7 February 2012

Assessing the value of print and ebooks for academic libraries

How does one assess a library collection, especially the relative value of print and electronic book collections?

Among various recent studies of academic libraries, this one from Tina E. Chrzastowski of the University of Illinois at Urbana-Champaign Library study, sets out to value ebooks in terms of their cost, use and cost-per-use; and also their perceived “esteem” by users.  It concludes that ebooks represent better value for money than print (Table 1 p.4), and that ebook "attributes seem to signal a "win-win" for libraries and library users" (p. 14).  However, I would advise caution in accepting this findings as applicable to other institutions, and will explain why.

This research, funded by Elsevier, looks at the cost and usage data for the ebooks held by the UIUC library in 2008-11, during which time their collection increased from 292,002 to 614,203 titles.  As is often the case with ebook statistics, the data on use is limited to the number of title hits; and in this case covers the majority, not the total, of their collection.  As the author admits, we cannot extract more information that this, so we don't know whether the hits are from unique or returning users, nor how many returned information which the user sought, nor how many users stayed for, say, more than 5 minutes in any ebook.  And, as the data was provided by publishers, a second, impartial analysis of the data would have been able to confirm the findings.

With that in mind, let’s look now at the analysis of the cost per view (Table 4).  Over the four-year period, this fluctuates between $1.48 (£0.94) and $.68 (£0.43).  Good value, you may think.  But buying ebooks in large discounted packages would be bound to yield a reasonable rate of return; the point is surely how much of what the library's users wanted was actually available in the ebook collection, and how much was not?  And good value ... in comparison to ….?  As far as I am aware, no one has yet collected data on the number of times students open a print book and read it, whether the whole book or just a few pages, and it is difficult to see how comparative figures could be accurately collected.  There are very good reasons for buying ebooks, but we simply can’t claim on this evidence that they represent better value to a library than print.

The second measurement used in the study, the “esteem” of ebooks, is drawn from the activities and responses of 129 students using a collection of Elsevier ebooks only, 114 of whom identified themselves as being at doctoral level, and 77 (of the 129) as studying physical sciences (fig 4).  Now, I'd suggest that postgraduate physicists are one of the groups of students who would be most comfortable with using electronic resources and therefore ebooks, so we can't assume that these figures would be necessarily representative for students of other subjects or at other levels.  

Fig 7 (p.14) reveals the students' attitudes to the information in Elsevier ebooks.  While 54.8% said the ebooks were "nice to have" over a quarter (27.1%) said that they “could have done without it".  I'm not sure that "nice to have" is a sufficiently explicit statement to tell us much, and at just over half the sample it isn't a ringing endorsement, as the author admits.  I should point out that while respondents were asked questions about the advantages of ebooks they were either not asked about their disadvantages, or the results were not included in the published report, which suggests an unfortunate bias in the research.

Having assessed the value of ebooks, in Table 1 (p.4), the study goes on to illustrate the relative costs of maintaining ebook and print collections.  To establish the costs of maintaining a print library, the author draws on the work of Courant and Nielsen. While this is very clear and readable, its application to UK circumstances should be very closely examined.  Here are some reasons why.

A nice clean library

Firstly, the authors propose that a print library has costs associated with space, cleaning, maintenance, electricity, staffing, circulation and access which are less than those associated with e.  Courant and Neilson analyse print costs per item, based on these books being retained in perpetuity, but certainly the Library where I work is engaged in ongoing and rigorous weeding, so any libraries with the same kind of collection management would need to revise the figures.

Secondly, their figure for construction costs assume that buildings must be completely replaced every 40 years.  Our Library was built in the late 1920s and there is no prospect of it being replaced for at least a decade so again, our figures would differ significantly from those in the report. 

I have looked more closely at the figures Courant and Neilson quote for cleaning costs, as these are the easiest to compare accurately with the Library where I work. They assume an average cleaning cost over the lifetime of a library of $3.64 (£2.30) per book.  This figure is obtained from the University of Michigan’s Buhr Shelving Facility, a vast remote store, using five year’s worth of data, averaged, estimated at the cost per square foot and calculate at a cost per book based on the storage capacity of various shelving types.  I find that using this figure, the cleaning costs for our Library would be something around £92,000.  However, a few calculations reveal that the actual cleaning cost per book for our Library is £0.029, and at this rate, we would have to retain
all the books for 79 years to achieve Courant and Neilsen’s sum..  Now we do have some books published in 1931 or earlier, but at a guess I would say that at least half of our collection is much more youthful. 

Lastly, we should also consider the content of the library with which we make the comparison.  In our case, it would be inappropriate to compare the cost of the ebooks in our current collection, which represent a proportion of reading list material plus more from various collections, against our print collection, as this comprises a mix of research, undergraduate and principally textbook material.  When/if e-textbooks become available to us the cost of ebooks is very likely to increase dramatically, thus revising costs in favour of print (should print textbooks still exist).

Therefore, I would suggest that all Courant and Neilson's costs should be recalculated for local circumstances before assumptions are made about ebooks being cheaper than print. Recently I carried out a comparison of the print/e costs for a small collection of Elsevier titles based on purchase/subscription price alone.  This revealed that over a period of three years (the likely lifespan of a print edition) the costs of purchasing print across the University were less than 3 years' annual subscription to e.  It seems to me that it is only relevant to add in the library building and maintenance costs to the comparison if an institution plans a completely new purpose for the building, and I wonder whether, for example, an administrative building would cost less than a library to maintain, or be as productive?

As Courant and Nielsen conclude (10), libraries often occupy attractive sites in the centre of University towns.  Institutions which are tempted to repurpose or sell off library buildings and make provision for their readers through electronic resources are applying a solution which doesn't fit the problem.  There are extremely good arguments for academic libraries to develop ebook collections, but savings from print collections are only part of the picture.  Antagonism to the ejournal "big deals" (c.f. A Man of Badly Encoded Character's blog post) and the current volatility of ebook publishing suggest that a measured transition, taking into account a wide range of factors, would be wiser.

Further UK studies will be of interest. But whatever decisions are made about the future of academic libraries, they must be based on accurate evidence, not glib assumptions, especially about the relative costs and benefits of both print and electronic provision.

No comments:

Post a Comment