In a case that has attracted worldwide attention, Apple and five book publishers were sued this spring by the U.S. Department of Justice on the grounds that they colluded to fix prices for e-books sold on Apple's iBookstore website. The government suit charged that by adopting a sales strategy known as "the agency model," Apple and the book publishers (HarperCollins, Hachette, Macmillan, Penguin and Simon & Schuster) reduced competition in the e-books industry, raising prices and harming consumers. Although three of the five book publishers have since signed settlements with the DOJ, Apple, Macmillan and Penguin are still preparing to defend their cases in court.
The case has cast a bright spotlight on the agency model -- otherwise known as the "platform model" -- of e-commerce, in which an increasing range of manufacturers are selling their products directly to consumers over the web, while setting their own prices for those products. The agency model differs from the more traditional "wholesale model" of e-commerce in which "e-tailers," such as Amazon.com, purchase goods from manufacturers and then resell those goods to consumers at prices that the e-tailers themselves set.
The legal controversy over Apple's efforts to use the agency model when selling books directly to its consumers offers a good opportunity to examine the benefits and costs of adopting that approach, both for online retailers and their customers, notes Wharton marketing professor Z. John Zhang, who has done numerous studies in this area with co-authors Vibhanshu Abhishek, a current Wharton PhD student, and Kinshuk Jerath, a professor at Carnegie Mellon University. Does the agency model of selling online harm consumers -- and society in general -- by raising prices for numerous products, not just for e-books? Or might the agency model offer significant, if often overlooked, economic and social benefits over the wholesale model for manufacturers, retailers and the general public? The researchers' most recent paper exploring these questions is titled, "To Platform-Sell or Resell? Channel Structures in Electronic Retailing."
Selling More Kindles
The origins of the conflict date back to the early days of the e-book industry, when Amazon.com, one of its pioneers, used the wholesale model to sell all of its e-books at a price of $9.99, says Zhang. "Amazon was using e-books as a 'loss leader,' apparently because it calculated that if it made e-books cheap enough for consumers, more people would buy its Amazon Kindle," the only hardware device on which Amazon's e-books could be read. Amazon's longer-term strategy seemed to be based on the idea that as Kindle's share of the e-book reader market expanded, the company would eventually make more money from the device and control the gateway to the e-book market. However, book publishers grew fearful that if Amazon.com charged only $9.99 for all of its e-books, they would be unable to charge more than that for their own e-books. "Publishers worried that Amazon might also [force them] to lower their prices on e-books at some point, as the mammoth Walmart does to its suppliers," Zhang adds.
That's why the book publishers got together, as alleged by the Justice Department, with Apple to adopt the same agency model that Apple has traditionally used for selling its own software applications -- rather than the wholesale model used by Amazon. Apple software application developers have long set their own online prices for their products, while giving Apple a 30% cut of the gross proceeds. Thanks to the agency model agreement between Apple and the publishers, the publishers were now able to set their own prices for their e-books on Apple devices, just as Apple has set its own prices for software.
Predictably, Apple e-books became more expensive than those sold by Amazon.com, and book publishers began urging Amazon to adopt the same agency model to sell their e-books on Amazon.com. In 2010, Amazon gave in to the pressure, and e-book prices on Amazon.com began to rise as a result. Many buyers of e-books on Amazon.com were irked by the price increases that ensued.
Avoiding a 'Huge Markup'
On balance, though, has the agreement between Apple and these publishers really hurt consumers? Or does the agency model have hidden benefits that the Department of Justice has overlooked or disregarded? To address such issues, Zhang and his collaborators have done extensive economic analyses, some published and some not, on the impact of the agency model on competition. A starting point in their research, notes Zhang, was their realization that the wholesale price model used by traditional e-tailers "is not very efficient" to start with because both the wholesaler and the retailer want to maximize their own profits, leading to a big markup in the prices paid by customers. "Economists agree that the huge markup doesn't help consumers or manufacturers or even retailers [including e-tailers]," he says. "Lots of economists identify this problem with the wholesale model as the 'double marginalization problem.'" When this model is used, "end prices are too high to maximize channel profitability. The interests of the manufacturer and retailer are not aligned in a way that maximizes social welfare, either."
For years, people have been trying to find a way to align those conflicting interests. One obvious solution is for a manufacturer to integrate forward -- that is, opening its own stores and selling its own products directly to consumers. Apple does just that, thus eliminating retail price markups. However, few companies can do this in the world of brick-and-mortar stores, notes Zhang, because "most manufacturers lack the resources" to set up their own stores, and most manufacturers -- unlike Apple -- "don't produce enough varieties of products to sell them in their own store."
If, for example, you sell only toothpaste, you can't open a store selling just toothpaste, even if you make it in several flavors. The best most manufacturers could do when selling through traditional channels is to use a so-called store-within-a-store format -- essentially, an agency model for brick-and-mortar retailers. This format is widely adopted for virtually all product categories in department stores in Asian countries such as Japan, Singapore and China. According to Zhang, there are telltale signs that U.S. brick-and-mortar retailers are also moving in this direction.
