Lemley & McKenna: Is Pepsi Really a Substitute for Coke? Market Definition in Antitrust and IP

Traditionally, antitrust law has viewed product markets as relatively static domains where products compete solely based on price and quality. But intellectual property rights complicate this simple picture. For example, while basic intuition would make one assume that Pepsi and Coke compete in the same market (the market for colas or sodas, for example), reality suggests otherwise. Though research indicates that consumers prefer Pepsi over Coke in blind taste tests (better quality), consumers still prefer Coke over Pepsi when the blinders are removed—the “Coke” trademark, on its own, differentiates the product such that Coke and Pepsi may, in reality, compete in separate markets. Using this observation as a springboard for their recently-released working paper, Professors Mark Lemley of Stanford Law School and Mark P. McKenna at Notre Dame Law School ask us to consider the impact of IP rights on market power generally; and, in so doing, they provide some useful insights about the use of market definition in IP cases.

Noting that market definition is an inescapable or implicit part of various IP analyses, including the functionality inquiry in trademark (requiring one to determine elements of a product that are “essential to competition”), lost profits and non-obviousness determinations in patent, and analyses of fair use and the idea-expression dichotomy in copyright, the authors express concern that judges do not employ any clear methodology in defining markets in these IP contexts. On the functionality front, for example, one court denied trade dress protection to “Dippin Dots” ice-cream, holding that the ice-cream design was functional and part of a distinct product market for flash frozen ice cream (and not part of a broader ice cream market); in contrast, another court determined that a “Ring Pop” ring-shaped lollipop was non-functional and part of a broader market for lollipops. And Judge Kozinski, after considering the idea-expression dichotomy, famously denied copyright protection to Bratz dolls on the ground that such dolls fell on the “idea” side of the dichotomy—specifically, he believed that there was a separate market for “fashion dolls with a bratty look or attitude.”

These rulings may have all been based upon reasonable intuitions. But the opinions’ rationales for adopting various market definitions, to the authors, appear arbitrary. Thus, Lemley and McKenna suggest that judges assessing market definition in such IP contexts take a page from antitrust, which has long-standing methodologies for defining markets. Antitrust is willing to consider supply substitution, cross-elasticities of demand, and other economic analyses in deciding whether products are in the same market. At minimum, there should be some inquiry into the actual interests of consumers when assessing market definition in these IP contexts.

Meanwhile, the authors make clear that traditional antitrust market definition methodologies are not entirely suitable in the IP frame—thus, they advocate a more flexible market definition in this space. They note that antitrust often thinks of market definition in a binary sense—either a product is “in” the market or it is “out.” But this oversimplifies the issue, as the question of whether or not a product competes with another product is a matter of degree. For example, courts have found in the antitrust space that brand name drugs and their generic equivalents are in separate markets. This, however, appears to be a bold conclusion given that (1) the generic product probably has at least some influence on the price of branded drugs; and (2) we frequently see efforts by branded pharmaceutical manufacturers to block the release of generic drugs. If market definition and product differentiation were assessed along continuums—rather than in yes/no buckets—we would move towards a better understanding of markets, viable substitutes of products within those markets, and the appropriate outcomes in various cases.

While market definition is important in analyzing the implications of various IP rights, Lemley and McKenna also highlight various IP doctrines that could be tweaked to enhance or diminish competition in markets involving IP. In copyright, doctrine could require much more “substantial similarity” between a potentially infringing work and the copyrighted work before upholding an infringement claim; narrow the derivative works right; or broaden the fair use rights of competitors to enhance competition. In trademark, the likelihood of confusion test could be applied more stringently. And in patent, claim construction can be narrowed, “means for” function claims can be constrained, and the experimental use defense can be broadened. Of course, such doctrinal adjustments are necessarily policy judgments—as they indicate that the costs of certain IP protections are “too high” or “too low” to warrant particular doctrinal adjustments.

Ultimately, Lemley and McKenna seek to emphasize that we need to acknowledge the inherent tensions between IP and antitrust. They add a valuable contribution to the literature by placing the courts on notice of various areas of IP that implicate market definition; if we are to take economic analysis seriously in the antitrust space, there is no reason to abdicate that responsibility in the IP space. By emphasizing that consumer demand may suggest that products like Pepsi and Coke may actually be in separate markets, meanwhile, the authors allude to aspects of behavioral economics that could be applied to IP analyses of market definitions. There may be certain cognitive biases that seem fanciful—even irrational—that we need to analyze when assessing the anticompetitive character of certain IP rights. Meanwhile, the authors provide policymakers and judges with some doctrinal levers that could be employed to enhance competition in markets that are often affected by intellectual property. Of course, such levers should not be pulled without a better understanding of the markets that will be affected by such policy changes.

This leaves us with several difficult questions. The less rigid market definition analysis proposed by the authors will be complex and difficult; given the “continuous” nature of this analysis, it may be difficult to articulate judicially manageable standards in specific cases. Specifically, if product differentiation is a matter of degree, how can judges assess infringement claims and other IP doctrines, which generally produce yes/no answers? Will the market analyses merely be used to assess degrees of liability in damages calculations? Will certain thresholds of substitutability need to be crossed to confer liability in specific cases; in different markets?  Additionally, such market definition analyses will increase litigation costs where they are employed, and there is a question of whether the benefits of these analyses will outweigh the costs to the litigants and to consumers. It is possible that the seemingly ad hoc, common-sense analyses of the past should remain the rule of the day. Nevertheless, the authors have done an excellent job in starting a much-needed conversation, and one thing is certain: IP rights add a fascinating wrinkle to our traditional understanding of market definition, and we need to better understand how they affect competition—and consumer behavior—in various IP spaces and markets.