INTRODUCTION I. THIS IS NOT MY BEAUTIFUL MARK II. WELL, HOW DID WE GET HERE? III. MATERIAL AND IMMATERIAL CONFUSION A. Justifications for Expanding Confusion 1. Consumers, producers, and the quality of unrelated goods 2. Uses that confuse consumers about quality 3. Consumer perceptions in pure sponsorship cases B. Costs of Expanding Confusion C. Weighing Costs and Benefits IV. DISTINGUISHING TRADEMARK INFRINGEMENT FROM FALSE ADVERTISING V. WELL, HOW DO WE WORK THIS? CONCLUSION
In 2006, thousands of soccer fans showed up to the World Cup game between the Netherlands and the Ivory Coast wearing pants in the colors of the Dutch national team. The pants had been given out as promotional gifts by a beer company. FIFA, the governing body of international soccer, objected. It claimed trademark rights in the team colors, and giving out pants in those colors was in FIFA's view "ambush marketing" that was likely to confuse those who saw (or even those who wore) the pants into thinking that the soccer team had sponsored the pants. And in FIFA's view, not only was giving out the pants illegal, but individuals wearing them were falsely suggesting some affiliation with the Dutch national team. (1) Prohibited from wearing the pants into the stadium, more than one thousand fans dutifully took their pants off and cheered the Dutch team to victory in their (largely orange) underwear. (2) This was Europe, after all, and it was an important match.
Trademark law centers its analysis on consumer confusion. With some significant exceptions, the basic rule of trademark law is that a defendant's use of a mark is illegal if it confuses a substantial number of consumers and not otherwise.
As a general matter, this is the right rule. When it works well, trademark law facilitates the workings of modern markets by permitting producers to accurately communicate information about the quality of their products to buyers, thereby encouraging them to invest in making quality products, particularly in circumstances in which that quality wouldn't otherwise be apparent. If competitors can falsely mimic that information, they will confuse consumers, who won't know whether they are in fact getting a high quality product. Indeed, some consumers will be stuck with lemons. (3)
Unfortunately, as the FIFA case illustrates, trademark law has taken the concept of confusion too far. Over the middle part of the twentieth century, courts expanded the range of actionable confusion beyond confusion over the actual source of a product--trademark law's traditional concern--to include claims against uses that might confuse consumers about whether the trademark owner sponsors or is affiliated with the defendant's goods. This expansion began for plausible reasons: consumers might be confused to their detriment in at least some cases in which the plaintiff and the defendant do not actually compete directly. But sponsorship and affiliation confusion has taken on a life of its own, leading courts to declare as infringing a variety of practices that might be confusing in some sense, but that do not affect consumers' decision-making process.
We think trademark law needs to refocus on confusion that is actually relevant to purchasing decisions. Specifically, it should anchor once again to the core case of confusion regarding the actual source of a defendant's product or service, the type of confusion most obviously related to consumer decision making. Most cases of confusion regarding actual source will involve competitive goods, but consumers may also mistakenly believe a mark owner is the actual source of noncompetitive products that are closely related to the mark owner's. We think genuine source confusion causes the same problems whether or not the parties' goods compete directly, so trademark law should treat as infringing any use that is likely to cause confusion about the actual source of a product or service.
Some uses now viewed through the lens of sponsorship or affiliation raise concerns analogous to those posed by source confusion, particularly those that are likely to cause consumers to believe that the trademark owner stands behind or guarantees the quality of the defendant's goods or services. Even if consumers understand that individual franchisees, rather than the McDonald's Corporation, actually make their hamburgers, they are likely to expect that McDonald's assures the burgers' quality. But in those cases it is the fact that consumers believe the brand owner guarantees the quality of the product that leads to consumer harm if their belief is misguided. We therefore would define the category of trademark infringement to include cases involving confusion as to whether the plaintiff is responsible for the quality of the defendant's goods or services in addition to those involving actual source confusion. And because we believe these types of confusion will impact consumer decision making with sufficient regularity, we argue that courts should presume materiality in cases that fit in the trademark infringement category.
Cases that involve allegations of other forms of confusion, many of which are now lumped into the "sponsorship or affiliation" category, should not be regarded as trademark infringement cases. This does not mean, however, that no other forms of confusion should ever be actionable. To the contrary, we believe the law should regulate some statements that create confusion regarding other types of relationships, but that claims directed to this other type of confusion should be analogized to false advertising claims. False advertising law, like trademark law, is designed to protect the integrity of markets by allowing consumers to rely on statements made by sellers. While trademark law prevents competitors from misrepresenting the source of their products by mimicking another's brand name, logo, or trade dress, the law of false advertising prevents false or misleading statements about the nature or qualities of one's own or a competitor's products. We think uses that cause confusion about things other than control over quality are more like the statements regulated by false advertising law than those traditionally regulated by trademark law. Thus, uses that cause non-quality-related confusion should be treated more like false advertising.
Importantly, false advertising law requires the plaintiff to demonstrate that the misrepresentation is of a particular type (4) and that it is material--that it is likely to affect consumers' purchasing decisions. We think requiring proof of materiality is desirable in the absence of confusion regarding responsibility for quality, both because confusion regarding other types of relationships is less likely to impact consumer purchasing decisions and because these claims are responsible for so much of the cost of trademark law. Indeed, we think it is no accident that the cases that pose the greatest threats to speech interests are sponsorship or affiliation cases, as are many of the most troubling cases involving new technologies, like the recent suits against search engines for returning paid search results in response to search queries involving trademarks. (5)
Our argument unfolds as follows. In Part I, we identify a number of examples of "confusion" alleged by mark owners, many of which courts have found actionable even in circumstances in which that confusion was unlikely to matter to the operation of the market. Part II explains how we arrived at this unfortunate pass. In Part III, we argue that courts can begin to rein in some of these excesses by focusing their attention on confusion that is actually relevant to purchasing decisions. Uses of a trademark that cause confusion about actual source or about responsibility for quality will often impact purchasing decisions, so courts should presume materiality and impose liability when there is evidence such confusion is likely. Uses alleged to cause confusion about more nebulous relationships, on the other hand, are more analogous to false advertising claims, and those uses should be actionable only when a plaintiff can prove the alleged confusion is material to consumers' decision making. We address the scope of such false advertising-like claims in Part IV. We continue the discussion in Part V, which explores some of the implications of distinguishing between different types of confusion and conceiving of some claims as false advertising rather than trademark infringement claims, and we discuss how to handle some close cases.
I. THIS IS NOT MY BEAUTIFUL MARK (6)
Pantsless soccer fans (and those sitting next to them) are far from the only victims of the broad modem conception of sponsorship or affiliation confusion. In 2008, Major League Baseball began to crack down on the longstanding practice by local Little Leagues of naming kids' baseball teams after major league franchises. MLB's theory was that people watching the twelve-year-olds play for the Tinley Park Cubs would wrongly assume that the Chicago Cubs had granted permission to or otherwise sponsored their eponymous Little League counterparts. Faced with the prospect of suit, Little League teams everywhere began renaming their teams. (7)
In 2006, back when it was good, NBC's hit show Heroes depicted an indestructible cheerleader sticking her hand down a kitchen garbage disposal and mangling it (the hand quickly regenerated). It was an Insinkerator brand garbage disposal, though you might have had to watch the show in slow motion to notice; the brand name was visible for only a couple of seconds. Emerson Electric, owner of the Insinkerator brand, sued NBC, alleging the depiction of its product in an unsavory light was both an act of trademark dilution and was likely to cause consumers to believe Emerson had permitted the use. NBC denied any wrongdoing, but it obscured the Insinkerator name when it released the DVD and Web versions of the episode. (8) And not just television shows but also movies have provoked the ire of trademark owners: Caterpillar sued the makers of the movie Tarzan on the theory that the use of Caterpillar tractors in the movie to bulldoze the forest would cause consumers to think Caterpillar was actually anti-environment, (9) and the makers of Dickie Roberts: Former Child Star were sued for trademark infringement for suggesting that the star of the absurdist comedy was injured in a Slip 'N Slide accident. (l0) Even museums aren't immune: Pez recently sued the Museum of Pez Memorabilia for displaying an eight-foot Pez dispenser produced by the museum's owners. (11) And forget about using kazoos on your duck tours: Ride the Ducks, a tour company in San Francisco that gives out duck-call kazoos to clients on its ducks, sued Bay Quackers, a competing duck tour company that also facilitated quacking by its clients. (12)
Most of these examples involve threats of suit, and they could be dismissed simply as overreaching by a few aggressive trademark owners. But these threats were not isolated incidents, and they shouldn't be quickly ignored. The recipients of all of these threats, like many others who receive similar objections, (13) knew well that they had to take the asserted claims seriously because courts have sometimes been persuaded to shut down very similar uses. In 1998, for instance, New Line Productions was set to release a comedy about a beauty pageant that took place at a farm-related fair in Minnesota. New Line called the movie Dairy Queens but was forced to change the name to Drop Dead Gorgeous after the franchisor of Dairy Queen restaurants obtained a preliminary injunction. (14) The owners of a restaurant called the "Velvet Elvis" were forced to change its name after the estate of Elvis Presley sued for trademark infringement. (15) A humor magazine called Snicker was forced to pull a parody "ad" for a mythical product called "Michelob Oily," not because people thought Michelob was actually selling such a beer (only six percent did (16)), but because a majority of consumers surveyed thought that the magazine needed to receive permission from Anheuser-Busch to run the ad. (17) And Snicker might face more trouble than that; another court enjoined a furniture delivery company from painting its truck to look like a famous candy bar. (18)
