E-books, P-books, and the Durapolist Problem

This proposed paper provides a novel explanation to some controversial recent and historic practices in the book trade. Last year, antitrust regulators in the US and the EU challenged the e-book distribution agreements between Apple and leading book publishers. At the heart of these challenges was the publishers’ concern about Amazon’s discounted pricing for e-books. The lawsuits have alleged that the publishers sought to move away from the wholesale pricing model, which allowed Amazon to buy e-books and resell them at whatever price it chose, and institute an “agency model” instead, under which the publishers set the retail price and the retailer gets a commission. This was not the first time that book publishers are unhappy about discounting resellers, and not the first time that they had colluded to prevent that. A century ago, an industry-wide two-level cartel of book publishers and booksellers was formed in order to prevent the discounting of books. The arrangement was designed to counter the discounting practices of the retail innovators of that time, such as Macy’s, and culminated in two decisions of the US Supreme Court: in Bobbs-Merrill v. Straus (1908), the Court established the first-sale doctrine and limited the ability of publisher to fix the retail price of books, and in Straus v. American Publishers' Ass'n (1913) the Court held that the agreements would violate the antitrust laws. There are remarkable similarities between the turn of the 20th century case and the present one. Both of them involve resellers who revolutionized retail, and whose ability and willingness to sell books at discounted prices was seen by publishers and other retailers as a serious threat. In both cases publishers had to collude among themselves and with some retailers in order to eliminate that perceived threat. There are some differences, of course. The old case involved a retail innovator and collusion between publishers and resellers of yore, whereas the current case involves, in addition to publishers, two innovators (Amazon and Apple) and two innovations (retail innovation and product innovation). Those two incidents, however, are far from unique or isolated. Rather, they reflect a much more common pattern. In fact, similar agreements between publishers and retailers to fix the retail prices of books have been a characteristic of the book trade in many countries for more than a century. In the UK, the Net Book Agreement existed for nearly the entire twentieth century, until its collapse in the mid 1990s (Thompson 2010, pp.50–52), and similar Resale Price Maintenance (RPM) arrangement—often bolstered by legislation—exist in Germany (Bittlingmayer, 1992) and many other European countries (Ringstad, 2004). Yet from an economic perspective, the persistence of systemic to fix the retail price of books—and now e-books—is puzzling. It is easy to see why publishers would be interested in fixing the wholesale prices of books, or why retailers might be interested in fixing the retail prices, and the literature on vertical restraints may explain why individual publishers might be interested in imposing vertical restraints on retailers. But arrangements that combine a wholesale cartel with a retail cartel are rather peculiar, and even more puzzling is the prevalence and endurance of such arrangements in the book trade. After rejecting some possible explanations, the paper will examine the hypothesis that these arrangements reflect the book industry’s solution to the durable monopoly problem (or following Orbach (2004), the “Durapolist” problem). The standard economics of copyright emphasizes the large gap between the high fixed cost of creation and the very low marginal cost of reproduction. Copyright, then, is supposed to encourage creativity by protecting the publisher from competing unauthorized copies. Copyright, however, might be effective in preventing competition from identical copied books (and somewhat effective in protecting the publisher from “substantially similar” competing books), but it is ineffective in protecting publishers from competing, yet non-infringing, books (Katz, 2013). This means that notwithstanding copyright, sustaining supra-competitive prices is a challenge because every new book has to compete at least with numerous other new titles in the same genre, as well as with numerous old books that are often just as good as new titles (think about all the great classic and non-classic novels that you have never read and probably never will). What aggravates this challenge is the fact that a book is a durable good, in the sense that typically a consumer buys it only once and because reading it does not consume it. This suggests, following Coase (1972), that notwithstanding its copyright, the publisher, in effect, is in price competition with itself over time, and that without a workable strategy that allows it to commit to higher prices, the price of books might drop towards marginal cost rapidly. The Coase Conjecture offers an explanation to the historic tendency to fix the retail prices of books through horizontal means, and for the publishers’ anxiety over e-books in general, and Amazon’s pricing in particular. I will show how the various historic RPM schemes for books were designed to generate credible commitments to high book prices. I will argue that committing to such a strategy was easier in the analog era, when the industry relied on printed books and brickand-mortar bookstores, but has become more difficult when printed books are sold online, and might be impossible with e-books. In essence, the business model for the distribution of trade books was based on making new books available at high prices for limited times, and then making them unavailable. Since printing a new edition is costly, waiting until the price drops would be a risky bet for high-demand consumers. Thus, the technological constraints of printing help