Friday, July 30th, 2010

The magical $0.50: Why ebook economics don’t work in libraries.

ebooks in libraries

A publisher counts its library ebook money.

Ebooks are hot, and libraries are noticing.

The public library conversation about ebooks is also heating up. Unfortunately, much of the conversation ignores a critical factor that makes ebook lending problemmatic. That factor is not luddism, but the simple fact that, for libraries, ebook economics doesn’t make sense.*

The $0.50 Circ. I’ve written on this topic in the past. My contribution today was for me to spend some time trying to figure out what it costs to take out a paper book from a public library. What, in other words, is the cost per circulation?

I did some quick work at the Institue of Museum and Library Services’ Library Statistics, which collects statistics from some 9,000 public libraries, and presents them in a simple, searchable interfact.

The data is pretty sparse, but it does include two critical data points: Total Collection Expenditures and Total Circulation. From these you can derive a third statistic, Cost per circulation.

For example, if the South Portland Public Library spends $78,038 per year on materials, and those materials are taken out 249,431 times, we can see deduce that the cost per circulation was $0.31.

All told, I surveyed 25 libraries, running down town names—eg., Portland, ME ($0.54), Portland, OR ($0.29), Portland, CT ($0.51), Portland, IN ($0.31)—as well as including a few big systems—eg., Dallas, TX ($0.48), Charlotte Mecklenberg ($0.50). You can see my spreadsheet here.

The average of all comes out at $0.50. I don’t think a much larger survey would change that by more than a dime in either direction.** (If you want to run the full numbers, and have a copy of Microsoft Access, ILMS has the data files.)

Unfortunately, this doesn’t get us all the way to a book circulation number. But it gets us close. Print circulation still makes up the majority of circulation transactions. And books are not the most expensive materials libraries circulate. So whatever the number, it’s near $0.50.

Come and get your half-buck! Cost-per-circulation is a critical number because it shows how much libraries are spending to satisfy their patron’s reading needs–ignoring staff, building, etc. If the switch to ebooks causes that number to rise, we can expect library budgets to rise as well. If it rises a lot, library budgets will rise a lot.

Unfortunately, to keep library budgets the same, ebook sellers will need to accept $0.50 for each library loan. That’s not just the publisher’s money. The same $0.50 must compensate the publisher, the author and the ebook seller—all of whose costs were figured into the previous $0.50 calculations. (You can perhaps begin to understand why publishers and authors have never liked libraries, although few will say it openly.)

This princely $0.50 is compared to what ebooks go for, on average, in the consumer market. Of course nobody knows the answer there, but Amazon has tried to get everyone to adopt a $9.99 price point for popular titles and the average must be on this order. So, in order to keep library budgets in line, ebook sellers will have to be willing to accept 1/20 of the revenue they could get from the consumer market.*** Even assuming some library buyers buy less than they would have loaned, that’s a daunting ratio.

The magic is over. Will ebook sellers accept the $0.50 deal? Of course not. To think optimistically in the face of such numbers is to avoid the problem. Publishers, authors and ebook sellers aren’t stupid.****

Libraries at least are smart. Libraries are so great because aggregated lending is a wonderfully efficient thing, wringing high value from low cost. By buying books for the whole community, and making them generally available, libraries managed to reduced the cost of reading to a paltry half-dollar, far less than a paper book ever cost. That was a magical time.

With ebooks, that time may be over.

Reply to “How is this in anyway different from the case with pbooks? Hardbacks normally retail well over $9.99, yet library shelves are filled with them. Somehow the publishers manage to survive with this.”

The difference is that publishers have never been able to stop lending–by libraries or anyone else. Like reselling, giving and inheriting, lending is a basic right we have so long as its not taken away. The only way to take it away is through a law (as applies to software rental) or a contract. Publishers actually did try posting notices inside books, preventing resale (see Bobbs-Merrill Co. v. Straus), but it was stopped. A real contract requires both sides to explicitly agree, and notices like that were ruled not to constitute a contract. In theory, bookstores could start requiring real contractings, only selling you a book on the condition you didn’t sell or lend it to someone else. Obviously, this would outrage people, so it’s never been tried.

eBooks make theory a reality. All the ebook licenses constrain gifting, reselling and lending, and most prohibit it entirely. The media change has largely obscured the shift. People know they can’t lend software, and indeed it makes a certain amount of sense. Digital goods are so frictionless an unlimited ability to lend, even constrained by a one-at-a-time rule, would quickly give rise to a massive lending clearinghouse, and a publisher would only sell as many copies of a book as were being read simultaneously at its peak.

