Patron-Driven Re-Acquisition

At first blush, a relationship between deselected print monographs and patron-driven acquisition (PDA) is far from obvious. In a typical deselection project, the prime candidates are books that have not circulated for 10-20 years. Recent imprints are excluded since they have not had enough time to prove themselves. In short, withdrawal candidates are the books patrons don't want. 

PDA titles from the major eBook aggregators, on the other hand, are intended to offer rich discovery and evaluation metadata for books that patrons might want. When the record set is further refined by application of an approval profile, the result should be books that patrons are most likely to want, given the curriculum and research interests of the institution they serve. Best case, those criteria might be further enhanced by analysis of historical purchase and use patterns.

And never the twain shall meet. Well...not so fast.

The prospect (never mind the act) of removing books from library shelves makes everyone nervous.This is a good thing, until we as a community are certain that all content is archived to our satisfaction. But for collections and public services librarians, there is also that primal belief that it would be better to have immediate access to a withdrawn title, just in case a patron ever wanted it. For most titles, this need will be accommodated by shared print storage agreements, or inter-library loan.

But there are other options. In cases where a Hathi Trust public domain title is available, that offers essentially a free online replacement. For that availability to be made clear and convenient, the Hathi URL is simply added to the cataloging record for the print version. This can be done simultaneously with the removal of the item record for the print version being withdrawn. In this example, that converts the library's catalog record to a do-it-yourself PDA record.

Clearly, the same principle can be applied to any eBook provider with PDA capability. Since the bibliographic, item, and holdings records for a withdrawn book have to be changed or removed anyway, the insertion of another URL would add almost no time to the process. Instead of suppressing or deleting a record for a withdrawn title, the library converts it to a PDA record, retaining its discoverability, but with a different delivery option. That option might be one-click electronic access, short-term rental, or re-acquisition through purchase (in addition, of course, to calls from storage or ILL).

By combining this record conversion with print deselection, the shelf space is freed, and the cost of maintaining an unused print version on the shelf is eliminated. But the potential for access remains intact, and in some cases actually improves. It's a solution that could satisfy both deselection and service objectives, a rare combination. Let's call it patron-driven re-acquisition (PDR). It's a technique that could work with eBooks, but also with print-on-demand (POD) providers. The key would be to have relevant URLs available at the time the withdrawal decisions are made, ready for batch matching and insertion into candidate file bib records.

At present, the degree of overlap between withdrawal candidates and titles available through PDA programs is probably miniscule. It would be interesting and useful to quantify that. But even if it's small now, that overlap will continue to grow.  The possibilities are quite interesting.

What We Need To Know

Box Lunches:   $42/each
Room Rental:   $250/day
Audiovisual:     $200/day
Coffee:             $75/gallon
Soda, Teas       $  5/each

Total cost:  more than you'd think

Informed conversation about monographs deselection: priceless

During ALA Midwinter in San Diego, R2 conducted three focus group sessions around "Sustainable Collections Services (SCS), the data-driven deselection tool we are developing. In all, R2/SCS heard from from 40 librarians, representing individual libraries of all sizes, as well as several consortia. A full summary of those discussions will be available shortly, but here are some excerpts, preceded by a glimpse of the Mad Men-style conference room in which two of the sessions were held:

SCS Use Scenarios
We asked attendees (a mix of directors, collections, and technical services people) to enumerate the ways in which a deselection project might first manifest itself in their institution. In our experience to date, this can vary widely, and has implications for project design and management. Grouped thematically, responses included:

• Space issues: the need to free space for a teaching center, learning commons, more room for students. In some instances, this becomes a "rapid response" scenario, where budget woes have scuttled existing expansion plans.

•  Joint consortial or regional action: As noted in a previous post, much effort is going into rationalizing shared print collections. Issues of duplication within the group and expression of archival commitments figure prominently here. For titles held in very small numbers, candidates for digitization might be identified. 

• Continuous deselection versus projects: several libraries noted that integrating deselection as a routine part of operations would be preferable to large-scale standalone projects. In some cases, it may be necessary to meter deselection activities to match the capacity of technical services units to perform necessary record maintenance.

•  Resistance to deselection: in cases where resistance is high, SCS could be used to demonstrate criteria and outcomes (to the title level) before taking any action. Use as an educational tool for stakeholders, when combined with educational presentations to faculty and staff, was seen as desirable. 

• Targeted collections or locations: Examples included the need to close a Biology branch, or tackle computer science books. 