The Mutual of Omaha Insurance Company persuaded a court to stop Franklyn Novak from selling T-shirts and other merchandise bearing the phrase "Mutant of Omaha" and depicting a side view of a feather-bonneted, emaciated human head. (19) No one who saw Novak's shirts reasonably could have believed Mutual of Omaha sold the T-shirts, but the court was impressed by evidence that approximately ten percent of all the persons surveyed thought that Mutual of Omaha "[went] along" with Novak's products. (20) The creators of Godzilla successfully prevented the author of a book about Godzilla from titling the book Godzilla, despite clear indications on both the front and back covers that the book was not authorized by the creators. (21)
The Heisman Trophy Trust prevented a T-shirt company called Smack Apparel from selling T-shirts that used variations of the word HEISMAN, such as "HE.IS.the.MAN," to promote particular players for the Heisman Trophy. (22) This was not Smack Apparel's first trademark lesson: a court previously ordered it to stop selling T-shirts that used university colors and made oblique references to those universities' football teams because the court believed the designs created "a link in the consumer's mind between the T-shirts and the Universities" and demonstrated that Smack Apparel "inten[ded] to directly profit [from that link]." (23) Respect Sportswear was denied registration of "RATED R SPORTSWEAR" for men's and women's clothing on the ground that consumers would be confused into thinking the Motion Picture Association of America sponsored the clothes. (24) A street musician who plays guitar in New York while (nearly) naked was permitted to pursue his claim against Mars on the theory consumers would assume he sponsored M&Ms candies, since Mars advertised M&Ms with a (naked) blue M&M playing a guitar. (25) A legitimate reseller of dietary supplements lost its motion for summary judgment in a suit by the supplements' brand owner because the court concluded the reseller might have confused consumers into thinking it was affiliated with the brand owner when it purchased ad space on Google and truthfully advertised the availability of the supplements. (26) Amoco persuaded a court that consumers might believe it sponsored Rainbow Snow's sno-cones, mostly because Rainbow Snow's shops were located in the same area as some of Amoco's Rainbo gas stations. (27) The National Football League successfully sued the state of Delaware for running a lottery based on point spreads in NFL games, even though the Lottery never used the NFL name or any of its marks for the purpose of identifying or advertising its games. (28) The court was persuaded that the betting cards' references to NFL football games by the names of the cities whose teams were playing might cause consumers to believe the NFL sponsored the lottery game. (29) And the owners of a Texas golf course that replicated famous golf holes from around the world were forced to change their course because one of the holes was, in the view of the Fifth Circuit, too similar to the corresponding South Carolina golf hole it mimicked. (30)
Whatever fraction of the total universe of trademark cases these cases constitute, there are enough of them that recipients of cease and desist letters from mark owners have to take the objections seriously. Indeed many simply cave in and change their practices rather than face the uncertainty of a lawsuit. The producers of the TV show Felicity changed the name of the university attended by characters on the show after New York University, the school originally referenced, objected to the depiction of those students as sexually active. (31) The producers of a movie originally titled Stealing Stanford changed the title of their movie after Stanford University objected to the movie's storyline, which centered on a student who stole money to pay tuition. (32) It's possible that the producers of the show and the movie would have had legitimate defenses had they decided to use the real universities' names despite the objections, but in light of the case law outlined above, neither was willing to defend its right to refer to real places in their fictional storylines. (33) And anecdotes like these are becoming depressingly common. Production of the film Moneyball, which was based on Michael Lewis's best-selling profile of Oakland Athletics General Manager Billy Beane, was halted just days before shooting was set to begin in part because Major League Baseball disapproved of the script's depiction of baseball and therefore objected to use of its trademarks in the film. (34) Apparently Major League Baseball believes it can control the content of any film that refers to real baseball teams.
What unifies all the cases that have given these creators such pause is that courts found actionable confusion notwithstanding the fact that consumers couldn't possibly have been confused about the actual source of the defendants' products. No one could think a (probably) dead Elvis Presley was running a kitsch bar in Texas, Dairy Queen had produced a beauty pageant comedy film, the Chicago Cubs were playing Little League baseball, (35) or the coastal golf course they played in South Carolina had moved to the Texas plains. And while it is possible that some consumers would think a soccer team was selling pants with a beer logo on them, the makers of Godzilla had written a book about him, Michelob was making a new beer called "Oily," the South Carolina golf course had opened a new branch in Texas in cooperation with the seventeen other golf courses the Texas course also mimicked, or Louisiana State University was selling T-shirts that read "Beat Oklahoma" and "Bring it Back to the Bayou!" (but which lacked the LSU name or logo), the plaintiffs in those cases either couldn't or didn't try to prove any such source confusion. And under modern law, they didn't have to.
The actionable confusion, according to these courts, was not confusion that would have led consumers to buy the wrong product, or even to wrongly think they were buying from the trademark owner. Rather, the theory in all of these cases was that consumers would think there was some relationship between the trademark owner and the defendant based on the defendant's use of the trademark. The problem with this formulation is that it fails to specify the types of relationships about which confusion is relevant or the harm that supposedly flows from confusion about those relationships. It is therefore impossible to establish meaningful limits on what sorts of confusion are actionable. In Part III, we address the possible harms that might flow from confusion as to sponsorship or affiliation, as well as the harm that results from expanding trademark law to cover such confusion.
First, though, we pause to consider how trademark law came to deem confusion actionable even when it is entirely unrelated to the source of the products. We do this because understanding why courts felt the need to expand the scope of trademark law gives a context against which to evaluate modern standards. It also helps us to see more clearly where the doctrine went off track.
II. WELL, HOW DID WE GET HERE? (36)
Trademark law traditionally aimed to prevent competitors from diverting consumers who, had they not been deceived, would have purchased from the trademark owner. (37) Because they could only be confident that the confused consumers otherwise would have gone to the mark owner when the defendant offered the same goods or services as the mark owner, courts in this era focused on uses of a trademark by direct competitors. And courts put heavy emphasis on the direct part of this formulation. In one prototypical case, Borden Ice Cream Co. v. Borden's Condensed Milk Co., (38) the court rejected the plaintiff's claim that use of the BORDEN mark for ice cream infringed its rights in BORDEN for milk and related products. (39) The court acknowledged that simultaneous use of BORDEN for milk and ice cream might confuse consumers, but it nevertheless denied the claim because the plaintiff could not show the defendant's use of the same mark for noncompetitive products would divert consumers who otherwise would have bought from the plaintiff. People who want milk don't buy ice cream by mistake.
Cases like Borden seem anachronistic by modern lights, but the results in these cases were a function of the limited purposes of early American trademark law and the economic climate in which courts developed traditional doctrine. Specifically, because producers in the nineteenth and early twentieth centuries sold relatively few products in limited geographic markets, (40) focusing on trade diversion naturally led courts to confine rights to uses by parties in close competitive and geographic proximity. Doctrinally, courts maintained these limits by finding infringement only when the defendant caused confusion about the "source of origin" of its products, and they interpreted "source of origin" quite literally. Since consumers were accustomed to encountering only a limited range of products from any particular producer, they were unlikely to believe that unrelated goods--even ones bearing the same or a similar mark--came from the same source. Thus, a liability standard that required evidence of confusion as to source of origin was essentially equivalent to asking whether confusion would result in trade diversion.
But this tight fit between the requirement of source confusion and the focus on trade diversion depended critically on the assumption that consumers would not think unrelated goods came from the same source. That assumption became increasingly problematic in the early- to mid-twentieth century as producers began serving much wider geographic and product markets. (41) Consumers were more frequently exposed to producers selling a variety of goods (or at least a wider variety), just as they were beginning to understand that companies didn't always themselves produce the products that bore their marks.
These new market dynamics put significant pressure on the traditional doctrinal structure because confusion about source of origin was no longer a perfect proxy for trade diversion if consumers believed producers made a variety of products. But the tension also presented an opportunity for trademark owners who wanted courts to protect their marks against noncompetitive goods and services so they could expand into new markets. And by about the 1920s, mark owners started having success convincing courts that trademark rights should be expanded to account for changed circumstances. In Aunt Jemima Mills Co. v. Rigney & Co., (42) for example, Aunt Jemima, which had used the AUNT JEMIMA mark for syrup, alleged that Rigney's use of the AUNT JEMIMA mark for flour infringed its rights in the mark. Aunt Jemima would have lost this case if the court had applied traditional trademark principles since Rigney was not diverting customers who were trying to purchase syrup. Indeed, the court acknowledged that "no one wanting syrup could possibly be made to take flour." (43) Nevertheless, the court found infringement on the ground the products were "so related as to fall within the mischief which equity should prevent." (44)
Similarly in Yale Electric Corp. v. Robertson, (45) the court refused to allow registration of YALE for flashlights and batteries in light of the plaintiff's prior use of the YALE mark for locks. The court acknowledged that the decision "[did] some violence to the language" of the Trademark Act of 1905, which defined infringing uses of a mark as uses on goods "of the same descriptive properties." (46) But the court claimed "it ha[d] come to be recognized that, unless the borrower's use is so foreign to the owner's as to insure against any identification of the two, it is unlawful." (47)
Both Aunt Jemima and Yale Electric Corp. involved defendants using marks identical to the plaintiffs' and for fairly closely related goods. Thus, consumers in those cases might reasonably have believed that the plaintiffs were the actual sources of origin of the defendants' goods, even if the plaintiffs did not in fact sell those goods. Consumers might, for example, have actually believed that Aunt Jemima sold both flour and syrup, given the complementary nature of those products. Consumers who believed that, of course, still would not have been deceived into buying flour when they intended to buy syrup. But as a purely doctrinal matter, a court could plausibly conclude that the junior user of the AUNT JEMIMA mark confused at least some consumers about the "source of origin" of the defendant's products.