As far as surviving, my post made no claims whatsoever about surviving. Publishing has always known that lending and resale cut into revenue, just as people who make snow blowers know that, if nobody could lend or sell a snowblower, more people would buy them. In the world of physical sale, lending and reselling are normal and, ultimately, built into the prices we pay. Just as snowblower manufacturers don’t go under even though my wife and I exclusively borrow our neighbors’, Publishers don’t either.

* This post only deals with public libraries. The issues are different for an academic library. Academic publishers have a much smaller “consumer” market, and academic libraries have much high cost-per-circulation. I suspect that academic libraries will transition to ebooks more easily.

** Obviously, real circulation involves books bought in previous years, but assuming neither number swings wildly around, you can estimate the cost per circulation very well from the collection expenditures and total circulation.

*** The only escape is in the notion that libraries don’t canibalize consumer sales. There’s certainly some effect there. If everyone had to buy their books, some people wouldn’t buy as many books as they used to take out for free at the library. But is this effect sufficient to overturn a 20-to-1 pricing disparity?

**** We have, of course, completely ignored the cost of the devices themselves. If libraries also lend the devices, their costs will go up still further.

The conventional rejoinder here, after calling me a luddite, is to argue that publishers are already willing to let Overdrive lend out some ebook titles in libraries. I contend that they were willing to experiment with it, especially considering how limited Overdrive is as a platform, and that in any case, when you add up all the fixed and variable costs, and divide by actual usage, Overdrive is actually quite expensive. It’s also worth noting that a recent COSLA report on ebooks noticed Overdrive is now refusing larger agreements, speculating:

“OverDrive won’t sell to the LYRASIS consortium and has begun to balk at statewide purchasing groups. Maybe this is for the comfort of nervous publishers who view ebooks as frictionless, ripe for piracy, or long-term profit losses as library products”

Labels: ebooks


  1. pd says:

    This is also assuming there isn’t a massive separate online database of ebooks created, eliminating the need for public libraries to stock them–Google tried to do this via Google Books a while ago; I believe copyright law in the U.S. stifled a large part of that effort.

    Also, they could charge for using their e-book service: most libraries these days lend out digital media for a nominal fee (and academic libraries usually charge for non-current students anyway).

  2. jonathan says:

    On those numbers it doesn’t even make sense economically for publishers to let us lend $15 or $20 paper books either. They’re still far better off selling individual copies to every reader.

    If what you say about authors & publishers never liking libraries is true, it’s not really that they’re getting a lower percentage with ebooks, it’s more that the game is changing and they have the chance to stop something they never liked.

  3. jonathan says:

    Oh, and there’s also Public Lending Right (do you have that in the States?) although I’m not sure how much it is and it is obviously still less money than they would get from pure sales.

  4. reading_fox says:

    Tim “This princely $0.50 is compared to what ebooks go for, on average, in the consumer market. Of course nobody knows the answer there, but Amazon has tried to get everyone to adopt a $9.99 price point for popular titles and the average must be on this order. So, in order to keep library budgets in line, ebook sellers will have to be willing to accept 1/20 of the revenue they could get from the consumer market”

    how is this in anyway different from the case with pbooks? Hardbacks normally retail well over $9.99, yet library shelves are filled with them. Somehow the publishers manage to survive with this. A lower ebook cost is good for libraries – their cost per circulation has just gone down. Of course for publishers it would be better if everyone bought their own copy. But this is true at any price point. What’s needed is to remove the atifical restrictions that make it easy for librarie to do this.

  5. elenchus says:

    I enjoy these posts, as sobering as they are.

    Regarding footnote ****: Any ideas as to how many eBooks are lent on a device also owned by the library, versus eBooks lent using the patron’s own device? (I’m not even sure, does that happen? I don’t own an eReader yet.) Any chance that would alter the basic dynamic you describe?

