Ebook economics: Are libraries screwed?

"Kindling" by Flickr user oskay

The advance of ebooks will no doubt bring much good. As often with technological change, we probably can't even predict what wonderful new things will emerge! But we can see some serious dangers ahead, and try to deal with them. I see three major areas of concern: to libraries, to physical bookstores and to the freedom to read in unfree countries.

This post explores the first of these—the danger to libraries. There are, of course, arguments to be made about the viability of physical libraries in a digital age—that while libraries aren't just buildings, the building still define much of what they do. That is not my point here.

Instead, I want to advance a pricing argument: that ebooks will end up costing libraries far more than paper books ever did.

Premise: Libraries will need a "library model" for ebooks.

A few libraries, such as NCSU have been experimenting with ebooks. Without exception, they are following a "consumer model," buying a large pool of devices and then buying books locked to individual devices in the pool.

This model is great for experimentation—to test what patrons think of ebooks and figure out what to do with them—but it's not a long-term solution. Digital books locked to individual physical devices are worse than physical books. That is, when you take out a physical book, one book is unavailable. When you take out a Kindle with 100 books on it, 100 books are unavailable. NCSU has bought extra copies when students need another copy in circulation. Obviously that's not a long-term solution.

Because the "consumer model" won't work, libraries will need—and publishers and ebook providers—will create a "library model." The library model will involve a "site license" model—a pool of books, with rights to use them on X devices at a time. Publishers are already talking about this.

Thus, libraries and consumers will be using different models. The market will "split." (I understand that Netlibrary and Ebrary, two library-centered ebook vendors, already used by many libraries, work this way now.)

Economic effect: Libraries are screwed.

1. With regular books, libraries took advantage of the same deal regular people got, but extracted a lot more value of that deal. That is, a regular person mostly got a single use out of a book; libraries got many more uses. We didn't think of it this way, but libraies had a "site license" of sorts—the so-called "first-sale doctrine."
   
   With the first-sale doctrine sidelined by digital rights management (DRM), publishers will seek to extract the higher value of their books within a library context. This will cause prices to rise.


2. With physical books, library price discrimination was impossible. Libraries and regular people bought the same stuff, and paid the same prices. If a given edition was pitched to libraries, its price was held in check by the availability of non-library editions. As a result, only purely academic titles had run-away libary prices—think Brill with its $300 monographs.
   
   Once the market is "split," price discrimination is possible. Publishers will charge libraries more for the extra value they get because they can do so without hurting the consumer market. This will cause prices to rise.


3. The cost of paper books have traditionally been held down by the existence of a secondary market. Copyright is, of course, a legal monopoly on the production of a given work, but once paper copies have been sold, new sales compete to some degree with the used copies out there. If you don't want to pay $242 for Brill's Collected Papers on Greek Colonization, BookFinder lists 25 used copies under $215.
   
   Because ebooks are non-transferable—and if such ability is added, it surely won't allow a consumer to pass an ebook to a library under library terms—no secondary market will exist. Until copyright expires, libraries will have to go to a single source—the publishers who have the copyright monopoly. This will cause prices to rise.


4. The "library model" will be inevitably pushed toward "rental" not "ownership." As many have remarked, ebooks are already more like "renting" than "owning," with no right of resale and at least the technical ability for the book to vanish at whim. Libraries, afraid of buying goods that a technological change or company bankruptcy will obliterate, will seek to avoid the "lock in" of ownership. Publishers will also see opportunity in offering large "packages" to libraries—packages that provide rental access to a collection that would take years to build up in a traditional buying-and-owning model.
   
   This logic is how libraries were pushed to renting their journals. It's also at work in enterprise software, either de jure or—through regular version upgrade payments—de facto. Libraries will rent, not buy, their ebooks.
   
   The combination of monopoly and rental is dangerous. It's how journal subscriptions have risen faster than inflation for 40 years, and spiked precipitously upward in the last decade. (The classic ARL graphic can be found here.)
   
   The logic of journals is the logic of the site-licensed ebook. Prices will rise unchecked. Some relief may come if the open-access movement goes past scholarly journals into other scholarly publishing—there's really no reason Brill books need to cost $300! But this will take a while, and it will only affect scholarly titles.
   
   Rental means prices will rise.


5. In the past, libraries could "coast." Collection development was a long-term thing, and libraries could, if necessary, restrict their acquisitions budget in line with financial realities. When times are bad, you buy less. When times are good, you buy more. As long as you have both ups and downs, the library as a whole stays healthy.
   
   Rental will change this. Libraries will only be as good as their last subscription check. This will change the nature of collection development (in both good and bad ways), and give politicians new opportunities for both unsustainable budget growth and budget-cutting during crisis. This may not cost libraries more, but it will put their value on the knife-edge.

What do you think? I've started a discussion topic in the "Librarians who LibraryThing" group.

---

I'm sure there are lots of good arguments against this post. Here are two that came up as people read earlier drafts.

Jason Griffey argues (by Twitter) that prices will be kept in check by wide availability of pirated versions. This is a good argument. The counter-argument is corporate software. It's not hard to get a free copy of InDesign or Photoshop, but corporations continue to shell out nearly $1,000 for each, because the penalties are so steep.

Another correspondent suggested the "dawning age of biblioplenty"—a world in which "millions of books will be available from almost anywhere"—will act to hold down prices, presumably through what economists call indirect competition.