This approach to expanding the scope of trademark rights by interpreting "source" confusion more broadly was pragmatic, particularly as compared to the radical reconceptualization Frank Schechter had proposed. (48) By continuing to focus on confusion as to source rather than adopting an entirely new conceptual framework, courts could act as though they were breaking no new ground, even as they were finding infringement when there was no risk of trade diversion. (49) But while effective in capturing cases like Aunt Jemima, in which consumers might have believed that the mark owner was the actual source of the defendant's noncompeting goods, the doctrinal formulations courts adopted were divorced from trademark law's traditionally limited purposes and had no obvious limiting principles.
We doubt that was an accident. Courts had good reasons to prefer an ill-defined standard in the face of rapidly changing economic circumstances, and an open standard was certainly attractive to mark owners, who wanted courts to protect their ability to enter new geographic and product markets. While it was true that consumers were becoming more accustomed to producers offering a variety of products, many consumers at this time still would not necessarily have assumed that different products bearing the same mark came from the same source, particularly when the products were very different. (50) Nor was it clear that consumers would have fully understood the varied production arrangements mark owners were increasingly employing. Thus, a strict insistence that the plaintiff demonstrate confusion regarding the actual source of the defendant's goods may not have allowed for much expansion beyond cases like Borden. It might also have stood in the way of potentially efficient outsourcing. (51)
To capture a broader range of conduct, courts began to find confusion actionable when it caused consumers to think either (1) that the plaintiff actually produced the defendant's goods or (2) that the plaintiff somehow sponsored the defendant's goods or was affiliated with their producer. (52) In Vogue Co. v. Thompson-Hudson Co., (53) for example, the court held that the defendant's use of "The Vogue Hat Company" to sell hats infringed Vogue's rights in the VOGUE mark for magazines because "[the] course of conduct by the defendant manufacturer and its retailers created a very common alternative impression--first, that these hats were manufactured by the plaintiff; or, second, that, although some knew that plaintiff was not manufacturing, yet these hats were in some way vouched for or sponsored or approved by the plaintiff." (54)
Courts sometimes acknowledged that, by recognizing confusion regarding sponsorship or affiliation as actionable, they were broadening the scope of unfair competition law. (55) Yet most appear to have regarded that expansion as unremarkable; the cases reflect no significant reservations about expanding trademark law to cover noncompetitive goods. (56) This was, however, unmistakably a significant change, and it was this change that set in motion the current crisis.
III. MATERIAL AND IMMATERIAL CONFUSION
As we have seen, the move to prevent confusion as to sponsorship and affiliation began with cases involving related goods that consumers might reasonably have assumed the trademark owner actually made. It expanded to include products that were not made by the trademark owner directly but that consumers might reasonably have believed the trademark owner stood behind or guaranteed, and ultimately it extended to include cases in which there was at most a business relationship between the trademark owner and the product maker but no belief whatsoever of a relationship between the trademark owner and the defendant's goods.
We think the concept of sponsorship or affiliation, introduced to accommodate these broader claims, is to blame for much of what ails modern trademark law. This is not to say that all of trademark law's expansion beyond competing products was unjustified. To the contrary, we think, as we explain further below, that trademark rights ought to extend far enough to cover uses that confuse consumers about who is ultimately responsible for the quality of the defendant's goods or services. Some of those cases will be situations where consumers may not believe the plaintiff actually produced the goods or services at issue but nevertheless believe the plaintiff has played a role in guaranteeing quality. But "sponsorship" or "affiliation" could refer to virtually any relationship between the parties, (57) and we believe it is precisely the vagueness of these terms that has led to the problems we described in the Introduction. Confusion about some relationships simply shouldn't matter because it doesn't affect consumers' decisions to purchase the defendant's goods or services. Yet the "sponsorship or affiliation" formulation allows for no such distinctions, threatening ultimately to swallow up all uses of another's mark.
We therefore propose to do away with the "sponsorship or affiliation" terminology altogether and to reframe the trademark infringement question in terms of whether the defendant's use is likely to confuse consumers about who is responsible for the quality of the defendant's goods or services. Uses that cause such confusion should be deemed trademark infringement; those that cause confusion regarding other types of relationships should be dealt with, if at all, through something analogous to a false advertising claim.
A. Justifications for Expanding Confusion
At the core of trademark infringement are, and always have been, cases in which consumers are confused about who actually produced and/or is offering the defendant's product or service--confused, that is, about the actual source of the defendant's products or services. These are most frequently cases in which the plaintiff and defendant sell the same type of goods, but this category now might also include cases in which the defendant sells goods closely related to those of the plaintiff--cases like Borden. Somewhat further afield, but in our view still justifiably included within trademark infringement, are cases in which consumers are likely to think not that the plaintiff actually supplied the defendant's products or services, but that the plaintiff somehow guarantees their quality. The clearest example involves franchising. Franchisors often do not themselves make the products sold under their brands, and we suspect that a significant portion of the public understands as much. But consumers nevertheless understand the franchisors to stand behind the products sold by its franchisees, and we think it is reasonable for them to do so. It is reasonable for a consumer to assume that the Baskin-Robbins ice cream cone they have today in Denver will be similar to the one they had last week in Seattle, and that if it isn't, it is the national company, not the Denver producer, who is ultimately to blame. (58)
In both of these cases--those in which consumers actually believe that the plaintiff produced the defendant's goods and those in which consumers understand that the plaintiff did not produce the goods but nevertheless believe the plaintiff assures their quality--consumers derive important information from the defendant's use of the mark, and failure to regulate such uses would have serious negative consequences in the commercial marketplace. If Borden sells ice cream, but not other milk products, the use by a defendant of the Borden mark on, say, condensed milk won't actually cause Borden's to lose a sale; they don't sell condensed milk. (59) But it is quite plausible that consumers will assume the Borden's that makes condensed milk is the same Borden's that makes ice cream.
Two arguments conventionally have been used to support the view that trademark law should prevent parties from creating such misimpressions. First is a quality feedback argument: Borden's (the condensed milk company) will be harmed if consumers attribute the ice cream to it and the ice cream turns out not to meet the consumers' quality expectations. (60) Second is the claim that consumers will be harmed because their belief that the same company stands behind both products might have induced some to purchase the ice cream expecting something more than they received.
We think the evidence supporting the first argument is pretty underwhelming. (61) Producers simply aren't likely to be harmed directly by noncompetitive uses except in unusual circumstances. Consumers, on the other hand, do have strong interests at stake in at least some noncompeting-goods cases.
1. Consumers, producers, and the quality of unrelated goods
Research regarding brand extensions suggests, somewhat surprisingly (to us, at least), that producers aren't often harmed by consumers' mistaken association of unrelated products with them. (62) Specifically, the research suggests that consumers generally do not alter their global evaluations of brands (i.e., their assessments of the brand's quality) when they encounter negative information about related products offered under the same mark. (63)
They may think badly of the related products, but that negative view generally doesn't alter the positive view they had of the core product. (64) Indeed the only studies that show any feedback effects involve products that are extremely closely related--such as toothbrushes and toothpaste--and which are explicitly tied together in the market. (65) Consumers, in other words, are smart enough to distinguish different products and hold different impressions of them. (66)
Importantly, even in those few cases where negative information impacts brand image, it does so only in an abstract sense. Negative information about an extension appears not to impact consumers' assessments of the parent brand in the context of the goods the parent previously offered. Thus, for example, Joseph Chang found that both of two unfavorable extensions (Sprite orangeades and Sprite dish-washing detergent) affected consumers' attitudes towards the overall Sprite brand. (67) At the same time, however, neither unfavorable extension diluted the image of the original brand of Sprite lemonades. (68) Thus, the impact of unfavorable extensions is much more hypothetical and contingent than immediate or certain.
Extension information is also unlikely to negatively impact specific brand beliefs, even when the extension is incongruent with those brand beliefs. (69) After reviewing the relevant literature to distill "main tendencies," Henrik Sjodin and Fredrik Torn concluded that negative evaluation of incongruent extension information will not affect evaluation of the parent brand. (70) Moreover, any impact an extension has on specific brand beliefs is likely limited to the parent brand generally; just as with global brand beliefs, an extension has little or no impact on the brand in the context of particular products. Thus, even if an extension affects consumers' general view of the Neutrogena brand as "mild," it is unlikely to affect their belief that Neutrogena hand lotion is mild. Moreover, any such effect on consumers' specific brand beliefs is unlikely to matter because evaluations of specific brand beliefs generally don't impact consumers' decisions: consumers evaluating a new product tend to rely on global attitudes towards a brand rather than attempting to recall and process specific brand attributes. (71)
Combined with the evidence we reported earlier regarding unsuccessful brand extensions, this may mean that companies that are perceived as having failed at brand extension in the past will get less of a brand boost for future extensions, even if they do not suffer a harm in their core market. This is an injury to a trademark owner only to the extent a foregone benefit constitutes a harm, and the case that consumers are injured in that circumstance is more tenuous.