• Project management and workflow design: While some libraries simply wanted access to data and the ability to run "if, then" lists, many recognized that they do not have staff or management capacity to handle additional work of this nature. There was strong interest in project management services, and perhaps consulting on workflow design related to record maintenance and disposition options.

Andy Breeding (l) and Rick Lugg (r) in full showman mode

At the outset, we described our work and results to date, focusing in particular on the Excel dashboard assembled for Grand Valley State University, one of our three partner libraries. In this action photo, apparently excerpted from surveillance video, you can practically feel the innovation!  Here some additional issues arose:

• Some libraries have immediate deselection needs, and may need to act before the "FDIC Layer" is fully in place.

•  Libraries want to define their own risk tolerance and criteria, and use SCS data to gauge the effects of various scenarios--which may vary by subject or material type. In the memorable words of one participant: "Build a tool, not a moral compass -- we'll provide that."

• Concerns were raised about misleading impressions caused by the use of Paul Courant's  $4.26 cost figures published in "On the Cost of Keeping A Book." Most believed that those figures were valid, but since this money cannot be directly recovered by deselection, great care must be taken in managing expectations of Provosts and University administrators.

•  Similarly, it is vital to have a clear plan for use of any space freed by deselection, and the funding lined up to convert that space.

On behalf of my partner Ruth Fischer, and our new SCS business partner Andy Breeding, a hearty thank you to the librarians who took the time to attend and to share their perspective on these important issues. The learning continues.

The FDIC Layer



L-R: Ivy Anderson (CDL and WEST); Bruce Hulse (WRLC); Aisha Harvey (Duke); Judy Russell (Florida); Bob Kieft (Occidental); Marie Waltz (CRL); Lizanne Payne (CRL and WEST); Rachel Frick (CLIR); Mark Sandler (CIC); Mark
Watson (Oregon/Orbis Cascade Alliance)                                                                                                                                                                    

One of the biggest fears related to deselection is the inadvertant loss of an important work from the scholarly record. This is a legitimate fear, as evidenced by the academic library community's efforts to prevent such occurences. The principle that no content shall disappear from the record has become the cornerstone of legacy print collections management. It takes a village to assure secure and accessible archives, especially when we drill down into what's required.

To paraphrase Ithaka's What To Withdraw framework (developed in relation to print journals), content security and accessibility is best assured by adopting multiple strategies:

• A secure digital archive
• An accessible digital surrogate
• A dark print archive, representing an agreed number of copies and conditions
• A light print archive, representing multiple copies distributed regionally

For monographs, the needs are very similar. But the infrastructure to support print archiving to these levels is still developing. Many talented people and relevant organizations are committed to developing this "FDIC layer", as I like to call it. The role of this FDIC layer is to provide overall archival security that allows individual libraries to act locally and independently, safe in the knowledge that the scholarly record is intact. At present, that security is incomplete, but growing:

• Hathi Trust provides a secure digital archive for more than 5 million monographs.

•  Hathi Trust public domain titles are freely viewable in full-text. Millions of other eBooks can be licensed to provide accessible digital surrogates.

•  A dark print archive for books does not yet exist, and it is possible that a completely dark archive may not be necessary, provided sufficient copies remain in the collective collection. There is some thought that monographs now in high-density storage may represent the beginnings of a print archive, but it is essential that retention commitments be standardized and disclosed clearly. Greater coordination is needed.

• A light print archive exists only informally at present. The distribution of copies throughout a consortium can be seen in shared catalogs. The distribution of copies throughout the collective collection can largely be seen through WorldCat holdings. Here again, disclosure of print archiving commitments and a coordinated approach are needed.

Historically, this level of preservation, and the security of the scholarly record, have been the province of large research libraries. Smaller libraries have counted on the larger libraries to perform these functions--to run the FDIC layer.  But increasingly, even the largest of these institutions, (even when acting in concert) cannot support the entire corpus and community by themselves. There is a need for every library to contribute to the integrity of the collective collection--by committing to archive specific titles and publicizing those commitments. The most promising avenue for such disclosure is use of the MARC 583 field in a WorldCat record, and work is underway to test the viability of this approach.


Meanwhile, organizations such as WEST (Western Regional Storage Trust), CRL (Center for Research Libraries), ASERL (Association of Southeastern Research Libraries), OhioLINK, the Center for Institutional Cooperation (CIC), ReCAP (Research Collections and Preservation Consortium), Ithaka Strategy + Research, and OCLC's Office of Research and numerous others have come together to plan and build a network of trusted repositories that will provide the foundation of this FDIC layer. A great deal of work has already been done under planning grants from the Institute for Museum and Library Studies (IMLS). WEST, in particular, has made enormous strides, with 89 libraries expressing interest in joining its efforts to coordinate print archiving and storage.

