The Agency and Wholesale Models in Electronic Content Markets

I investigate strategic interactions and market outcomes in the “agency model” and “wholesale model” of sales. Adopting the agency model initially raises prices, and can (but need not) raise the profits of rival retailers. Nonetheless, consumers prefer the agency model. I relate my results to events in the market for electronic books. I investigate the “agency model” and “wholesale model” of sales, which are two distinct ways of structuring relations between suppliers and retailers and of determining final retail prices. The wholesale model is very typical: suppliers set per-unit wholesale prices to retailers, who then set retail prices. The agency model is very different: suppliers set retail prices and then split revenue with retailers according to pre-determined shares. Adopting the agency model changes the operative strategic incentives in the market, leading to a different equilibrium market outcome. The agency model can raise retailer profits, which may encourage retail entry and benefit consumers. But even when entry and investment levels are fixed, consumers prefer the agency model—despite the fact that retail prices increase immediately following its adoption. My analysis provides additional insight for regulators assessing markets that utilize the agency model. In particular, my results are relevant for the pending antitrust case in the US, in which Apple and some book publishers (who adopted the agency model) are accused of conspiring to raise prices for electronic books (“e-books”). The EC is also considering legal action. In addition to being recently adopted by e-book retailers Amazon and Apple and publishers supplying them, the agency model is also commonly used by companies that support marketplaces for applications (“apps”) usable on mobile devices such as smartphones and tablet computers, as well as the Amazon Marketplace (a channel for many retail goods). A variant of the agency model is also present whenever a salesperson vends goods of multiple producers who set the market price for their own goods and pay a commission for each sale. Among my main results is that a simple condition determines retailers’ preferences over the two pricing methods: whenever the differentiation between suppliers is sufficiently high I thank Michael Waldman, Henry Schneider, and particularly Yossi Spiegel for his extensive and helpful comments. Johnson Graduate School of Management, Cornell University. Email: [email protected]. The terms “agency model” and “wholesale model” are not mine, but rather are used in practice.