In short, the evidence suggests that producers aren't likely to be harmed in a direct way by the sale of unrelated products bearing their trademark. (72) Certainly producers are not likely to be harmed frequently enough to justify a rule that assumes harm whenever there is confusion in a case involving noncompeting goods. The only plausible case of injury involves likely entry into a market using the same brand.
2. Uses that confuse consumers about quality
Despite the lack of a compelling producer interest in trademark protection against noncompeting goods, we think consumers have a strong interest in protection when the defendant's use of a mark suggests the plaintiff controls the quality of the defendant's products or services. The ability to rely on statements of quality is critical to consumers' ability to evaluate products or services. In cases of this sort, consumers are getting real information that affects their decisions about which products or services to buy. In the franchising context, for example, even if consumers understand that McDonald's Corporation is not the actual source of their hamburgers, trademark enforcement allows consumers to connect the McDonald's brand name to product quality by modulating the reputation of the company ultimately responsible for controlling the quality of those hamburgers. (73) Likewise, consumers are likely to derive important information from use of the APPLE mark in conjunction with the iPhone, information that affects their evaluation of the quality of the iPhone.
Consequently, even if the harm to mark owners in the context of noncompeting goods is uncertain, quality-related messages are important enough to consumers' ability to evaluate products or services that uses of a mark that cause confusion about responsibility for quality ought to be actionable as trademark infringement. (74) Moreover, we think confusion about responsibility for quality affects buying decisions frequently enough that such confusion ought to be actionable without any need for evidence of impact on consumers' decision making. In other words, confusion regarding responsibility for quality ought to be presumed material.
Even here, it is worth distinguishing between cases in which consumers really will be confused about who is responsible for quality despite the unrelated character of the goods and cases in which differences between the goods or the brand may dispel any such confusion. And it bears emphasizing that confusion is to be judged not based on abstract notions of similarity but in the context in which consumers actually see the goods or services. For example, producers often distinguish their goods with a house mark, a product-specific brand, a logo, a slogan, product packaging, and perhaps product color or configuration all at once. (75) Confusion is less likely in the case of unrelated goods when a defendant copies only one (or a few) of these elements rather than all of them. (76) And brand owners are even less likely to be harmed by any confusion that does arise when there are other branding elements by which consumers can differentiate the parties' products or services: several studies have concluded that differentiating an extension product from the parent brand by adding to or altering the stimulus is effective in preventing any feedback effects on the parent brand. So, for example, use of a sub-brand name such as "Ultra by BMW" was sufficient to protect the BMW parent brand from any reputational harm that might otherwise have arisen from consumers' negative perceptions of a lower-priced model offered under that name. (77)
It is also worth emphasizing that consumer, rather than producer, interests are driving protection in cases of confusion regarding responsibility for quality. Producers serve as an imperfect proxy for consumers here because they are motivated to sue in ways that consumers aren't. (78) But the fact that it is consumers we are primarily protecting in these cases means that trademark law should eschew rules that undermine consumer quality expectations. For example, producers that are nominally protecting consumer interests in quality should not be able to undermine those interests by licensing their mark without quality control. (79) Indeed, whatever the justifications for the general rules against assignments of trademarks "in gross" and naked licensing, (80) those rules seem to have particular force here. We think, in fact, it is a significant virtue of focusing trademark infringement on responsibility for quality that it would harmonize the infringement standard with the requirement that licensors exercise quality control in order to claim the benefits of licensed uses. (81)
3. Consumer perceptions in pure sponsorship cases
In contrast to the types of uses we put in the trademark infringement category, consumers get no quality-related information from the defendant's use of a mark in cases that involve confusion about other types of relationships that might exist between the plaintiff and defendant. Consumers, for example, might believe the presence of Coca-Cola cups in front of the American Idol hosts suggests some kind of product placement agreement between Coca-Cola and the producers of American Idol, but no reasonable person thinks Coke controls the quality of the American Idol television show. (82)
This is not to deny that consumers obtain some kind of information from these uses regarding the brands' personalities. Consumers may, for example, learn about the image Coca-Cola is trying to project through its association with a popular show (and probably about the target audience of American Idol through its association with Coke). The fact that consumers can use brand relationships to build such image connections might be thought to benefit them, though even that is open to debate. But this image-related information is qualitatively different than the information consumers derive from quality-related messages, and we believe the benefits from protecting such information are lower and the costs of protecting it are higher. (83)
On the benefits side, it seems clear to us that the producer-harm arguments for claims in the context of these other relationships are particularly weak. The marketing literature dealing with brand alliances is instructive here. Brand alliances, as defined in this literature, are "partnership[s] between two entities in which efforts are combined for a common interest or to achieve a particular aim." (84) These partnerships can take many forms, but the two most common forms are joint promotions (McDonald's using Kung Fu Panda toys in its Happy Meals) and co-branding arrangements (Edy's[R] Loaded Cookie Dough Ice Cream with Nestle Toll House[R] cookie dough). The lesson of this literature is clear, if somewhat counter-intuitive: "consumers do not routinely blame a host brand for its partner's mistakes." (85)
In one significant study, the authors focused on the consequences to a clothing company of negative information about its supplier and a celebrity endorser of the clothing company's products. (86) More specifically, the authors attempted to measure the change in consumers' attitudes toward the fictitious clothing company (the host brand) from information that the company's partners had behaved immorally or had been incompetent. (87) There wasn't any. Putting their finding in context with other research on brand associations, the authors concluded that negative information does not have any feedback effect on the partner absent some additional information about the partner's culpability for the failing, regardless of whether the information relates to competence or moral failings and regardless of whether the information is about another company or a person with which the partner is associated. As the authors noted,
[A] host brand may generally be quite impervious to negative publicity surrounding its partner brand; the host brand was only affected [in the study] when participants were led to believe that the host knew of and condoned the partner's behavior. Spillover from the partner brand to the host brand did not occur unless this condition was present. (88)
Recall that this finding comes from a study in which respondents were told explicitly that the host brand had relationships with the partners about which the negative information was provided. Thus, there was no ambiguity about affiliation--respondents understood that the host brand was affiliated with the partners. What this suggests is that there is unlikely to be any negative feedback absent some information--additional information, beyond the mere fact of association--demonstrating the host brand's specific culpability. In other words, consumers generally do not impute responsibility based solely on the fact of association. The important implication here is that consumers generally do not view alliance relationships themselves as indicative of a partner's control.
These studies suggest that any harm to producers from confusion about sponsorship or affiliation is quite attenuated: producers suffer no lost sales, and they are unlikely to suffer any reputational consequences absent additional information suggesting control over the partner. If a mall cookie vendor advertises that its cookies contain M&Ms, for example, consumers might or might not assume that Mars had entered into a deal with the cookie company, but whether they do or not they are unlikely to blame Mars if they don't like the cookies. The only sense, then, in which a mark owner is harmed by third-party uses that suggest sponsorship or affiliation is that third-party uses might interfere with the mark owner's own ability to develop and derive value from such relationships. In other words, the only likely loss to trademark owners from affiliation confusion is the loss of revenue the trademark owner could have made by licensing the mark to the putative affiliate. This is a claim to market control, not a claim of harm resulting from confusion or even an injury to consumers at all. (89) We think this circular claim to licensing revenue is insufficient to justify trademark protection, particularly in light of the significant costs such protection entails. (90)
The marketing literature has implications for the merchandising right as well. Merchandising cases involve the use of brands not to identify the source or quality of goods, but instead as desirable products in and of themselves. The sale of brands qua brands on T-shirts, hats, and the like presents difficult problems for trademark theory. (91) There is no obvious source relationship between, say, a university or a professional sports team and T-shirts or hats that feature the logo of that university or team. There might be a presumed franchising-type quality relationship; universities and sports teams today do license the manufacturing of clothing featuring their logos, and it is possible that consumers both assume that the mark owner is serving as a guarantor of the quality of those clothes and that any clothing featuring the school or team name is in fact licensed by the university. On the other hand, those consumer assumptions will not be present in every case. Outside the school and sports context, and perhaps a few others (Nike, say), it is not clear to us that consumers assume that any T-shirt with a trademark on it is necessarily licensed, much less quality-guaranteed, by the brand owner.
Our point is not that consumers can never be harmed by confusion regarding non-quality-related relationships. Rather, the point is that the sort of attenuated confusion at issue in sponsorship and affiliation cases does not necessarily or even often harm consumers or the market for quality products. The benefits of expanding confusion law to this class of cases are correspondingly smaller.
B. Costs of Expanding Confusion
If the benefits of treating sponsorship or affiliation confusion as infringing are low, the costs of that protection are high. We see at least four potential types of costs to extending trademark rights to cover any perceived relationship unrelated to the quality guarantee.
First, the expansion of trademark infringement to include any claim of affiliation or relationship necessarily expands the rights of all mark owners beyond the goods and geographic regions in which they sell or into which they are likely to expand. As a result, it produces any number of conflicts between legitimate mark owners that have coexisted for years under traditional trademark rules but who cannot share a mark without some risk that someone will think the identity of the marks implies that they are somehow related or affiliated. Dell the bookseller predates Dell the computer company, (92) but if the lack of relationship between the products is no longer to matter--as it increasingly doesn't under the broad form of sponsorship or affiliation confusion--then whether those two marks should be allowed to coexist will depend on what a court concludes about the percentage of people who think they are related. The same is true of United the airline company and United the moving company, of Apple the computer and electronics company and Apple the travel agency, and thousands of others. (93) Whether or not the trademark owners will sue in these cases (and it is far from clear which would sue the other), the fact is that consumers are likely just as "confused" in these cases as in the ones in which they do sue. Taking mark similarity in different markets to imply sponsorship or affiliation, as many courts have done, means that only one company should be allowed to use any given mark (or indeed any term too close to that mark). Either we should be awash in lawsuits, or we are not really taking seriously the idea that confusion as to sponsorship or affiliation matters.