convincing buyers that production will be limited. Moreover, since easily located brick-and-mortar bookstores have limited shelf space, they necessarily can stock only a limited number of books. The combined effect of those two features is that consumers know that if they don’t buy the book at a high price today, the book may not be available tomorrow, and it may be uncertain if and when a new edition is released. As the paper will explain, the horizontal elements of the pricing schemes were necessary to stabilize them. The growth of the large retail bookstore chains changed some aspects of the system, but did not alter its fundamentals, but the rise of Amazon in the 1990s led to significant changes. Amazon did not face the same physical constraints on shelf-space that brick-and-mortar bookstores face, and more importantly, competing on price and making as many books available as possible, were essential elements of its strategy (Thompson 2010). Amazon’s rapid growth and the resulting bargaining power against publishers made commitment to high prices more difficult, but as long as printed books are concerned, the fixed costs associated with printing each edition may have enabled some commitment to high prices. In addition to showing how the Coase Conjecture explains some key characteristics of the traditional book trade, the paper will show how the move to ‡‡‡‡‡ Except on special order, through libraries, or as second-hand. The first option may not be available at a reasonable price and within reasonable time, the second may be highly inconvenient, and the third may be highly unreliable, or inconvenient, or costly. §§§§§ This is similar to the example of an artist who commits to limited production of a lithograph by destroying the plates (Bulow, 1982). ****** Bulow suggested that “an author may be able to establish that he will not put out a low-cost edition (paperback) for at least a year after the hardcover comes out. However, it may be difficult to convince anyone that a paperback will never come out” (Bulow, 1982; 326). Nevertheless, the expected availability of cheaper paperback editions is merely a variation of the strategy, whereby instead of one opportunity to buy the book there are two. But there is no guarantee that the book will always remain available. †††††† A related change was the emergence of online secondhand retailers, such as Biblio, AbeBooks, or Alibris, which was a matter of growing concern for publishers (Thompson, 2010; 44). This concern is consistent with Waldman’s observation regarding the effect of secondhand markets on durable good monopolists (Waldman, 1997). e-books might have threatened to undermine the trade’s basic fundamentals. In essence, the distribution of e-books does not face the same constraints that facilitated commitment to high prices of printed books. Since e-books don’t have to be printed in limited editions, the marginal cost of an additional copy is always close to zero. This means that there are no physical constraints on the available quantity at the production stage and credibly committing to a limited quantity becomes more difficult. In addition, since e-books don’t have to be stored, shelved, or transported from one place to another, there are no shelf-space limitations or other retail-level constraints on quantity similar to those that exist in physical bookstores. As a result, as long as there is demand, it can be supplied instantly. Therefore, a distribution model based on limited availability of highprice copies might become unsustainable for digital books. These insights may explain some of the recent developments in the book and ebooks trade: from publishers’ reluctance to embrace e-books, to their angst over Amazon’s pricing of e-books. Indeed, their attempt to impose the “agency model” might be explained as an attempt to recreate a distribution system for ebooks that resembles that of printed books, or at least as an attempt to stall the demise of their existing business model in the hope of figuring out another model before others precede them. These insights may also explain publishers’ objections to projects of mass digitization of books, and to proposals to revive orphan works. While the vision of easy access to all books ever written have sparked the imagination of many, it might be the nightmare of an industry whose profitability has traditionally depended on making only a small selection of books available, and only for short periods. I am also planning to discuss some of interesting legal implications of these findings. For example, if commitment to supra-competitive prices of e-books becomes impossible, prices might be set at (or close to) marginal cost even if copyright is enforced perfectly and all piracy is eliminated. This interesting result may suggest that focusing on preventing unauthorized reproduction by legal or technological means might be completely futile, and further suggest that copyright itself might be futile. Alternatively, it may suggest that copyright might still be useful if it can serve as a basis for other business models, not those based on distribution of copies, but on providing convenient access to different sorts of “libraries”. These insights might also have important implications for competition law and policy. Horizontal agreements to fix the retail prices of books are highly suspicious from competition law perspective, and where books RPM schemes were permitted, they have often been justified on cultural rather than economic grounds. But this paper’s hypothesis may invite revisiting these positions. If ‡‡‡‡‡‡ Gans (2012), reaches a similar conclusion, but for other reasons. RPM is indeed necessary for sustaining investment in the production of new books—not an obvious assumption—and if it cannot be sustained without horizontal components, then competition law might have to rethink its position, and book RPM schemes might actually be defended on economic, rather than cultural grounds.