Market Efficiency, or, Arguing with Rick

Let’s have a little argument with Rick Anderson. His recent article from Ithaka S+R makes a lot of good points. The primary thesis is that libraries should concentrate on building up their special collections (non-commodity collections, he calls them) because that is what makes each library unique and building these collections will also result in the preservation of the widest possible selection of our intellectual record. I don’t argue with that. No sense in libraries all developing the same collection of commercially published products.

The backstory of Anderson’s thesis, however, (the backstory of all Rick’s articles) is that we live in an information-rich environment (check) and that (commodity) library collections are of declining value because so much of the information they might contain is more cheaply and readily available elsewhere. BONK!

Anderson says:

Today’s more efficient online marketplace features much lower prices and much lower barriers to personal collection-building, a pervasive full-text searching capability that makes traditional cataloging less obviously necessary, and widely distributed storage and access points that undermine traditional approaches to preservation and curation. (p. 3)

Some elements of that are partly true. There is a tremendous growth in full-text searching capabilities (“pervasive,” I don’t know), the Internet generally offers many storage and access possibilities that did not exist before, and the cost of some things is lowered by the 1000-pound-gorilla-ness of Amazon e-book pricing. But here’s my point: only a fraction of the scholarly communication record is affected by these factors.

Anderson suggests (p. 2) that because it is more possible (technologically) for scholars to share articles with one another, the need for such materials to be held and distributed by libraries is diminished. Clearly, it is faster and easier for scholars to share articles with one another today than it was 40 years ago, but there is no evidence I have seen in the literature that says this amounts to a majority of uses of scholarly articles or that this kind of sharing is responsible for a decline in library access. In fact, presumably scholars could by-pass use of the library copy in only three situations: for articles they wrote themselves (easily shared with others), for articles from journals for which they hold a personal subscription (easily accomplished but of dubious copyright and ethical status), or for articles that are freely available in an open repository (the sense of sharing in that case is not quite an accurate representation). It is implausible, in the first instance, that EVERY use of a scholarly article would be mediated by the author of the article. If that were true, the need for scholarly journals would cease to exist altogether. In truth, a great deal of sharing within the scholar community is accomplished, I would suggest, from downloading PDF files from library holdings. Obviously, the open access movement plays a significant role in the enhanced sharing of publications outside the commodity environment, but does anyone believe this is now the primary way scholars share information? The continued promotion/tenure scramble to be published in “top-tier” journals suggests otherwise.

The idea of lower cost and lower barriers to personal collection-building being the drivers of diminished library use (or at least relevance) certainly warrants study, but I don’t think it has been demonstrated and I doubt it would hold up under careful analysis. Do scholars today personally buy a greater percentage of the resources they use compared with scholars of the past? Perhaps marginally. Are scholars acting as intelligent consumers in the scholarly communication marketplace, finding the lowest price in all instances? Doubtful. At any rate, if they were behaving in that way, a free library copy would always be the best option. Interlibrary loan anyone?

The cost of almost all scholarly publications have, in fact, continued to go up relentlessly. Journal subscription prices outstrip the CPI every year. Even purchase-by-the-article services tend to run in the $25 to $50 range per article. Scholarly monographs are rarely cheaper than $50 per title. The existence of efficient distribution channels and digital options have not changed that very much. As an example, here are a couple of items my acquisitions team has dealt with recently:

  • Evolution and Human Sexual Behavior. Harvard University Press, 2013. We paid $33.24 through our regular academic book vendor. The hardcover edition is listed on Amazon at $35.67. There are 11 used copies listed there as well. None is cheaper than $34.50. There is a Kindle edition available at $31.16. No great savings in any case. What may make the Kindle version appealing is the instantaneous access rather than the price, but Kindle would not serve every user need.
  • Guidelines for Chemical Process Quantitative Risk Analysis. American Institute of Chemical Engineers, 2000. Again our vendor is selling it for $254. Amazon lists it at $190. Used copies are more expensive than that. It doesn’t exist in the commercial, individual-user e-book market. Knovel offers it through a subject package and several other library platforms have it at $254 or more. This is content that some of our users really want, but they are unwilling to pay $200 for it themselves.
  • Plessy v. Ferguson (Landmarks of the American Mosaic). Greenwood Publishing, 2012. $58 from our vendor, $37 from Amazon, $35 for a Kindle edition, $32 for a used copy. Some savings over our commodity-priced version, but not much. This is the kind of semi-academic title that might serve student needs very well, but they wouldn’t want to pay $30 for the right to quote a few lines. (Although that IS a use some students might find for the Google Books edition, provided that the appropriate snippet could be found in the limited-preview.) Faculty might be interested in using the book as a teaching text, but it probably wouldn’t serve their research interests. Few would pay for it themselves. Thus, the savings through Amazon is still not that appealing.

