Libraries have always been expensive to operate. Acquiring, managing and storing large collections of print books and bound journals, done on any significant scale, have substantial associated costs. But one of the fuzzier aspects of cost analysis for a library has involved determining the ongoing, annual expense of keeping items in the “warehouse.”

Physical storage is not free. Shelf space has a measurable value, as do the associated support functions of climate control, lighting and inventory tracking. But those indirect costs have always been “below the line” for library operating and acquisitions budgets. Furthermore, physical collections have significant inertia in that they persist in place unless concerted action is taken to remove materials on a regular basis. The upper bound on any given physical collection is determined by square footage and linear feet of shelving in a library building and its associated auxiliary storage facilities.

In the digital world, libraries face an entirely different situation. Each year we license enormous amounts of digital information for academic purposes, much of it at great expense.

A quick look at Falvey’s lists of databases and electronic journals makes the scale of this investment quite clear. These digital materials are very heavily used, so that the per use cost and per capita cost of this material are well justified. Two quick statistics demonstrate this: There were 1.9 million licensed e-journal articles accessed by Villanova users in academic year ’07-08, and 1.6 million database searches completed by Villanova users during that same year.

With some notable exceptions (for example, JSTOR journals), acquisition of digital materials does not come with any guarantee of persistent “storage.” These digital materials don’t automatically end up on shelves in our buildings or on disk drives in our data centers. Of course, most of our content licenses for such materials include assurances of permanent access.

However, those assurances have not often been tested in practice. Big questions remain regarding how libraries should commit to sustaining collections that remain, in effect, in commercial hands. The recent announcement of an independent digital archiving initiative, the HathiTrust, from the libraries participating in the Google Books project, is one indicator that library professionals continue to feel collective unease about the idea that the long-term preservation of digital library holdings should be dependent upon the business plans — or philanthropic commitments — of for-profit corporations.

This reality is driving a move toward the development of digital storage and preservation strategies, some of them voluntary and relatively economical, while others are organized by new non-profit agencies and consortia and are distressingly costly. There’s a national study underway, under the aegis of the National Science Foundation and the Mellon Foundation, to develop an economic sustainability model for digital preservation.

At this point, however, it is only the most costly solutions that promise true archival durability of digital content. And, here’s where cost comes back into the picture, “above the line” for library operating budgets rather than below it, as in the case of physical storage in a building.

At a minimum, new entities such as Portico and the LOCKSS Alliance impose additional annual subscription fees on libraries for what amount to storage insurance policies for licensing digital resources. In the case of LOCKSS, the costs are modest but the system is in part dependent upon large-scale voluntary participation by libraries. With Portico, the costs amount to a surcharge on top of annual content investments, which can result in substantial new ongoing expenses.

Perhaps even more alarming than these potential surcharges for storage of licensed content are the emerging cost models for developing “dark archives” that will guarantee persistence of access to locally-created digital content, such as in our own Digital Library. Though in the early stages of business development for this important preservation arena, this is a challenging economic problem, as outlined by Brian Lavoie is his March / April 2008 D-Lib article, “The Fifth Blackbird: Some Thoughts on Economically Sustainable Digital Preservation.”

For some sense of actual costs in the present environment, the OCLC Digital Preservation service now operational would result in costs to Falvey of $23k per year to handle our digital library content. At present, we have online about 3 terabytes (costs approximate $7500 per terabyte), with the collection projected to grow at about a terabyte per year. There are no less expensive alternatives that provide full preservation and back-up guarantees.

In this new environment, only one thing is certain: library storage costs will not go away in the digital age. These costs may diminish on a per unit basis over time, but will continue for the foreseeable future to impose an economic burden on all libraries that take their preservation mission to heart.

Photograph by John Welsh