    The digital divide means there needs to be some lending of devices, but I assume it’s similar to libraries lending DVDs and VHS tapes: some patrons also borrow the players, but most don’t.

  6. BamaGriz says:

    No matter what the future holds… no eBook will ever replace the feel of opening up and engulfing yourself in a hard copy. Some things were meant to be, and paper and ink in a bound volume can never truly be replaced by electrons and plastic.

    “No book is really worth reading at the age of ten which is not equally – and often far more – worth reading at the age of fifty and beyond.”
    — C.S. Lewis

  7. Quasar says:

    That’s just depressing, and makes me wish even more for a legislative approach so that the public library is not excluded.

    And really it just makes me continue to refuse to buy ebooks. Of course growing up in a library culture I’ll admit my book reading is probably 90/10 in favor of library books.

    Interesting that overdrive is mentioned. I managed to get my local library to look into that, but they found it was too expensive for to small a collection.

  8. bowerbird says:


    i love the fact that you’re proceeding in a rational manner,
    using objective criteria, especially the cold hard money one.

    so understand that it’s only out of respect for your method
    that i say that you are trapped in the thinking of the past…

    the world is changing… it _will_ change, no matter what.
    publishers can’t stop the change. they can only refuse to
    admit it, in which case they’ll be swept aside by tsunami.
    (or, if you want less poetic imagery, they will go bankrupt.)
    or they can try to swim with the tide. those are the choices.

    (the corporations will choose to seek easy profits elsewhere.)

    the reality is that electronic-books have zero variable costs.
    once an e-book has been created, it can be put into _every_
    library across the country (indeed, around the entire world)
    at virtually no additional cost. anyone who ignores _this_
    part of the equation ignores their most important salvation.

    so rather than charging a fixed cost per e-book, a producer
    (who i would assume would be an _author_, since i see little
    if any need for publishers in the future) should instead seek
    reimbursement on a per-use basis. we can argue all day
    about what a “fair” per-use cost would be. i put it at $.02.
    that’s right, two cents; authors should get their 2 cents worth.
    cumulated over every library in the world, it will add up fast,
    and that will quickly come to be established as “_the_ rate”…


    p.s. if you want to up your $.50 figure, considerably, you
    should factor in the cost of _reshelving_ a book. look it up.

  9. TweetReflex says:

    Doesn’t DRM undermine this argument somewhat? By ensuring the one-copy/one-user model and limiting how long the file is accessible to a given user, DRM makes the circulation model for eBooks pretty much the same as for pBooks.

    A $.50 circ is quite different from a $9.99 purchase, because while a purchase happens only once, a circ happens many times. DRM ensures that circulation is necessary by making the checkout temporary. Moreover, as with print books, people who get short-term access through the library often decide to purchase their own copy outright.

    Given that the majority of authors in a publisher’s stable are relative unknowns (not everyone is Grisham), and that the library is the place most willing to stock books by such writers, it would make sense for publishers to sell their more popular materials available to libraries at a lower-than-pure-market price (they can make up the loss through sales to individuals, anyway) in order to build recognition for their their less-popular, up-and-coming authors.

  10. Jackichan says:

    It seems obvious to me that the answer is to purchase a $10 copy and be able to lend that ebook out to people 20 times. Then be required to purchase another ebook.

    The real issue is when pirates and hackers figure out the procedure to steal and give away ebooks. Publishers, authors and the public in general will be sorry then.

  11. bowerbird – Have you read Scott McCloud’s “Reinventing Comics”? He has a nice visualization of digital media cutting out the middleman. Based on this, I’d take your scenario a step further and say there would be no libraries, nor bookstores, just a direct author-to-reader download.

    Tim – I wonder how long this transformation will take? Do you have any stats or ideas about how long we have until Bibliopocalypse? Thanks for sharing your thoughts on these developments.