Let’s face it, the nature of library collections and collection use is changing due to an expanded digital infrastructure, but scholars and students continue to rely for the most part on the same kinds of resources: scholarly journal articles, scholarly monographs, and reference resources. The fact that these are available digitally and in a piecemeal fashion does not change their overall cost or make them more collectible by individuals. Their pricing, I suspect, still puts them beyond the reach of most users. But I’d be willing to be proven wrong. The mere existence of Amazon and e-book readers or Google and its Books and Scholar platforms does not necessarily demonstrate that these are the causes of declining library use or expanded information-consuming power for users. I’m not sure correlation is evident, let along causation. I think we need a lot more ethnographic studies of user behavior and its subsequent impact on scholarly publishing before we can make those kinds of assertions. But I’m totally on board with asserting that we should expend more effort and money on developing locally relevant special collections. Go for it.



Georgia State E-Reserves

A verdict has finally been issued in the Georgia State University e-reserves case (Cambridge University Press et al v. Patton et al). Several publishers were suing GSU over their electronic reserves practices. The judge’s decision is mostly favorable to libraries. Most of the particular claims of infringement were rejected. The case, however, may establish some specific guidelines or safe havens that may not be exactly what librarians would want.

Further summaries of the case:


Tech Timeline

I’ve been working on a tech timeline (both personal and library) that I thought more appropriate over on my personal blog. Tell me about the technological change you’ve experienced in your career.

Academic Kindle

ALA Midwinter 2012: I don’t know if this is common knowledge, but a source at Ebsco tells me that Amazon will soon start offering Kindle downloadable content through academic e-book aggregators like Ebsco, Ebrary, and EBL. I didn’t hear a timeframe on this other than “soon.” Obviously, the Kindle program through OverDrive was considered a beta project. Amazon seems ready to expand library lending beyond the public library market.

As it was described to me, the Kindle academic lending would operate somewhat differently than the OverDrive program, which actually redirects users to affect the loan through the Amazon store. It sounds like the academic plan is to allow a seamless and direct load to the Kindle Fire. Other Kindle devices would require download to a computer and then transfer to the device. That procedure sounds a lot like Adobe DRM-protected files, where an e-book is downloaded and authorized through Adobe Digital Editions and then transferred to a registered device. One assumes that Amazon won’t using Adobe DRM. So, what is the transfer mechanism? Just a drag and drop file management process also seems unlikely. So, I don’t know how much all this is vapor ware , but my source seemed pretty confident it would hit the shelf fairly soon. Looking forward to seeing how it works out.

It’s time for productivity not publicity

I’ve written here a time or two about apps developed by libraries to highlight unique collections. Three prominent examples:

All of these are well-designed and effectively promote their respective libraries to a readership that is increasingly online and mobile. As a first step to demonstrate, at least partly, what is possible in the mobile world, they are impressive.

Suddenly, however, I begin to have issues with these sorts of library products. Part of my misgiving about them is that they are demonstrations. They are tools for promoting the library collections but they are not the library collections themselves. They are surrogates and even billboards for tangible collections sitting on library shelves.

Obviously, a great deal of the digitizing that libraries have done to date has been focused on unique and valuable collections. Digitization has always been advertising for what was great and special about any particular library. It has been a means of creating online exhibits, as it were. But it is precisely that goal that makes library digitizing and app development of limited value to library users. Advertising and exhibits can be interesting, informative, and motivational, but, in the end, they do not take the place of actually using the library. They do not even especially aid or enhance use of the library in any immediate sense.

