Business Idea #2: eBooks are not Books

After finishing my article on the economic reasons that libraries exist, I briefly considered titling this followup "2030: No More Libraries". But as I've said before, I'm an optimist about the future of libraries, and my message is that a new sort of public library will thrive in 2030, even if the current sort don't survive.

To quickly sum up my previous article, in the 90's, Prof. Hal Varian derived an equation (pdf) that describes when library-like sharing can benefit both producers and consumers. Varian's equation boils down to comparing the transaction cost of sharing to the producer's marginal production cost. When the cost of producing and selling an additional book is higher than a library's cost to loan a book, society will benefit from the existence of libraries. This is the environment that has allowed libraries to thrive for hundreds of years.

eBooks are not like printed books. The cost to produce an additional book is essentially zero. The cost to deliver an additional book is essentially zero. The only marginal costs a publisher is likely to have are author royalties and the financial overhead of the sales channel. Thus the likelihood that public libraries can generate economic value by mere sharing of ebooks is minuscule.

I want to emphasize that I'm not saying that public libraries don't generate economic value, just that libraries generate a lot of value in the print economy that won't be generated in the digital economy. There's been a lot of reaction to a short post by Seth Godin in which he asks "What should libraries do to become relevant in the digital age?" Some of the umbrage is directed as his seeming unawareness of what libraries are already doing in the digital age, but to say, as another blog does, that that the answer is "nothing" is just head-sand-sticking.

Another economic benefit of circulating libraries described by Varian's equations kicks in when consumer access preferences are heterogeneous. This benefit remains in the digital book economy, but oddly, it is tied to the inconvenience of using libraries. To put it another way, libraries help to segment the ebook market only if ebook access through the library is sufficiently inconvenient to shield publishers' higher priced access options. If you don't believe me, go and look at how ebooks are offered at the New York Public Library; you'll have to pretend that ebooks have all the inconvenient characteristics of real books.

Demand aggregation is a function that libraries have fulfilled in the print economy. If 10,000 people are willing to pay $10 for shared access to a book collection, they can buy 1,000 books that cost $100 each. But local libraries lose their advantage in collective acquisition when books become digital because there is no longer a necessity for users to be geographically close to books. Smart publishers will want in on this action, as will other entrepreneurs. As Evan Schnittman, a senior executive at Oxford University Press told me, "Lending models of scale are coming... but I doubt consumers will turn to libraries en masse to get their ebooks, as capitalism has a funny way of turning demand (at any price) into financial opportunity."

There's no doubt there will be collective acquisition of ebooks, the question is the shape it will take. Today I read about the problems of tropical fish farmers in Florida. This group no doubt has an distinctive set of information needs. They need access to information about tropical fish, weather forecasting, pond digging, Florida real estate, agricultural regulations and international trade. Few of them have enough money to access all the types of information they need, and they're too small a group to attract the attention of a publisher-sponsored "ebook-club". As a collective, however, it should be theoretically possible to aggregate their demand, improve information access for members, and increase publisher revenue.

Alas, the Florida Tropical Fish Farms Association would have a number of difficulties in building the ebook library it needs. Making deals with all the different publishers would be prohibitively difficult; technical infrastructure to support the use and administration of such a library may not exist. Perhaps Gluejar can make it possible for communities of interest to nucleate and create libraries of ebooks.

Locally funded public libraries will continue to be efficient demand aggregators to the extent that they focus on the particular interests and needs of their communities. In this respect they will repeat the experiences of academic and corporate libraries that have seen their importance increase with the transition to electronic information. But on the whole, public libraries will need to reprioritize and reduce their budgets, reconfigure their physical assets, scrap expensive infrastructure exclusively focused on inventory management and adopt infrastructure that supports community involvement and community development.

They'll need our support.

Why Libraries Exist

Imagine a world in which you share every book you buy with nine strangers, and you don't buy nine of every ten books you read. What would that world be like?

If you're a librarian, you're probably thinking it would be a reader's paradise. If you're a publisher, you're probably thinking it would be an author's nightmare. If you're an economist, you're probably computing demand curves, and concluding that both librarian and publisher have got it completely wrong.

Demand curves are used by economists to characterize the market for products. They plot the the amount the market would buy for a given price. For example, the market for a book would show that many people will buy books at a low price, and few will buy the book if it has a high price. If you are an exclusive seller in a market, as most publishers are, you can choose any price you like, but your revenue will be maximized if you choose a price somewhere in the middle of the curve. For the demand curve shown, the best price for the book is $10, which results in 1000 sales, for a revenue of $10,000.

What happens to the demand curve in our hypothetical everybody-shares world? The sales go way down of course, but the people willing to get the book for $1 will get to read books priced at $10. The people willing to spend $10 will be able to read the book even if it's priced at $100.

If the publisher took economics in college, he would look at the demand curve and price the book at $100 instead of $10, and guess what? The publisher revenue would be $10,000, or exactly the same as before! Since the publisher doesn't have to incur the printing costs of the larger print run, he makes more profit.

Is this realistic? Librarians can't have missed the fact that books meant for the library market are invariably priced at five times what the book would command in the trade market. On the other hand, if publishers were better at math, well... they'd have become bankers, and we'd probably all be better off in 2010 than we are!

The real world makes things complicated, and economists like UC Berkeley's Hal Varian studied these situations in the 90's. They wrote lots of interesting articles, including ones filled with math (pdf) and others fit to be read by librarians and publishers. Varian included the effects of transaction costs, production costs and the different values of owning and sharing, and found that library-like sharing benefits both publishers and consumers when the transaction cost of sharing is less then the marginal production cost:

1) more books will be read; 2) consumers will pay a lower price per reading; 3) the sellers will make a higher profit; and, 4) consumers will be better off.

To put it another way, libraries can be economically justified if the cost to lend a book is less than the cost to produce and sell a book. As I discussed in my previous article, the Institute of Museum and Library Services publishes a survey (available here) that tells us that US public libraries performed 2.17 billion circulation transactions in 2007 at an operating cost of $8.86 billion dollars. If we ignore all the other services that libraries provide, that gives us a cost per transaction of $4.09. So public libraries are inherently beneficial if it costs more than $4.09 for publishers to make, sell, and deliver an extra book.

Varian also notes that consumer sharing also yields benefits when consumer preferences are heterogeneous. Some people want to read the book the day it comes out; other people are perfectly happy to wait until the paperback comes out, and others will prefer to get it from the library. This heterogeneity allows market segmentation, which improves the efficiency of the market. The book publisher can charge $30 for the hardcover edition, but can still make money off of people who are only willing to pay $10 by issuing a paperback a year later. The publisher even makes money from library sales that allow use by people not willing to pay anything for book!

Libraries benefit society in many ways, but it's important for both publishers and librarians to understand the economic role that libraries have played in the book industry- they benefit everybody, publishers, authors and readers, by aggregating demand and helping to segment the market.

Next... ebooks.