  12. Eric Hellman says:

    50 cents is not the only number that we should be looking at. We should also look at the TOTAL cost of running a library, divided by the number of circs. If you do that calculation , you get $4.09, using the IMLS data. That number should be looked at as a ceiling on what people in the US are willing to pay for access to books in libraries. That’s right, the library overhead rate is 88%. Let’s suppose we create a system for making ebooks available to the public. Surely such a system could be made more efficient. Let’s suppose that we can cut that overhead in half. That means we can make more than $2 per circ available to authors and publishers. That doesn’t seem like quite as much of a stretch, does it?

  13. Mark Kohut says:

    Essay’s reasoning is entirely wrong. It assumes the budget
    drives the buying BECAUSE it is what patrons want (and use it for)…..

    Budget is based on size of library and patrons therefore will always find an average as the writer did. It is because books, etc., have an average price. Were that price lower, cost per circ would be too.

    Cost per circ goes UP when materials budget is cut IF patrons continue growing.

    If ebooks are much more heavily circulated—can be downloaded from home, for example, circ could go way up, driving down cost per circ….

  14. Nancy Ledeboer says:

    I would urge you to reconsider the cost per circulation transaction. Be sure to include the salaries to check out, check in and reshelve the materials. Libraries already pay per circulation costs for staffing. With e-books why not compensate the authors for each circulation. We need an e-book consortium to deal directly with authors. Given then 25 cents every time someone checks out their books. This will add up much faster than what their publishers pay them now.
    For hot best sellers the transaction might be higher for the first year. Let’s hope we can work togetehr on this so libraries are not left out in the cold.

  15. Quasar says:

    Jackichan – what do you mean when? Pirated ebooks are as plentiful as every other media type and amazon drm got broken ages ago iirc.

  16. Stanley Kalemaris says:

    Publishers are trying to avoid the the loss of sales and market control that the record companies have experienced. File sharing of e-books and lawsuits by publishers are inevitable.

  17. Dale Ricklefs says:

    I’m not sure these are the right datapoints. It assumes that the new acquisitions were what circulated, but the circulation stats most likely reflect the ENTIRE collection circulating unless the reported circulation numbers are for the new purchases only. And, it isn’t just the cost of the item that libraries must manage– it is the cost of the infrastructure.

    To get a true cost of circulation, and as most of what we do supports circulation, dividing circulation into total cost of service might give you the real cost of supporting hard copy circulation. For my library, at $2.4 million, divided by 800,000 circ, the cost per circulation is actually $3.00. Still a far cry from $9.99.

    BUT, if you are using the Overdrive model with shared checkout of a book that is $15.00 cheaper than the hard copy version, then your economics improve– no processing of books (staff and materials), no staff to intervene (unless you have self-checks), and public convenience (download it at home– priceless).

    Reference has dropped 70% over 10 years. Google and websites address common questions, as do the databases we offer that are remotely accessible. E-Book readers are becoming affordable, and Kindle and I-Pad may be left in the dust if the Overdrive type models prevail. And, in a techologically advanced area like central Texas, offering E-Books, E- Audio, movies, and music through downloads may reacquaint taxpayers who haven’t “been in” for a while. ANd we need those taxpayers to support the library.

  18. Dale Ricklefs says:

    Eric Hellman:

    If I paid $2.00 per circ, with 800,000 circ, if I close my library I can afford it. A total budget of $2.4 million and a hard copy materials budget (all media) of $300,000 only, we’re closed for business, basically.

    At 25 cents, if that is ALL we paid (no cost for the material from the get go), it is doable.

    Overdrive is already out there for downloadable materials. It obviously doesn’t address all materials. It will be the library community being behind some model and adding economic “force” that will change the limitations and begin to force a common format.

  19. jonesy says:

    There’s just a fundamental lack of sound reasoning here. Ebooks circulate in libraries in exactly the same way paper books do. The library buys a copy and that copy can be leant only to one person at a time. You want more circs? You must be more copies. The 9.99 price point, and consumer demand for lower prices on digital media in general, is merely a reflection of the fact that it costs less – in materials and supply chain operations – to publish ebooks, and consumers know that. If it does cost less, and a lower price only reflects that fact, publishers will make no less money than they do now. Not only is this article a waste of time, it’s actually sort of dangerous, in that it can stir up various counterproductive reactions among readers who lack the training to see where it goes wrong.