Part of what makes the Google Book project so interesting is that it has the audacity to digitize the entire library (many libraries, in fact). It is the library in a way that most library digital projects and products are not. Aside from lacking a collection scope that would serve more user needs, most library digital products are also short on functionality. They are not intended to enable users to do things with information. I think it is time that we start to build apps that will help users do things with information.

Some of what I envision has, up to now, been left to vendors and commercial interests to develop. Some may be beyond the financial means of libraries. Some is just plain hard to do. I think we shouldn’t let those things stand in the way of trying. Collectively, we ought to be able to push the envelope more than we have thus far.

Some of the things library mobile apps ought to enable:

  • Discovery of library collections across all kinds of formats
  • Authentication for and use of licensed digital collections
  • Annotation and note taking
  • Citation management
  • Sharing and conversation via social media
  • Remixing and mash-up of content

We can talk a lot about the digital library, the virtual library, but until the tools we offer to users actually enable use, we are only advertising for the physical library and hypothesizing about a digital future. It’s time to give users real productivity tools and make the digital library a reality.


Project Muse Beta

Project Muse is rolling out a beta site that demonstrates their future integration of journals and ebooks. (You’ll need a campus subscription to Project Muse to get to the beta.) The ebook product should go live in January 2012. We can certainly expect the ebooks to be a quality product, like the journals. Muse is collaborating with the University Press Content Consortium to provide “digital books” from about 65 different publishers. According to Project Muse newsletter, the digital and print versions of books will be released by the publishers simultaneously. The ebooks will also offer unlimited simultaneous use and will have no DRM. The files will be PDF.

If the beta site is any indication, each book will divided up into individual PDF files for each chapter. I know the intent is to keep the file sizes manageable and to make each book consumable in smaller bites. I think this actually works in a negative way towards download of books to mobile readers. Does anyone actually want to download and manage 15 different files in order to read a single book? There should be a choice to download the entire book. In fact, I’m rather disappointed there won’t be a file option other than PDF. EPUB would work more seamlessly with mobile readers (well, Kindle excepted).

One of the points I’m interested in is whether users will really embrace a journals-n-ebooks collection mix. I know several scientific publishers have already been offering this for several years. (See Elsevier and Springer for example.) And the great push in libraryland is to develop discovery layers that integrate everything into one search interface. Still, platforms like Project Muse and JSTOR that have done such a good job with providing journal access are venturing into new territory. (JSTOR will have an ebook product available on about the same timeframe as Muse.) Platforms like this have always been a selective subset of journals. In such a selective environment, does it make sense to integrate journals and books? Of course, their users will love these platforms. The user population for Muse comes primarily from book-centric humanities disciplines. They will likely love having more content and having books added to the mix.

Mobeus Trip

yin yangThe other day, I was at a retreat for my library’s administrative team. We were doing some strategic visioning. I managed to get some of my ideas about collections embraced. Well, actually, they were all pretty prepared to accept a lot of what I want to have as part of our collection vision. In fact, our head of discovery services (cataloging) was pretty quick to recognize the key concept: discovery and collections are one and the same. We spent some time coming up with metaphors for that idea. It’s the yin/yang of the library. Or maybe its a mobeus strip. Collections and discovery on different sides of the strip…wait a minute!

There are a couple of other key concepts to this vision. Discovery is seamless, which is to say all discovery looks the same. There are no separate interfaces. It’s all one interface, one discovery platform (though perhaps with many customization and personalization features). The other concept, however, is that the collection is a hybrid. Some of it is an owned physical collection, some of it is subscription-based, some free, some curated local data, much of it is digital. A lot of it is not owned or subscribed to in any normal sense of those words, but access to it is mediated instantaneously or on-the-fly. Demand driven collections that happen at the point of need.

The discovery mechanism is what allows your library users to have access to information. In this environment, library staff will be engaged in pointing the discovery mechanism in the direction of collections or repositories that are useful. Although I think users should also be able to point it anywhere they like. More to the point, library staff will be engaged as well in creating metadata for information objects that will make them more discoverable, regardless of whether they are already owned or subscribed to. I guess this also means that cataloging and acquisitions are more or less the same process.

