The Federal Trade Commission (“FTC”) and 17 state attorneys general on September 26th filed their long-awaited antitrust complaint against Amazon.com in federal district court in the Western District of Washington. The complaint (a redacted version of which is available here) is a foreseeable ambition of FTC Chair Lina M. Khan, who as a Yale law school student authored an influential law review article titled, “Amazon’s Antitrust Paradox.” The article itemized conduct Ms. Khan characterized as “anticompetitive,” which, when committed by a dominant firm with market power, can be unlawful under the antitrust laws.
Establishing in court whether Amazon’s conduct is “anticompetitive,” or that Amazon possesses market power, can pose challenges, particularly in a “new economy” industry such as Amazon’s digital marketplaces for goods and marketing services. The court’s analysis is likely to depend heavily on the FTC’s definition of the relevant antitrust markets in which it claims Amazon possesses market power and harms competition.
The complaint defines two relevant markets: the “online superstore market” and the “online marketplace services market.” The FTC alleges that Amazon is a monopolist in both markets and engages in anticompetitive conduct to maintain its two monopolies (Counts I and II). Although both counts allege violations of Section 2 of the Sherman Act, which prohibits monopolization, they are drafted principally as violations of Section 5(a) of the FTC Act, which makes unlawful all “unfair methods of competition in or affecting commerce, and unfair or deceptive acts or practices in or affecting commerce.”
The complaint alleges that “a core strategy” pursued by Amazon to maintain its monopolies is to punish sellers on its platforms that offer lower prices anywhere other than on Amazon. As a result of these “anti-discounting” policies, the complaint alleges, Amazon can deprive rival online superstores of sufficient scale and scope to challenge Amazon. Amazon, of course, will argue that its policies are not anticompetitive at all, but merely requirements meant to ensure that Amazon customers are offered the lowest available price.
Faced with these contrasting viewpoints, the court is likely to conduct a detailed antitrust analysis, starting with the alleged relevant markets. The market described in the complaint as “online superstores” seems to describe a particular kind of multiline retail distributor. This is an economically coherent relevant market as long as it makes sense to consider the cross price-elasticity of demand as between two or more candidates for inclusion in such a market. The (theoretical) cross elasticity variable comes into play in the hypothetical monopolist test. If a hypothetical monopolist in a candidate market can profit from a non-transitory price increase of, say, 5%, the candidate market is a relevant antitrust market. But if a price increase causes enough buyers to switch to some other product or service, that product or service belongs in the relevant market and the candidate market definition must be enlarged.
Here, the FTC and the states seem to focus on competition between multiline online retailing websites. The key issue is whether a price increase on Amazon could drive online shoppers to Google Shopping, Walmart, Target, Home Depot, BestBuy, eBay, Wayfair, or any number of other online retail destinations that are potential substitute outlets for products sold on Amazon.
To be sure, there is ample precedent for defining a particular type of distribution channel as a relevant market. For example, in FTC v. Sysco Corp. and US Foods, 113 F. Supp. 3d 1 (D.D.C. 2015), the court accepted a market defined as “broadline foodservice distribution services.” In that case, the court found that hotels and restaurants were confronted with only a limited choice of distributors who could offer their customers centralized ordering of a wide variety of foods and supplies that such institutions regularly purchase. It is reasonable to test whether customers of broadline distributors would have anywhere else to go if a hypothetical monopolist imposed a non-transitory price increase. Similar reasoning applies to the “office superstore” mergers, in which the relevant market is comprised of multi-line office supply stores that cater to large organizations that regularly purchase a broad line of office supplies.
In the Amazon case, the court is going to want to know the economics behind the alleged “online superstore” market definition, so it is likely to examine whether the definition makes economic sense when it comes to retail sales to the general public. For large offices and institutions with recurring supply orders, it is clear why the choice of supplier can depend on whether the seller is a broadline distributor that offers one-stop shopping. It is far less clear that consumers care whether Amazon is a “superstore.” If the significance of Amazon’s scope is solely to increase the likelihood that Amazon will be able to sell to a particular consumer, then being a “superstore” means little more than Amazon is likely to carry the sought-after product and Amazon being a superstore offers no additional efficiency to the consumer beyond any other online seller of the product. Since the fact of being a superstore may be of little or no importance to an individual consumer for a given transaction, the court may look skeptically on the superstore market definition and conclude that narrow-line online retailers also belong in the relevant market because the cross elasticities that matter occur on a product-by-product basis.
To make the point another way, starting with Amazon as the only firm in an initial candidate market, how will the FTC establish whether or not few enough customers would purchase elsewhere in the event of an (across-the-board) non-transitory price increase so that such an increase would be profitable for Amazon? Surely, an impossible task on the individual product level and a poor proxy for the “price” of using one online superstore as opposed to another, which is the market in which Amazon is alleged to be unlawfully maintaining its monopoly.
The conduct attempted to be reached by defining the superstore market, Amazon’s policies of prohibiting price discounting on non-Amazon sites, however, is likely to be harmful to competition even if the alleged market definition fails. Enter Section 5 of the FTC Act, which offers a path for the court to condemn the conduct without necessarily finding that Amazon is engaged in maintaining a monopoly or has violated Section 2 of the Sherman Act. Of course, this will depend on whether the court believes that Section 5 establishes a violation for conduct that does not itself violate Section 2 or any of the other antitrust laws. Thus, the Amazon case may be a crucial test of the usefulness of Section 5 to reach anticompetitive conduct in circumstances in which monopolization cannot be proven by a preponderance of the evidence.
The second market definition in the complaint, for “online marketplace services,” attempts to support the claim that by rewarding sellers that use Amazon fulfillment services and punishing those that do not, Amazon harms competition among providers of online marketing services. Even with the fallback of Section 5, this claim is likely to face substantial challenges. While the claim superficially resembles a tying claim, in which a monopolist in one market forces its customers to purchase a product in a different market, there would appear to be little to distinguish Amazon’s conduct from everyday “bundling” of related goods or services. To plead a tying claim, the complaint must define two markets, one that is monopolized by the defendant and another that is competitive but harmed by the monopolist’s tying conduct. Having alleged only a single market, however, the complaint confines itself to the theory that Amazon’s bundling of fulfillment services has the effect of maintaining its monopoly in the marketplace services market, not that it harms competition between providers of fulfillment services. Here, it is unlikely that even Section 5 will save the day for the agency and the states.
Traditional antitrust economics faces significant challenges grappling with the new economy industrial environment, and the Amazon case is likely to be a crucial test for antitrust and the usefulness of Section 5 of the FTC Act.