Second, and related, is the problem of coddling consumers. Consumers are pretty good in most circumstances at figuring out what they want to buy. (94) But their perceptions are shaped by the environment in which they find themselves. If they see trademarks that overlap, they will adapt and deal with that environment. And as long as legal claims remain available against explicitly misleading statements--e.g., "Coca-Cola is the official sponsor of American Idol"--there is every reason to believe that consumers will be able to do so relatively easily. (95) But if they become used to a world in which only one company has a right to refer to a brand for any reason, consumers may well become confused by uses that would otherwise seem perfectly reasonable. (96) Expanding trademark law to prevent remote prospects of confusion will change consumer expectations in ways that may make confusion on the basis of remote connections more likely, which might make still further expansion of trademark law necessary to stop critics, parodies, or gripe site Web pages from funding themselves with online ads or selling T-shirts. (97) Put another way, unless we are able to identify more specifically the types of relationships that could give rise to actionable confusion, there is no logical stopping point for trademark protection.
To see this point, consider a seemingly extreme example. (98) We think it utterly uncontroversial for a grocery store to locate generic colas on a shelf next to Coca-Cola. But why? One could say, perhaps, that such uses do not confuse consumers into thinking that Coke licenses the placement or sponsors the generic colas. But if the placement does not confuse consumers about Coke's relationship with the generic colas, it is only because the law has long permitted the practice, and so consumers accept and understand it. That is not an inevitable result, however. After all, cola companies do pay for the placement of their products on store shelves--they just pay grocery stores rather than mark owners. Had the courts said at the outset that trademark owners could sue to prevent such placement--reasoning that consumers might think that proximity implied association, as evidenced by the fact that the generic sellers pay for shelf placement--grocery stores might well have had to separate like products to avoid any risk of confusion. Further, even if a finding of confusion seemed unlikely, many companies would have agreed to change their behavior or take a license rather than pay to litigate a case all the way to trial and risk losing. This, in turn, would mean that consumers would not have gotten used to seeing all the colas grouped together and would make it harder for anyone else to arrange their shelves this way because, over time, the placement of generic cola beside Coke would be more surprising to consumers. And if no one else is putting genetic colas next to Coke, it is an easy mental step to conclude that a grocer that does so is free tiding on Coke's interest in being insulated from nearby competitors, particularly if the grocer is making money directly or indirectly from the placement or sales of generic colas. (99)
Arguably, something similar has already happened with T-shirts bearing university and sports logos. A use that originally confused no one came, over time, to confuse consumers as a few courts held that such uses were illegal. (100) These rulings led to widespread licensing, which made consumers assume that such T-shirts came only from licensed vendors. (101) As a result, it may be that today the law must enforce the claims of universities and sports teams to be the only ones to sell merchandise bearing the team logos, though that conclusion is certainly contested. (102) But if it must, it is not because doing so was the only way to prevent confusion. It is the law itself that will have created that confusion.
Third, sponsorship and affiliation cases may be more likely to reach the wrong result than other types of trademark infringement cases. Courts have developed multifactor likelihood of confusion tests to identify the circumstances in which plaintiffs should win trademark cases. (103) But those tests were designed to deal with cases like Borden in which consumers might believe that the plaintiff is responsible for quality, and few of the factors make much sense when the issue is confusion about some unspecified sponsorship or affiliation relationship. (104) Factors such as marketing channels, likelihood of expansion, and even consumer sophistication have little relevance to evaluating Coke's claim to be the exclusive soft drink associated in the minds of consumers with American Idol. The result is that more and more attention is focused on factors--notably the court's assessment of the defendant's intent and survey evidence--that are quite malleable and may tend to shift with the quality of the lawyers or experts arguing rather than the strength of the case. (105) In Balducci, for instance, the court's (unsupportable) conclusion that a parody of a Michelob ad was infringing was driven by spurious survey evidence; in that case, the survey asked a question (whether the defendant should have to get permission to publish the parody) that misstated the law. (106)
Barton Beebe's work has shown that courts tend to fall back on their assessment of a defendant's intent in deciding whether consumer confusion is likely. (107) Perhaps this is acceptable if the intent in question is intent to confuse consumers into buying the defendant's goods instead of the plaintiff's. But when the relevant intent is more amorphous--some sort of free riding--the fact that courts rely so heavily on intent becomes problematic because the concept of free riding is ultimately empty. (108) In the end, it may be that trademark law has worked out a set of rules that effectively distinguish confusing from nonconfusing uses in the run-of-the-mill trademark case. But those rules don't apply to sponsorship cases, and the factors courts substitute may be more prone to produce erroneous findings on confusion because the tools don't work well for determining whether this kind of confusion is likely.
Finally, sponsorship or affiliation confusion claims pose particular risks to free expression. Many of the examples we discussed in Part I involve not the sale of commercial products but the use of a mark as part of protected speech on issues of social relevance. (109) This should be no surprise because, with the importance of brand image in today's economy, trademarks "form an important part of the public dialog on economic and social issues." (110)
No one can talk about Barbie dolls and the role they play in popular culture without using the term "Barbie," and often the dolls themselves. Nor can they effectively make fun of trademark owners without using their marks. (111) And as many commentators have noted, modern expression frequently requires the use of trademarks in their role as social referents, whether or not the product itself is being discussed directly. (112) Satire or commentary on unrelated political issues may need to refer to advertisements to make a clear point in a culture in which advertising is ubiquitous. (113) And iconic brands may be needed for supportive messages as well; adulation may well prompt imitation (which is, after all, the sincerest form of flattery). One could hardly go a day in 2008 without encountering an ad for Obama memorabilia, for instance.
Restricting this speech is harmful to society. It is likely also unconstitutional. (114) And while all aspects of trademark law can pose First Amendment threats, we think it is no accident that the worst problems don't come from traditional suits against the sale of competing products, but rather from claims that try to stretch the concept of confusion to cover unrelated parody products or pure speech by using the rubric of sponsorship or affiliation. (115) Even if those claims don't succeed, the mere fact that the suits are brought may distort the use of trademarks as cultural referents, changing the nature of the social conversation about things, (116) as when trademark owners demand that brands be removed from pictures or movies depicting places in which they actually appear. (117) As Judge Kozinski has noted, "[m]uch useful social and commercial discourse w[ill] be all but impossible if speakers [are] under threat of an infringement lawsuit every time they ma[k]e reference to a person, company or product by using its trademark." (118)
C. Weighing Costs and Benefits
None of the costs we have identified is sufficient, alone or together, to defeat the rationale for trademark protection generally, or even with respect to uses that suggest responsibility for quality. Trademark law as a whole has benefits that far outweigh its costs. But it is significant--and not conincidental--that the extreme form of the sponsorship theory of confusion produces the fewest benefits and causes the most harm. Trademark law works best when it stays close to its traditional core. As trademark doctrine expands further a field from that core, it imposes more costs on society and accordingly requires more justification. (119) But as we have seen, many of those expansions have less benefit, not more.
What we need, then, is balance. Trademark law must extend beyond pure source confusion in some cases, or it can't effectively serve its purposes. But if it extends too far, it does more harm than good. In the Parts that follow, we apply these lessons to trademark doctrine.
IV. DISTINGUISHING TRADEMARK INFRINGEMENT FROM FALSE ADVERTISING
We believe that the best way to achieve balance in cases that do not involve confusion about responsibility for quality is not to try to fit them within the traditional trademark framework at all, but instead to think of them as akin to cases of false advertising. False advertising law coexists in the Lanham Act with trademark law, and both are directed at misrepresentations in the marketplace. But while trademark law has traditionally aimed at protecting against the use of the plaintiff's mark to misidentify the source of the defendant's goods, false advertising law targets a broader range of false or misleading statements about either the plaintiff, the defendant, or the plaintiff's or the defendant's goods or services. (120)
False advertising law's broader scope, however, is counterbalanced by limitations that do not apply in trademark infringement cases. (121) Most notably for our purposes, while trademark law presumes actionable harm from proof of consumer confusion, entitling plaintiffs to an injunction, (122) false advertising is actionable only if the representations made by the defendant materially affect consumer purchasing decisions. (123) In other words, false or misleading statements may or may not cause harm, depending on how people perceive those statements. Imagine, for example, that the defendant falsely states that the plaintiff has 11,500 employees, when in fact they have 11,600. The claim is one of fact, and it is provably false, but it is hard to imagine that the difference would matter to consumers at all. False advertising law accordingly treats it as not material. (124)
We think that logically trademark law can be conceived as a specialized subset of false advertising law. False advertising law covers a broad range of misrepresentations, not all of which are actionable. Trademark law focuses on a subset of these misrepresentations--those that involve use of the plaintiff's trademark or a simulacrum thereof to brand the defendant's goods. Because of the centrality of those representations, courts in trademark cases have not required proof that confusion is material. Rather, they have presumed materiality.
That presumption makes sense in the context of traditional trademark infringement: if consumers are confused into thinking the defendant's goods are the plaintiff's, it is logical to think such confusion will materially affect consumer purchasing decisions, at least when the confusion arises and is not dispelled before purchase. (125) We think the same can be said of uses that cause confusion about responsibility for quality. But the expansion of trademark law to cover confusion about other types of relationships stretches the general presumption too far. The types of confusion alleged in the cases we discussed in Part I may or may not affect consumer purchasing decisions; indeed, for the reasons we explained in Part II, in most cases we think it unlikely that it will.