    • Tim says:

      “There’s just a fundamental lack of sound reasoning here. Ebooks circulate in libraries in exactly the same way paper books do. The library buys a copy and that copy can be leant only to one person at a time.”

      You will excuse me for noting that you have no idea what you’re talking about. Libraries are not allowed to buy ebooks on the same terms as individuals, and anyone who tells you they are is uninformed. Existing ebook licenses sharply differentiate between private, personal use and what can be done by institutions. Libraries buy ebooks on different terms, from a small number of sellers, covering a small number of publishers. Existing terms are nowhere near as good as you believe them to be—most libraries rent ebooks, because “buying” is prohibitive—and are essentially trials. Publishers are on the record as unwilling to provide the licenses you think already exist.

  20. thegreattim says:

    Whatever previous posters suggest, there will be very little author-to-reader direct downloads for books. Only possibly with the exception of well established best selling authors, and probably not much even then.

    Everybody seems to be forgetting about those silly editors. Have you ever read a raw manuscript? It’s not pretty, nor is it that close to the finished product you’ve come to expect. It is certainly not a book.

    Not to mention marketing to get those new authors noticed in this imagined future sea of author-only connections. And then you still have the typesetters, the designers, layout and illustrations… the list is nearly endless. About the only thing you don’t need from the traditional publishing route is the printers.

  21. Anne says:

    Being in the computer industry and yet a huge supporter of public libraries (and sharing books in general), this issue of 1-to-1 licensing really bothers me. Even the model that OverDrive uses where I check out a book for 2 weeks, make it unavailable to other listeners, and I can’t check it in early is archaic and based on physical books. I bvelieve that instead of OverDrive requiring a library to pay for rental of a collection of audio books, they should pay for a number of downloads. Publishers and authors then would paid for the number of people actual downloading/reading/listening to their book, not having it sit in an “ecollection” unused; at the same time, the benefit of electronic resources would work… we could all be reading the same book at the same time… without waiting for virtual check-ins.

    Overall, I think the ebooks and other electronic media is going to cause a change in the way these materials are licensed… I hope public libraries, and not purely corporate capitalism, can help influence this.

  22. Mike says:

    To those that promote the cost-per-circ scenario: What happens if the library does a *great* job promoting their collection, the public loves it, and they go on to circulate millions of ebooks? It sounds like a great scenario — until the library gets the bill for their $.XX/circulation model.

    One could imagine that maybe they’d limit it. If the rate was, say, $.50/circ and the library had a budget of $100,000 — they’re paying for 200,000 circs. What happens if they hit that in June?

    One thing is for sure, these aren’t easy questions to answer…

  23. allen424 says:

    I also think that anything that helps establish e-Books as a viable method of delivering reading material is an advance. When I buy a physical, printed book, I have the right to use it forever. It is the wonderful opinion the things which are mentioned and used here are numerous.

    Resale Rights

  24. nmoore says:

    This article is of interest to me as including a section for e-book check out in school libraries is now something being discussed, as these are the books that many students are now reading. I wondered how well this was working in public libraries. This is good information to consider for school libraries as well.

  25. Kate says:

    I don’t think the $0.50 cost takes into account the most expensive part of circulating a book– the staff time it takes to check out, check in and re-shelve the book. Additionally, it costs staff time to retrieve overdue materials, handle damaged ones, and locate mis-shelved or missing ones. Outside of circulation costs, there’s also the cost of staff to process the book before putting it on the shelf. I’m wondering if there’s a way to make a comparison taking staff time into account.

  26. Tim says:


    So, strictly speaking, those factors are irrelevant to my calculation. They are what publishers are now getting per book rental, when it’s paper. That other money is spent, but the publisher isn’t getting it.

    That said, it’s possible libraries could pay more if their costs were lower. Ebooks could (in theory) allow them to cut staffing and redeploy the savings to higher prices for books.

    But there are limits. Take the South Portland PL, listed above. Their total budget is $669,291, yet they circulate items 249,431 times. If each circulation cost $2.50, they wouldn’t be able to afford a director, let alone heat, toilet paper or someone to clear the snow. The enormous good they do—ten books per man, woman and child in the town—nearly REQUIRES the First Sale doctrine.