Monday, August 12, 2013

A Rational Framework for Library eBook Licensing

Since the Redigi decision made it clear that there is no right of first sale for digital content in the US, it's been much easier to think up realistic doomsday scenarios for public libraries in the US. Why should a publisher let a public library lend an ebook if Amazon or some other competitor were to offer much better terms? How would our public library system, saddled with difficult-to-use systems and unfavorable contracts, ever hope to compete?

Back when HarperCollins first announced that it would only let libraries lend their ebooks 26 times before they would expire, there was widespread outrage from the library community. Looking back on that, it seems pretty clear that a lack of consultation and poor customer communication fueled the furor. By itself, the lending limit could have terrible long-term consequences for libraries, but as part of a wider, well-thought out framework, it could be useful component.

I've been doing a lot of thinking about this over the last 3 years, and I've decided it's time to float a comprehensive proposal for how libraries and publishers might work together on ebook distribution to benefit the entire reading ecosystem. eBook lending as implemented to date has been founded on a combination of irrational fears and outmoded processes. We deserve better.

Behind this framework is a set of assumptions.
  1. Library ebook distribution must sustain and increase the total population of readers; this is a prerequisite for a healthy book publishing industry.
  2. Patron discovery of ebooks in libraries must connect effectively to ebook sales.
  3. Library distribution must become much more efficient, and overhead must become much smaller for ebooks than it is today for print books and ebooks.
  4. Long term preservation of ebook availability must be a joint undertaking of libraries and publishers.
  5. The economic models used for library ebook distribution must provide incentives for libraries and publishers to promote points 1-4.
I don't pretend that people won't disagree with some or all of these 5 assumptions, but if any of them are false, then, I think there will be NO distribution of ebooks through libraries. I also recognize that not all books are alike; even if library distribution works for some ebooks, it's unlikely that it will work for every ebook.

So the fifth assumption is what this post is really about. Given 1-4, what should an economic framework look like? Here are the features of a model that makes sense to me:
  1. Decoupled pricing. An ebook license that allows for lending makes the ebook more valuable, so why shouldn't it cost more than an individual, non-transferable license? I can't say whether Random House's 300% markup for libraries is excessive, but why not let the marketplace decide? For new, super-popular ebooks, maybe 500% markup makes sense. On the other hand, maybe ebooks that need exposure should have an 80% markdown because libraries might turn them into bestsellers.
  2. Rate limits instead of DRM. Patron license embedding.  I've written about this before. This may take the most convincing, but in thinking about the imperatives of effective discovery, low distribution overhead, and long-term preservation, I've concluded that there are no alternatives to major change in library distribution technology.
  3. Circulation charges after an initial period. Most books are bought in the first year of publication. Today, libraries "deaccession" books to match their declining demand. But there's no reason for a library to deaccession an ebook, so for most books the global supply for any given ebook will eventually exceed global demand. If the library can cut its transaction cost from ~$2 per circulation to $0.20 per circulation it seems fair to reward the publisher with part of the difference for developing books with long term value. 
  4. License transferability/InterLibrary Loan. Libraries rely on interlibrary loan to expand the scope of their collections and meet special needs. But ebook loans can be instantaneous, so digital ILL can compete directly with backlist sales. If the transaction costs (currently ~$10) for ILL can be squeezed down to $1 or so, there's plenty of margin to provide a transaction payment to the rights holder for the privilege of doing so. 
  5. Patron-funded purchases. Libraries are tight on funding even as they need to completely transform what they do. Their biggest asset is a huge reservoir of public goodwill. At this pivotal juncture, their ebook offerings are characterized by long hold queues. Why can't a library patron buy an extra copy for the library and jump to the front of the queue? Why don't publishers offer "Buy for your Library" buttons on their catalog pages? The reasons are complex, but it's mostly a case of "we haven't done that before". But if it doesn't happen I just can't fathom how library discovery can effectively plug into publisher commerce.
  6. License durability. If libraries are expected to "buy" ebooks, it should be pretty much for keeps. If the publisher for some reason has to revoke a license without cause, the library should get a refund of the license price.
  7. Archival copies. Libraries need to do a lot of things with books other than lending. Indexing and archiving are good examples. The saddest thing about the most successful library ebook distributors today is that libraries don't get access to unencrypted ebook files. If libraries are to offer effective discovery and archiving of ebooks, they need access to the files. Seems a no-brainer to me.
There are a bunch of parameters to plug into this framework; here's my guess as to what they should be:
  • Rate limits: One authenticated user per two weeks.
  • Circulation fee: $0 for the first year, after the first year, 2% of purchase price or $1 whichever is greater. 
  • ILL fee (publisher share): 5% of purchase price or $2, whichever is greater. 