Rent or Own?

I have a running debate in my brain: rent or own? Libraries have been in the ownership business for a long time. You might say we are almost obsessed with containers. In the past, information was only available in containers: a particular copy of a book, a print subscription to a journal. Those physical items sat on our library shelves. In fact, their very physicalness gave libraries a lot of rights that made library collections possible and practical (see the right of first sale doctrine). As we move into an electronic environment, a whole new set of principles and laws apply. I don’t think libraries and publishers have quite worked out their relationship in a totally digital world. It is a time of chaos. Libraries would still like to hold some of those rights that inhere in physical objects. Publishers, however, think different terms should apply. Mostly, publishers have had their way. Digital library collections are largely driven by license agreements that dictate different terms than the ownership model, or the level of ownership is limited or restricted (try interlibrary loaning an item from your digital collections).

But the radical thought I’m having is whether we should give up the notion of ownership altogether. What if libraries and collections staff focused their energies on use? What if our entire raison d’etre was simply to media and enable use? Give up the entire ownership model? This seems especially relevant in a digital library collection. Does it really matter if we own something that isn’t really “anywhere” anyway? This might make some of the recent ebook debates seem irrelevant. No HarperCollins limit on the number of uses your ebook can have before you have to buy it again. You pay for each use individually. This is scary for both libraries and publishers. It can make it difficult to budget for usage. For the library there are no discrete costs; they are behavior driven. For the publisher, they might discover that large swathes of their catalog do not generate any library sales at all. (See Go To Hellman on the Pareto Principle–in fact all of Eric’s recent posts.)

What one hopes is that the as-yet undetermined pricing model is fair to both libraries and publishers. For the libraries, one would hope that the unit or usage cost would be low enough to enable use of a greater number of unique titles at a total cost that was not greater than what we are already spending. Obviously, that is a hard thing to work out. Publisher: “How much money you got?” Library: “$100,000.” Publisher: “Send it to us.”

What kind of math can we use to determine fair usage cost? I think the library world needs to spent some energy thinking about this. Otherwise, the models will be entirely driven by the publishing world. ALA now has a Presidential Task Force on Equitable Access to Electronic Content, but the thinking needs to spread out beyond the task force. It is something all collections and acquisitions librarians should be thinking about and talking about. Collectively, we will come up with more and better ideas.

Just Another Brick in the Paywall

Just over a month ago, the library world blew up when OverDrive announced that ebooks from HarperCollins would have a limit of 26 uses, after which libraries would be required to buy an additional copy. Librarians, understandably, were not pleased.

Last week another limit to online access took effect. On March 28, the New York Times implemented a paywall. After using 20 articles within a month, visitors to will be asked to take out some kind of subscription. Subscribers to the print edition will have unlimited access. Little mention of this is being made in the library world.

Nonetheless, there do seem to be library implications. One presumes that IP addresses and cookies are the means by which NYT will track users and uses, and that subscriptions will be monitored with individual IDs and passwords. As of yet, there is no model for institutional access.

All these mechanisms present problems for access on public computers in libraries (or even on staff computers for that matter). Will particular computers begin prompting users for subscription or login information? How will we keep track of public use? What will be our response to the angry patron demanding access? It is unclear what will be the best response. Perhaps deleting cookies will reset the usage clock. Perhaps switching to a different browser will work, or moving to a different computer. Or do we tell patrons to buy their own subscription? All of those seem rather impractical in a library.

Yet librarians, it seems, are not very concerned. We typically have other online access to NYT through database aggregations. I would bet, however, that a large majority of NYT use on library computers happens through We also have print subscriptions, which would entitle us to one online user per print subscription–we could just run around and log on each library user with our one password whenever the paywall kicked in. (Violation of TOS?)

Perhaps more than 20 NYT uses in a month on a particular computer will be rare, but I doubt it. It may be that NYT will quickly come up with a practical and affordable institutional solution. I doubt that too. Better start thinking about this.