Associate Professor, Notre Dame Law School. Thanks to Barton Beebe, Laura Bradford, Eric Claeys, Stacey Dogan, Abraham Drassinower, Laura Heymann, Ariel Katz, Mark Lemley, Bill McGeveran, Adam Mossoff, Rebecca Tushnet, and participants at a faculty workshop at the George Mason University School of Law, the University of Toronto Innovation Law & Theory Workshop, and the Eighth Annual Intellectual Property Scholars Conference, all of whose comments improved this paper. Thanks also to Erin Czerney and Merideth Mendenhall for outstanding research assistance.
Modern scholarship takes a decidedly negative view of the scope of trademark protection. Commentators decry recently developed confusion doctrines like initial interest confusion, which they regard as evidence that trademark law no longer works for the benefit of consumers.1 Whole conferences convene to address dilution,2 the innovation many contend most clearly crystallizes trademark law's over-expansion. And scholars frequently criticize trademark law for its deleterious impact on free speech, focusing most frequently on the outer edges of trademark law.3
Whatever the merits of these criticisms—and many of them do have merit—their focus on dilution and other modern doctrines is hugelyPage 67 disproportionate to the practical significance of those doctrines.4 More importantly, all of the criticism of these marginally significant doctrines obscures overwhelming acceptance of the most important determinant of the scope of trademark protection—the modern likelihood of confusion standard. As Robert Bone stated succinctly, "No one doubts [that the] central function of protecting trademarks [is to] benefitf] consumers by preventing deceptive and confusing uses of source-identifying marks."5
Bone is certainly right that few have questioned this "central function." But easy acceptance of the harmfulness of confusion is actually a problem because it operates at too high a level of generality, masking important differences among the types of relationships about which consumers might be confused. In fact, I argue that modern trademark law's excesses are primarily the results of courts' failure to differentiate among those different relationships. Mark owners are able to characterize almost every conceivable use of their trademarks in consumer confusion terms, and because courts have simply equated confusion with harm, they have lacked the tools to resist novel confusion claims.
In order to begin the task of identifying meaningful conceptual limits on the scope of trademark protection, this Article disaggregates consumer confusion and focuses on the alleged harms to producers from confusion regarding the source of non-competing goods. In particular, it evaluates common arguments about the consequences of this type of confusion for mark owners—arguments which sound in empirical terms but which havePage 68 never been empirically supported—against a growing body of marketing literature dealing with brand extensions and brand alliances.
There are two parts to this framing, both of which deserve some explanation. First, is the focus on cases in which the defendant's goods do not compete with the plaintiffs. Here I mean to differentiate between a claim by Nike, Inc. against another company selling NIKE shoes and a claim against a company using the NIKE mark for other goods or services, and to focus on the latter.6 This subset of trademark infringement cases merits specific attention because the harm to producers from confusion about the source of competing goods is relatively clear: a competing NIKE shoe company could use consumer confusion to divert customers who otherwise would have bought shoes from Nike, Inc.7 But since non-competitive uses do not pose the same risk of trade diversion, trademark protection against those uses requires a different justification. Courts' accommodation of claims against such non-competitive uses therefore reflected not simply a modest expansion of trademark law but a significant conceptual shift. Such an important and radical change deserves closer scrutiny than it has received, particularly because this conceptual change is responsible for most of the competitive and speech costs of modern trademark protection.8
Second, is the focus on producer-based justifications for claims against non-competing goods rather than the more conventional consumer-interest justifications.9 I focus on producer interests because that was trademarkPage 69 law's traditional orientation—nineteenth-century trademark law focused on producers' interests in preventing trade diversion10—10 and because the case for expanding trademark protection to non-competing goods was based almost entirely on claims about harms to mark owners.11
But this focus on producer interests is not intended to deny the influence of the consumer-interest narrative in shaping modern trademark law. Nor is it intended to deny that, in many cases, consumer and producer interests overlap. Instead, by highlighting the producer-based arguments, I hope to illustrate how commentators' singular focus on consumer search costs fails to capture the complex relationship between the consumer-oriented rhetoric and producer-oriented doctrinal structure that has developed in the last half century. Courts in the mid- to late-twentieth century reasonably understood trademark doctrine as producer-oriented, and they shaped doctrine explicitly in response to producer-oriented arguments.12 As time went on, however, the consumer interest narrative gained traction, and courts and commentators increasingly layered consumer search cost rhetoric on top of the producer-oriented doctrine. This layering enabled even further trademark expansion, as mark owners successfully mixed producer- and consumer-based arguments in support of their newest claims, often taking one from column A and two from column B, and avoiding rigorous scrutiny of any of the arguments.
This Article separates out the producer-based arguments analytically so that those arguments can be evaluated on their own merits. In particular, it evaluates the assumptions underlying the producer-harm arguments in light of the available evidence from marketing literature regarding brand extensions and alliances. This literature focuses on the effects of certain branding strategies, particularly uses of known brands for new products and joint efforts to produce products or offer services. Because these studies make explicit that the extension or alliance products come from the senior mark owner, they offer valuable insight into the consequences to mark owners even if a. particular use is likely to cause confusion.
The marketing literature paints a complex and sometimes contradictory picture of consumer evaluation of brand uses. On the whole, however, one thing is clear: no presumption of harm from uses of a mark for non-competing goods is warranted. Indeed the only sense in which one could confidently conclude that mark owners are likely to be "harmed" by uses for non-competing goods is that the later uses may interfere with the mark owner's ability to expand into new markets. In other words, the literature suggests that, from a producer's perspective, trademark protection against non-competing goods is analogous to a derivative work right: it serves primarily to protect mark owners' ability to use their marks in ancillary markets, not because non-competitive uses cause mark owners harm, but because of a sense that those ancillary markets rightfully belong to the senior mark owner.13
Market preemption arguments along these lines have long been offered in support of trademark rights against non-competing goods. But because they have generally been articulated as additional support for other primary justifications that focused on quality attribution, and because most modern commentators evaluate trademark law from a consumer search cost perspective, trademark theorists have not been forced to confront squarely the types of incentive-based arguments typically associated with copyright generally, and the derivative work right specifically.14 This paper suggestsPage 71 that, given the lack of strong empirical support for quality feedback arguments, market preemption and free-riding arguments need to be taken seriously if producer-based arguments are going to continue to influence the scope of trademark protection.15
If the market preemption and free-riding arguments are not persuasive, then consumer-based arguments ought to take on much greater relative significance with respect to claims...