The solution, in our view, is simple: not to categorically rule out cases involving those other forms of confusion, but to limit those claims so as to increase the benefits of those still actionable and decrease their costs. The easiest way to do so would be to import into trademark law the materiality requirement courts have created in the false advertising context and apply it in any case based on confusion that does not relate to source or control over quality. (126) Plaintiffs bringing cases then would face a choice: (1) bring a trademark infringement claim and be required to prove confusion regarding actual source or responsibility for the quality of the defendant's goods or services; or (2) bring a false advertising-type claim alleging that the use causes some other form of confusion and be required to prove confusion about that relationship and that such confusion materially affects consumers' decisions whether to purchase the defendant's goods or services. (127)
This formulation has an obvious analogue in antitrust law and its distinction between conduct deemed per se anticompetitive and conduct judged according to the rule of reason. Courts in antitrust cases have identified certain conduct, including conspiring to fix prices, rig bids, or divide markets, as the type of conduct experience has shown harms competition in most cases. (128) They therefore impose liability whenever a plaintiff shows that the defendant intentionally engaged in conduct of this type, and the plaintiff does not have to prove that the defendant's conduct harmed competition. Instead, harm is conclusively presumed. (129) Conduct with ambiguous effects on competition, on the other hand, is evaluated by a court more carefully, and plaintiffs alleging such conduct are required to demonstrate that the conduct at issue in fact harms competition. (130) What we propose for trademark law is much the same. True source confusion and confusion regarding responsibility for quality should be actionable and should be presumed to materially affect consumers' purchasing decisions. Other forms of confusion have more ambiguous effects on consumers, however, and those forms of confusion should only be actionable when they can be proven material to consumers' decisions in particular cases. (131)
Requiring materiality may also enable courts to tailor relief even when they find some remediable confusion. Because materiality is a sliding scale, not an all-or-nothing inquiry, courts could plausibly find some conduct to be material to purchasing decisions of only a few customers. In false advertising cases, the strength of the materiality finding is related to the remedy; the more problematic the deception, the more willing the courts are to act. (132) This makes sense as a matter of cost-benefit analysis. Thinking about sponsorship cases in these terms may permit courts to do the same sort of balancing of remedies here, for example requiring disclaimers as the cure for certain minor types of trademark harm. (133)
One objection to our proposal is that information about sponsorship or affiliation may be material in some respects to consumer perceptions or to the trademark owner's reputation without actually having any measurable impact on consumer purchasing decisions. Trademark owners enter into sponsorship relationships because they think the goodwill those relationships create will ultimately translate into purchases made by happy consumers. For example, 711 sponsors the Chicago White Sox, which has agreed to start all its home games at 7:11 pm. (134) Both parties hope the relationship signals something to consumers about the brands' personalities and that it impacts consumers' affective responses to those brands. And while both expect to benefit from the relationship, it can, at least in theory, work both ways: the linkage between the two may mean that when the White Sox have a bad year, or if one of their players takes steroids, 7-11's image suffers a bit too. (135) We recognize this possibility, but we don't think it should deter courts from requiring materiality. If information relating to brand personality is really important to consumers, it should be reflected in the final analysis in purchasing decisions. If the information does not influence purchasing decisions, any harm a trademark owner might suffer is de minimus, and is likely to be outweighed by the harms of expanding protection we detailed above. To put it plainly, attempting to protect these other forms of information through trademark infringement claims will ultimately eviscerate any meaningful limits on the scope of trademark rights. We should be extremely confident of real harm in these cases before we go down that road.
V. WELL, HOW DO WE WORK THIS? (136)
Treating some of the cases currently brought as sponsorship and affiliation cases as analogous to false advertising claims rather than trademark infringement claims would not necessarily require a change to the Lanham Act. Trademark law did not originally extend to sponsorship claims involving unrelated goods. It was courts, not Congress, that expanded the scope of trademark law, and courts presumably could undo that expansion.
While the seemingly simple approach would be to treat sponsorship and affiliation claims as false advertising claims, we can't actually push these claims into the false advertising system wholesale. The problem is that false advertising claims require proof that the plaintiff and the defendant are competitors. (137) Unless it was interpreted extremely broadly, such a requirement would eliminate sponsorship and affiliation claims, which by definition involve unrelated goods. Nevertheless, we think the principle of false advertising--that companies should not deceive consumers in ways that materially affect their purchasing decisions--applies with full force to implicit or explicit assertions about sponsorship or affiliation that do not imply responsibility for quality. Those assertions should be actionable if, but only if, they (1) are false or misleading and (2) materially affect consumer decisions.
The challenge, then, is to interpret the trademark infringement provisions of the Lanham Act to distinguish between cases involving confusion regarding responsibility for quality on the one hand, and cases involving confusion regarding other types of relationships on the other. Just as the Supreme Court has read distinctions into the statute in the context of trade dress (138) and reasonable distinctions into other statutes, (139) we think courts could distinguish between different types of alleged confusion and require evidence of materiality where the alleged confusion relates not to responsibility for quality, but to some other form of relationship. This particular distinction, we acknowledge, would be easier to reconcile with the text of the statute in the case of section 32 claims for registered trademarks since that section lacks specificity about the nature of the relevant confusion. (140) Because section 43(a) specifically refers to confusion regarding "the origin, sponsorship, or approval" of the defendant's goods, (141) courts would have to distinguish between different cases of confusion regarding sponsorship or approval, requiring evidence of materiality where the alleged confusion of this sort did not imply responsibility for quality. We do not think this type of distinction is inconsistent with the Lanham Act or that it is a qualitatively different kind of distinction than the distinction between product configuration and product packaging (neither of which is specifically referenced in the statute). (142) Indeed, the Supreme Court was willing to draw a similar distinction between authorship and source in Dastar Corp. v. Twentieth Century Fox Film Corp., (143) which seemed to presage our point. The Lanham Act, the Court said, "should not be stretched to cover matters that are typically of no consequence to purchasers." (144) And some courts already seem willing to consider materiality, evaluating evidence of actual confusion based on whether that confusion is actually relevant to a consumer purchasing decision. (145) And courts have already had to engage in similar legerdemain in interpreting subsection (B) but not (A) of the Act to require materiality; (146) they could easily require it in some subsection (A) cases as well. Given how entrenched the current understanding of trademark law has become over the last several decades, however, it may actually be easier to achieve this reform in Congress.
Whether courts or Congress are the actors, the change we propose is straightforward: the law should require that trademark owners claiming infringement based on confusion regarding anything other than source or responsibility for quality must demonstrate the materiality of that confusion to consumer purchasing decisions. That is, we should not presume social harm from likely confusion regarding other types of relationships, as we do with confusion as to the source of products. Instead, plaintiffs should have to prove that harm, as they do in false advertising cases. And the plaintiff should have to show that it is confusion about the relationship between the parties that is material to consumers' purchasing decisions; it is not sufficient that the trademark standing alone is material to purchasing decisions. To have a claim against the producers of a television show in which the characters used an APPLE computer, for example, Apple would have to demonstrate both that the presence of the computer is likely to confuse consumers about a relationship between Apple and the show and that consumers' belief that such a relationship exists is likely to impact their decision to watch the show. It would not be enough to show that the use of the APPLE computer captured viewers' attention.
Thus, the structure of trademark claims under our approach will look something like this: traditional trademark claims about source confusion among competitors will be subject to the traditional rules, which presume harm to the plaintiff upon evidence of likely consumer confusion, so long as the alleged confusion arises before purchase and is not dispelled. So too will certain classes of confusion that don't strictly involve competitors: plaintiffs who are thought by consumers to guarantee the quality of the defendant's products (franchisors, for example) and plaintiffs who have not yet entered an adjacent market but are sufficiently likely to do so that consumers regard them as quality guarantors. Owners of famous marks that can show likely harm to the mark in its core market from tarnishment or blurring can assert a dilution claim. Trademark owners who cannot satisfy any of these criteria can still bring a claim for confusion as to sponsorship or affiliation, but only if they can demonstrate that the defendant's false or misleading representations as to sponsorship or affiliation are actually likely to affect consumers' purchasing decisions.
We acknowledge that a finding of liability based on material confusion in a downstream market represents a windfall of sorts to trademark owners, particularly if courts award damages. As the evidence in Part III demonstrated, trademark owners are not normally hurt by uses of this sort. Nonetheless, because consumers are hurt in the downstream market, we think it makes sense to permit the trademark owner to sue in cases where a mistaken belief about the relationship between the plaintiff and defendant is likely to affect the purchasing decisions of a substantial number of consumers. (147) But we emphasize that it is injury to consumers, and not some putative right on the part of the trademark owner to own non-competitive uses of the mark, that justifies a lawsuit in these circumstances. (148)
Cases subject to the materiality requirement, we propose, would include not only the sports team and media cases we discussed in Part I, but also cases involving the so-called "merchandising right." (149) Indeed, the approach we advocate might actually offer a reasonable middle ground in the debate over the merchandising right, giving trademark owners control over the sale of T-shirts, hats, and other memorabilia if, but only if, consumers are actually influenced in their purchasing decisions by the belief that the trademark owner sponsored the goods. (150) We think it is unlikely that the Dutch soccer fans we mentioned in our Introduction actually took the pants from a beer company because they thought there was a sponsorship relationship between Bavaria and the Dutch team, but we concede it's possible. FIFA should have an opportunity to prove that they did, but we certainly shouldn't assume such a connection matters without evidence. And again, to be clear, what FIFA should have to show is that a relationship between the beer company and the Dutch national team would be material to the fans' decisions to wear the pants. It is not enough that the colors were important to the fans.
