost consideration. It will be, though, when the time comes to build the library addition.

In many, if not most cases, the only budget that the librarian can use to pay for an archive like JSTOR is the acquisitions budget. Assuming that the account used to purchase an item is related to the purpose of the purchase, reliance on the acquisitions budget to pay for JSTOR makes it a database, not an archive. In other words, by paying for JSTOR from the acquisitions budget, rather than from a combination of the acquisitions budget and funds that would normally be used to build new infrastructure for storage and archiving, institutions will have more difficulty recognizing, in economic terms, the capital cost savings associated with JSTOR participation. Existing financial structures make it hard to take all costs into account.

Who is in charge of archiving at academic institutions? What budget in higher education institutions is dedicated to preserving this literature in the future? If electronic access is no longer tied to geography, and there are no local motivations for supporting an archive, how can we ensure that electronic archives will be established and maintained? Are institutions going to be willing to contribute to centrally managed archives of important materials when the goal is specifically preservation, not access?

Conclusion

The challenge that lies ahead, if little-used electronic materials are to be preserved for future scholars and students, is to address directly the motivations for investing in the resources to perform this function. How will institutions justify the investments in supporting or providing a centrally-held electronic archive when there is no named building to point to, no beautiful library to show prospective students, no addition to the volume count that leads to a ranking among the top libraries? Will colleges and universities be able to justify investments in archiving for its own sake?

If there is to be a transition from paper to electronic publishing, the archiving challenge will have to be addressed. This is not a problem that is going to be solved through a complex array of disconnected local decisions. Because electronic publication by its very nature depends on centrally held resources distributed widely through communications networks, inevitably some parties will be "providers" of an archive while others will be "beneficiaries" or "users." We must determine ways to motivate the beneficiaries of the archive to support the costs of the providers. JSTOR will continue to test the question of whether support can be generated for a not-for-profit organization with archiving as its fundamental mission. But there are and will be other examples. One is The Andrew W. Mellon Foundation's Electronic Journal Archiving Initiative, which has provided support to seven universities to develop plans for archiving electronic content. These universities are working with a variety of publishers of scholarly content to test the technological, legal, and economic questions associated with archiving electronic content in a practical way.

This project and others like it will contribute to our understanding of what will be needed to preserve electronic documents. We hope they will also shed light on what motivates colleges and universities to protect and preserve scholarly literature for future users. The local motivations that have been the foundation of the current paper archive do not naturally generate the scale of resources that will be required to establish the more centralized model necessary for the preservation of electronic documents. There are examples, especially in the larger collecting institutions, of decisions made solely for the purpose of the long-term maintenance of the scholarly record. In those institutions, regular funding streams and even departments (e.g., preservation and conservation) have been established. The government, and in particular the National Institutes of Health, played a critical role in these cases through programs such as the Brittle Books program. It may be that government will have to play a role here as well. Alternatively, the kind of transformation that has happened in the financial systems and procedures related to computer technology to accommodate the transition from a mainframe-dominated to distributed computer infrastructure, may provide some clues to the path ahead. These are but a few examples, but they demonstrate the kind of budgetary recognition and allocation that will be necessary if the problems of electronic archiving are to be overcome.

If the documents of today are to be preserved for tomorrow, the budgets of academic and cultural institutions may need a new line item in the digital age, a line item that transcends existing capital and operating budgets and addresses all the costs of maintaining the electronic record; perhaps it should be called "e-archiving." Fortunately, one major benefit of the advances in information technology is that, from both a system-wide and an individual organization's perspective, if costs can be spread over a significant number of institutions, an e-archiving budgetary commitment can be both less expensive and more effective than the regular investments already being made in the maintenance of the printed record.

This is an abridged version of an article by Kevin Guthrie in Educause Review, November/December 2001, volume 35, Number 6.

Last updated on September 8, 2006