Despite the negative image of the agency model generated by the headline-grabbing lawsuit, the Internet has been a boon for this approach, notes Zhang. Apple's "platform" (or agency) model for apps allows software developers to run their own stores and sell their own products directly to consumers. "In theory, the agency model is a very good way to solve this double marginalization problem because there is only one layer," he points out. When, for example, a book publisher pays a constant 30% fee to Apple, "this doesn't distort the price at which the publisher sells to the end user." That's because when the publisher gets 70% of the pie, his interest -- like that of his agent (in this case, Apple) -- is to increase the size of the pie, not to increase his own share of it. When the channel members all try to get a bigger share of the pie, the pie shrinks and everyone suffers, including consumers.
On the Internet, it is much easier for all kinds of manufacturers to implement the agency model -- not just giant manufacturers like Apple, which have the product variety and strong brand to sell directly in their own online stores. A growing number of smaller manufacturers have been selling more and more of their products at competitive prices directly to customers over e-commerce platforms managed by such sites as Alice.com, Etsy.com and even Amazon's own Amazon Marketplace. Rather than emerge weakened as a result of the Apple e-book controversy, "over time, the agency model has a good chance to win the day. In fact, it should win the day for the good of the economy and for consumers," adds Zhang.
Short End of the Stick?
Why then, in the case of e-books, does the Department of Justice want to stand in the way of the agency model? For many, the most objectionable outcome of the alleged collusion between Apple and the publishers is that prices for Amazon's e-books -- and, in fact, e-books in general -- increased, thereby harming consumers. However, Zhang questions whether this argument examines all of the ways the agency model has had an economic impact on the market for e-books. For example, although prices of Amazon e-books have risen, prices of Kindle readers also went down over the same period. "You can't say that the consumer only gets the short end of the stick as a result of the agency model. You have to balance all of the interests of consumers."
Moreover, he suggests, the first wave of people who purchased Kindle readers were often more affluent consumers who were less sensitive to price and possibly more willing and able to pay somewhat higher prices for e-books. "For those people, there was very little impact when Amazon raised its e-book prices." Society as a whole also gained when consumers who could finally afford a now-cheaper Kindle gained access to the thousands of e-books still available at little or no cost on Amazon's website. Indeed, with the agency model taking hold in the publishing industry, "you will see that iPad, Kindle and Nook will all come down in price, competing even harder for new readers," notes Zhang.
For their part, publishers also have a valid argument that if e-book publishers were restricted to using the wholesale model preferred by Amazon, consumers might eventually become unwilling to pay the higher prices Amazon might charge after it gained more and more share of the e-book market. "If Amazon became successful in building its market share, it could become so strong that it could go to the publishers and dictate the prices for e-books," Zhang points out. If that happens, "who will have the incentive to produce e-books with the most advanced technologies," such as videos, interactive elements and links to the Internet? "All of these things cost money, and publishers want room to make investments in the technology they need in order to be innovative. The pricing flexibility and discretion they gain through the agency model will give them the room they need."
Initially, the agency model "may lead to higher prices for e-books, as we have observed," Zhang concedes. This is because Amazon has been using loss-leader pricing to achieve its market share objectives. In addition, the agency model does have a tendency to reduce inter-store competition: If a publisher sells a book through both Apple and Amazon, "you can bet the publisher will not try to lower its price at Amazon to cut into its sales of the same book at Apple." However, in that case, Zhang and his collaborators have shown that all bets are not off on price competition, as the agency model opens up inter-publisher price competition. "To sell a book, [one publisher] will have to compete much harder with other publishers with similar titles for readers."
In their most recent paper, Zhang and his coauthors address a key question that e-tailers are now facing: When should they use a platform -- or agency -- model instead of the more conventional reselling format? In developing a theoretical model for their study, the researchers focused "on the effects of two main factors on the resulting selling format in electronic retailing: competition among e-tailers, and reaction by the manufacturer due to the impact of the electronic channel on sales in the traditional channel (brick-and-mortar retailing)." The results of their work suggest that "whenever sales in the electronic channel lead to a negative effect on demand in the traditional channel, e-tailers prefer to set up platforms, whereas when sales in the electronic channel lead to a substantial stimulation of demand in the traditional channel, e-tailers prefer reselling contracts with manufacturers." Furthermore, "this preference is moderated by competition among e-tailers -- as competition between them increases, e-tailers prefer to set up platforms."
The researchers note that platforms have become widespread in industries as varied as consumer packaged goods (Alice.com), travel (Expedia, Travelocity) and vintage goods and arts (Etsy.com), as well as the "marketplaces" on Amazon Marketplace and Marketplace at Sears.com. They warn marketers, however, that "online retailing is still in flux," and that although the agency model has "become a pervasive phenomenon, the understanding of this phenomenon is limited." "This is one more reason why we should be extra careful not to throw the baby out with the bath water," Zhang says.