We think our approach would also ameliorate the problems with celebrity endorsement cases. (151) Even in those cases where use of a celebrity identity creates confusion about a relationship between the celebrity and the user, it's unlikely the perceived relationship entails control over the quality of the user's goods or services. These cases therefore involve confusion regarding the other types of relationships we've tried to distinguish, and that plaintiffs in these cases should have to prove materiality. But we can imagine that some uses of a celebrity's identity--those that really do suggest the celebrity's endorsement of the user's products or services--might be material to consumers' decisions to purchase the goods or services they think the celebrity has endorsed. These cases, then, are a good example of why we don't rule out non-quality-related confusion categorically.
Our approach may even provide a way to think about trademark use in virtual worlds. (152) The sale of virtual goods is a bit of an edge case. There is no real good here for which the trademark owner might stand as a guarantor of quality. So we think in the ordinary case that any confusion arising in the context of a virtual transaction involves an assumption about affiliation or sponsorship between the trademark owner and the virtual world. We can imagine, however, that over time circumstances might arise in which the quality or source of the virtual good is itself at issue. If Gucci virtual purses last longer or have features that other purses don't, selling a non-Gucci purse without those features as a Gucci purse would create the sort of harm trademark law has traditionally sought to prevent. If that happens, it shouldn't matter that the good is virtual rather than real. (153)
To be sure, there is some risk under our approach that plaintiffs will be unable to prove in court harm they did in fact suffer. But we think there are ways to mitigate that risk. For example, courts could draw inferences or create presumptions of materiality for certain classes of explicit statements like "Coke is the official drink of American Idol," just as false advertising law currently treats literal falsity more harshly than literally true but allegedly misleading representations. (154) And the opposite risk--that plaintiffs will prove materiality by dubious evidence--can be reduced by demanding a logical causal chain showing injury to the trademark owner. (155) In any event, the risk of false negatives must be balanced against the costs of sponsorship and affiliation litigation, which as we suggested in Part II are substantial.
Trademark law has expanded dramatically in the last century to the point where it now prohibits conduct by companies that seems unlikely to confuse consumers in any material way. The result is a long series of seemingly absurd decisions. We think the problem is that courts have presumed that if consumers are confused at all, that confusion is problematic. That presumption makes sense when the plaintiff and defendant are competitors and consumers are confused about the source or the quality of the products they are buying. But the same presumption makes no sense in cases of sponsorship or affiliation confusion because the evidence suggests that most such confusion does not in fact affect consumer purchasing decisions.
We suggest that trademark law can best deal with sponsorship or affiliation claims by taking a page from history and returning this subset of cases to its roots in false advertising law. Believe consumers have been injured because a Little League team uses your professional sports team name, because soccer fans are wearing your team colors without your permission, or because a TV show uses your soda in it? Our rule is simple: prove it.
([dagger]) [c] 2010 Mark A. Lemley, Mark McKenna, and the Trustees of Leland Stanford Jr. University.
(1.) FIFA appears to have objected to the pants because they were distributed by, and bore the name of, Dutch brewer Bavaria, while FIFA had entered into a partnership agreement with Anheuser-Busch making Budweiser the official beer of the World Cup. Dutch Fans Watch Match in Their Underwear, ESPN SOCCERNET, June 17, 2006, http://soccernet.espn.go.com/news/story?id=371466&cc=5901.
(2.) See Heather Smith, Goal Tending, IP LAW & BUS., Aug. 2006, at 28, 28; Dutch Fans Watch Match in Their Underwear, supra note 1. FIFA apparently did not object that the underwear too was orange.
(3.) Because of this uncertainty, consumers won't be willing to pay as much for that quality. See, e.g., GEORGE A. AKERLOF, The Market for Lemons: Quality Uncertainty and the Market Mechanism, in AN ECONOMIC THEORIST'S BOOK OF TALES: ESSAYS THAT ENTERTAIN THE CONSEQUENCES OF NEW ASSUMPTIONS IN ECONOMIC THEORY 7 (1984). We won't pay as much for an iPod if we think there is a chance it is a cheap knockoff masquerading as an iPod.
(4.) Indeed, the statute specifies the sorts of misrepresentations that are actionable. 15 U.S.C. [section] 1125(a)(1)(B) (2006) (forbidding misrepresentations as to the "nature, characteristics, qualities, or geographic origin" of goods or services).
(5.) See, e.g., Rescuecom Corp. v. Google Inc., 562 F.3d 123 (2d Cir. 2009); Am. Airlines, Inc. v. Yahoo!, Inc., No. 4:08-CV-626-A, 2009 WL 381995 (N.D. Tex. Jan. 16, 2009) (denying Yahoo!'s motion to transfer venue in trademark infringement case based on Yahoo!'s use of American Airlines as keyword); Am. Airlines, Inc. v. Google, Inc., 4:07-cv-00487 (N.D. Tex. Oct. 24, 2007) (order denying Google's motion to dismiss).
(6.) With apologies to the Talking Heads and to anyone who hates cute heading titles.
(7.) See, e.g., Tim Cronin, MLB to Youth: You're Out, HERALD NEWS, May 27, 2008, available at http://www.suburbanchicagonews.com/heraldnews/news/ 971947,4_1_JO27_LOGOS_S1.article; see also Katie Thomas, In Cape Cod League, It's Tradition vs. Trademark, N.Y. TIMES, Oct. 24, 2008 at B11 (discussing a similar claim brought by MLB against the amateur Cape Cod League); Michael Masnick, Major League Baseball Bullying Amateur Baseball in Trademark Shakedown, TECHDIRT, Mar. 13, 2008, http://www.techdirt.com/articles/20080312/013742509.shtml (same).
(8.) See Paul R. La Monica, NBC Sued over 'Heroes' Scene by Garbage Disposal Maker, CNNMONEY.COM, Oct. 17, 2006, http://money.cnn.com/ 2006/10/17/commentary/mediabiz/index.htm.
(9.) Caterpillar Inc. v. Walt Disney Co., 287 F. Supp. 2d 913,917 (C.D. Ill. 2003).
(10.) Wham-O, Inc. v. Paramount Pictures Corp., 286 F. Supp. 2d 1254, 1255-58 (N.D. Cal. 2003).
(11.) Museum Faces Legal Battle over Giant Pez Dispenser, KTVU.cOM, July 1, 2009, http://www.ktvu.com/print/19911637/detail.html. The museum was originally called the Pez Museum, but the owners changed the name in response to a previous objection from Pez.
(12.) Jesse McKinley, A Quacking Kazoo Sets Off a Squabble, N.Y. TIMES, June 3, 2009, at A16. Ducks are open-air amphibious vehicles that can be driven on streets and operated in the water.
(13.) The Chilling Effects Clearinghouse collects letters from trademark owners that make aggressive assertions of trademark (and other intellectual property) rights. See Chilling Effects Clearinghouse, http://www.chillingeffects.org (last visited Sept. 9, 2009). As of February 25, 2009, the Chilling Effects database contained 378 such letters. Among the many specious objections are an objection from the National Pork Board (owner of the trademark "THE OTHER WHITE MEAT") to the operator of a breastfeeding advocacy site called "The Lactivist" for selling T-shirts with the slogan "The Other White Milk," Pork Board Has a Cow over Slogan Parody, CHILLING EFFECTS CLEARINGHOUSE, Jan. 30, 2007, http://www.chillingeffects.org/trademark/notice.cgi?NoticeID=6418; from Kellogg to the registrant of the domain name "evilpoptarts.com," Kelloggs Poops on Evilpoptarts.com, CHILLING EFFECTS CLEARINGHOUSE, June 5, 2006, http://www.chillingeffects.org/ acpa/notice.cgi?NoticeID=4377; from Nextel to the registrants of the domain name "nextpimp.com," Nextel Says "Don "t Pimp My Mark", CHILLING EFFECTS CLEARINGHOUSE, June 22, 2005, http://www.chillingeffects.org/acpa/notice.cgi?NoticelD=2322; and from the owners of the Marco Beach Ocean Resort to the operators of "urinal.net," a website that collects pictures of urinals in various public places, for depicting urinals at the Resort and identifying them as such, Mark Owner Pissed About Urinals, CHILLING EFFECTS CLEARINGHOUSE, Jan. 4, 2005, http://www.chillingeffects.org/trademark/ notice.cgi?NoticeID=1576.
(14.) Am. Dairy Queen Corp. v. New Line Prods., Inc., 35 F. Supp. 2d 727, 728 (D. Minn. 1998).
(15.) Elvis Presley Enters., Inc. v. Capece, 141 F.3d 188 (5th Cir. 1998).
(16.) Anheuser-Busch, Inc. v. Balducci Publ'ns, 28 F.3d 769, 772-73 (8th Cir. 1994). That any consumers were confused was remarkable, and perhaps a statement about the reliability of consumer confusion surveys rather than the stupidity of 6% of the population.
(18.) Hershey Co. v. Art Van Furniture, Inc., No. 08-14463, 2008 WL 4724756 (E.D. Mich. Oct. 24, 2008). Hershey has also sued Reese's Nursery. Complaint at 1, Hershey Chocolate & Confectionery Corp. v. Reese's Nursery and Landscaping, No. 3:09-CV-00017 JPB (N.D.W. Va. Mar. 19, 2009).