For the past few years, an ad hoc group of people interested in these issues has met during each ALA conference. The group has no official name or organizational locus. Its meetings are generously underwritten by LYRASIS, which also organized a national discussion on the future of print monographs collections in October 2010. Bob Kieft, College Librarian at Occidental College, has provided excellent leadership, facilitating discussion and information sharing among a wide range of voluntary participants. At the ALA Midwinter meeting in San Diego (pictured above), more than two dozen organizations were represented.

In addition to those named in the caption above, this session included: Ed Shreeves (Iowa); Karla Strieb (ARL); Bryan Skib (Michigan); Ross Housewright (ITHAKA S+R); Dennis Massey (OCLC Research); Sam Demas (Carleton College & MLAC); Julie Gammon (Univ of Akron/OhioLINK), Rick Lugg (R2); Sharon Farb (UCLA); Peggy Seiden (Swarthmore); Joni Blake (GWLA); John Berger (ASERL); Rick Clement (Utah State); Kim Armstrong (CIC); Tim Cherubini (Lyrasis). There are at least another 20 people regularly involved in these discussions.

These people, and others, are in the early stages of creating the archiving infrastructure necessary to safely and responsibly manage the drawdown of duplicative, low-use monograph collections. Creating this network of shared print (and digital) repositories--this FDIC layer-- is of critical importance. As a community, we are fortunate to have so many good people attending to this issue. The groundwork done by these "volunteers" will ultimately help rationalize the massive capital investment represented by print collections.

The Case of Bertrand Russell

The road to monographs deselection is lined with difficult decisions. Among the most vexing is the relative importance of use versus the inherent value of a work. In R2's first analyses of library circulation data, it has quickly become apparent that good books--even seminal books--are well represented among titles that have not circulated for many years. This points out one of the limitations of use as a metric for deselection. Use patterns are sometimes ugly, and often deeply unsettling, even to those of us who believe that data-driven deselection is the most rational basis for addressing the pressing problem of crowded and expensive-to-maintain stacks.

A case in point: Bertrand Russell. In one recent analysis for a small liberal arts college library, we determined that Russell's Philosophical Essays had not circulated in the past 24 years. The Stanford Encyclopedia of Philosophy identifies this 1910 Longman, Greens edition as a "major anthology of Russell's writings." Russell himself is "one of the most important logicians of the twentieth century", as well as a Nobel Prize winner (for Literature in 1950). Even the smallest academic library would want this titan of analytic philosophy represented in its collection.

And yet...no one has touched it in 24 years. This may be because these essays are included in more comprehensive works the library does hold. It may be because Russell is not taught. But the facts are clear: it is an important work that does not appear to be important to contemporary users. This may be a sorry comment on our culture; it may be an accident of timing and curriculum. We can't know for certain without further investigation. But let's assume for a moment that the situation is as described: Philosophical Essays by Bertrand Russell has not circulated in more than 24 years. What do we do?

If we decide that significance trumps use, we simply ignore the data and retain the book, along with our self-respect as an academic institution. This is a valid decision, and supportable to some degree. But what if there are thousands of worthy books that have not been used? How do we make room for more active titles...or for more users?

If, on the other hand, we decide to remove it, and a future logician demands access to this seminal work, what options exist to replace access? Here things get interesting, as there are several indicators of secure and convenient access:

1. A public-domain full-text electronic surrogate currently exists in Hathi Trust.
2. Within the library's statewide consortium, three libraries hold the exact edition, and dozens of libraries hold later editions.
3. WorldCat holdings data indicates 1,866 holdings in 174 different editions worldwide. Within the library's home state, 98 libraries hold one of the 174 editions.
4. A 2009 Routledge paperback edition exists as part of the Routledge Classics series.
5. Alibris for Libraries reports 103 copies currently available for sale.
6. Commercial eBook editions are currently available from ebrary and other providers.

In short, Philosophical Essays is as secure and conveniently re-obtainable as any book is ever likely to be. Removal of a little-used copy is a decision whose logic Bertrand himself would approve. And yet...actually withdrawing and discarding a title so important may be something none of us can actually tolerate, no matter how compelling the logic. Many tough decisions lie ahead for all of us. And as an old Dutch proverb has it, "he who has the choice has trouble."