A rational ebook lending framework would mean big changes for both the book publishing industry and the library industry. Even if a HarperCollins decided today that this was an attractive way forward, it would be hard-pressed to find a way to implement it, because libraries just don't work that way. So it seems a bit far-fetched at this point. Based on the iBookstore fiasco, it appears to be illegal for big publishers to even talk to each other, let alone drive business model changes. It's good that a library group is still trying to figure it out.

Maybe some small startup company could try some sort of pilot program.


  1. This comment has been removed by a blog administrator.

    1. I messed up and deleted by accident. Here is Lesley's comment:

      You have clearly done way more practical thinking about this issue than I have and I salute you! I agree with the majority of your proposal and I'm glad to see something other than just crying over how badly we are being treated :) I have a question about circulation fees though. We currently offer both OverDrive & Freading to our patrons -- with one we pay a huge annual fee to have access to the platform and the collection, with the other we pay only per circulation. If the OverDrive model becomes both a large annual fee + circulation fees after the first year we purchase an ebook, I'm worried we would be paying a lot more than we currently are. I wonder if that kind of a system would change how the distributors charge for their services (since we don't buy directly from the publishers). Thanks for this thought-provoking post!

    2. Lesley-

      One of the main points I wanted to make was that the fat margins currently consumed by intermediaries is simply unsustainable. Publishers need to work together with libraries to create efficient distribution systems that allow fair payments to flow back to the rights holders.

  2. I would be very happy for our library to pay a publisher a fee of X+Y+Z for each eBook. X would be an initial payment for us to acquire the eBook, Y would be a per-circulation fee, and Z a payment to assure we have perpetual access to the license. I presume the publisher would be very happy to make the same deal with us. The only issue is the amounts of X,Y and Z that would be mutually acceptable to both parties.

    I would be very happy to weight the fee heavily on Y, so that I know that my payment for the eBook is primarily determined by its circulation. Another factor in this is that the publisher would have some skin in the game, too. I'm going to guess a publisher would want a lot more X and Z than I would want to pay, but that's what markets are for, aren't they, to find the overlapping interest/price point.

    In this model it would be in everyone's interest to allow multiple simultaneous circs of the item, and I could see a formula that had a higher Y in some initial period than in later periods.

    The model would also place the burden on us to manage our users' access and activity. A particular concern would be what constitutes a circ -- simply clicking on an item by a patron would cost us money, unless we managed access by patrons, as you've noted.

    Alan Kirk Gray

    1. Alan,

      Here's the difficulty. We know there's one solution to the X,Y,Z model. It's called Kindle-iBooks-Nook-Kobo. Y is about $10, X and Z are effectively 0. If you believe that a library model can coexist with K-i-N-K along the same axis of business model space, then no worries, but I don't think that's viable.

      So the key question is whether there's a second axis that libraries can live on with minimal interference with the K-i-N-K model. The traditional model has X=$50, Y,Z=0, and no simultaneous access to reduce channel interference. I think we can do better.


  3. Like your blog.

    I don't think the case alters much. Until the copyright law for Digital First Sale is changed by another Act of Congress (the current copyright Register of the US has written about the need for reform, linked on your site from 2011) the courts can do nothing but interpret the law as it is.

    1. I agree- much more would have changed if the ReDigi case had come down the other way.