(19.) Mutual of Omaha Ins. Co. v. Novak, 836 F.Zd 397, 397 (8th Cir. 1987).
(20.) Id. at 400.
(21.) See Toho Co. v. William Morrow & Co., 33 F. Supp. 2d 1206, 1206, 1212 (C.D. Cal. 1998).
(22.) Heisman Trophy Trust v. Smack Apparel Co., No. 08 Civ. 9153(VM), 2009 WL 2170352, at *5 (S.D.N.Y. July 17, 2009). Smack Apparel produced several such T-shirts, including one that substituted the number 15 for "IS" in the word HEISMAN and was printed in the colors of the University of Florida, clearly to promote Florida quarterback Tim Tebow's candidacy. See Smack Apparel Lawsuit, LSU TIGER TALLER NEWSLETTER (LSU Trademark Licensing, Baton Rouge, La.), Jan. 30, 2009, at 6.
(23.) Bd. of Supervisors for La. State Univ. Agric. & Mech. Coll. v. Smack Apparel Co., 550 F.3d 465,484 (5th Cir. 2008).
(24.) Motion Picture Ass'n of Am. Inc. v. Respect Sportswear Inc., 83 U.S.P.Q.2d (BNA) 1555, 1564 (T.T.A.B. 2007).
(25.) Burck v. Mars, Inc., 571 F. Supp. 2d 446 (S.D.N.Y. 2008) (denying Mars' motion to dismiss plaintiff's false endorsement claim).
(26.) Standard Process, Inc. v. Total Health Discount, Inc., 559 F. Supp. 2d 932, 941 (E.D. Wis. 2008).
(27.) Amoco Oil Co. v. Rainbow Snow, 748 F.2d 556, 559 (10th Cir. 1984). Rainbow Snow sold its snow cones from fourteen round, ten-by-six-foot booths, which were blue with a 180-degree, red-orange-yellow-green rainbow appearing on the upper half of the face of the booth and prominently displayed the name "Rainbow Snow" in white letters below the rainbow. Id. at 557. Signs at Amoco's Rainbo gas stations displayed the word "Rainbo" in white, with the word appearing against a black background and below a red-orange-yellow-blue truncated rainbow logo. Id.
(28.) NFL v. Governor of Del., 435 F. Supp. 1372, 1376, 1380-81 (D. Del. 1977). The lottery game was called "Scoreboard" and the individual games were identified as "Football Bonus," "Touchdown," and "Touchdown II." Id. at 1380.
(29.) The cards on which the customers of the Delaware Lottery marked their betting choices identified the next week's NFL football games by the names of the cities whose NFL teams were scheduled to compete against each other (e.g., Washington v. Baltimore). Id. The parties stipulated that, in the context in which they appeared, these geographic names were intended to refer to, and consumers understood them to refer to, particular NFL football teams. Id. This was enough for the court to find sponsorship or affiliation confusion because, "[a]pparently, in this day and age when professional sports teams franchise pennants, teeshirts, helmets, drinking glasses and a wide range of other products, a substantial number of people believe, if not told otherwise, that one cannot conduct an enterprise of this kind without NFL approval." Id. at 1381. The court therefore entered a limited injunction "requiring the Lottery Director to include on Scoreboard tickets, advertising and any other materials prepared for public distribution a clear and conspicuous statement that Scoreboard [was] not associated with or authorized by the National Football League." Id.
(30.) Pebble Beach Co. v. Tour 18 I Ltd., 155 F.3d 526, 526 (5th Cir. 1998).
(31.) Sara Lipka, PG-13? Not This College. Or That One. Or..., CHRON. HIGHER EDUC., June 26, 2009, at 1; William McGeveran, Trademarks, Movies, and the Clearance Culture, INFO/LAW, July 2, 2009, http://blogs.law.harvard.edu/infolaw/2009/07/02/tmmovie-clearance/.
(32.) McGeveran, supra, note 31. Apparently Harvard was less troubled about a student being depicted as having stolen money to pay its tuition: the movie was retitled Stealing Harvard.
(33.) See also Vince Horiuchi, HBO Disputes Trademark Infringement in "Big Love,' SALT LAKE TRIB., July 8, 2009 (discussing a lawsuit filed by the University of Utah over the three-second depiction of a fictional research report bearing the University of Utah logo).
(34.) Michael Cieply, Despite Big Names, Prestige Film Falls Through, N.Y. TIMES, July 2, 2009, at B1.
(35.) We mean really playing Little League baseball, not just playing little league-quality baseball. The latter, as any good White Sox fan knows, would certainly have been believable.
(36.) OK, we said we were sorry before, but now we really are.
(37.) See Coats v. Holbrook, 7 N.Y. Ch. Ann. 713, 717 (1845) (noting trademark law's purpose of preventing a defendant from "attract[ing] to himself the patronage that without such deceptive use of such names.., would have inured to the benefit of [the plaintiff]"). For a longer discussion of traditional trademark law principles, see generally Mark P. McKenna, The Normative Foundations of Trademark Law, 82 NOTRE DAME L. REV. 1839 (2007).
(38.) 201 F. 510 (7thCir. 1912).
(39.) Id. at 515.
(40.) See Robert G. Bone, Hunting Goodwill: A History of the Concept of Goodwill in Trademark Law, 86 B.U.L. REV. 547, 575-79 (2006).
(42.) 247 F. 407, 408 (2d Cir. 1917).
(43.) Id. at 409.
(44.) Id. at 409-10
(45.) 26 F.2d 972 (2d Cir. 1928).
(46.) Id. at 974.
(48.) Frank I. Schechter, The Rational Basis of Trademark Protection, 40 HARV. L. REV. 813, 833 (1927) (arguing that unique marks deserved property-like protection against any use in any context); cf. Sara Stadler Nelson, The Wages of Ubiquity in Trademark Law, 88 IOWA L. REV. 731, 735 (2003) (arguing that an implication of Schechter's theory is that companies that sell many types of products have diluted their own marks).
(49.) See, e.g., Edward C. Lukens, The Application of the Principles of Unfair Competition to Cases of Dissimilar Products, 75 U. PA. L. REV. 197, 199 (1927).
(50.) Some commentators simply assumed away this problem. Lukens, supra note 49, at 204 ("As commercial organization becomes more complex, it is becoming more usual for a corporation to manufacture or sell a wide variety of products. Many companies produce articles that have no similarity, nor any relationship beyond the fact that they are so produced. Such a concern frequently applies the same trade-name to all its products in the hope that the good-will of the older products will attach to the newer ones. The public has become so accustomed to the idea of dissimilar articles being produced by the same company that it is hardly surprised at any combination whatever." (emphasis added)); see also George W. Goble, Where and What a Trade-Mark Protects, 22 ILL. L. REV. 379, 388 (1927) (arguing against the requirement that the defendant's goods be of the "same class" as the plaintiffs: "It seems reasonable to suppose that ordinarily identity of trade name or mark in itself would sufficiently relate them to cause mental association as to the manufacture or origin of the goods, dissimilar and unrelated though the goods may otherwise be").
(51.) Licensing posed serious conceptual problems in traditional trademark law because courts in that era viewed "source" literally. When plaintiffs who had licensed production of products bearing their marks sought to enforce their trademark rights, courts were faced with two parties, neither of which was the actual source of the products bearing the mark at issue. It was difficult for courts in these cases to see how a mark owner deserved relief when it arguably was engaging in the same type of deception as the accused infringer. It was also difficult to see how the defendant's use diverted consumers who otherwise would have gone to the mark owner when the mark owner was not, in fact, the source of the products. For this reason, licensing traditionally was forbidden. See McKenna, supra note 37, at 1893-95.
(52.) Courts did something very similar to legitimate licensing practices. In order to distinguish uses by affiliated companies from infringing uses by third parties, courts began to hold that, even when it did not actually produce the products bearing its mark, a mark owner could be considered the legal "source" of those products if it exercised sufficient control over their quality. See, e.g., Keebler Weyl Baking Co. v. J. S. Ivins' Son, Inc., 7 F. Supp. 211, 214 (E.D. Pa. 1934) ("An article need not be actually manufactured by the owner of the trade-mark it being enough that it is manufactured under his supervision and according to his directions thus securing both the right of the owner and the right of the public."). Congress later codified this understanding of source in section 5 of the Lanham Act, which provides that use of a mark by "related companies" inures to the benefit of the mark owner. 15 U.S.C. [section] 1055 (2006). A "related company" in this context is one "whose use of a mark is controlled by the owner of the mark with respect to the nature and quality of the goods or services on or in connection with which the mark is used." 15 U.S.C. [section] 1127 (2006). Thus, in modern terms, the legal source of a product is not necessarily the actual producer of a product but instead the entity exercising control over the quality of products bearing a particular mark. That entity might be related to the actual producer only by contract.
(53.) 300 F. 509 (6th Cir. 1924).
(54.) Id. at 511. For reasons not entirely clear to us, many of the early sponsorship or affiliation cases involved the use of a mark previously known as the title of a magazine. See, e.g., Triangle Publ'ns, Inc. v. Rohrlich, 167 F.2d 969, 972 (2d Cir. 1948) (finding defendant's use of "Miss Seventeen Foundations Co." as the partnership name to make and sell girdles, and "Miss Seventeen" as the trademark for the girdles, infringing of the magazine publisher's fights: "[T]he defendants' use of 'Seventeen' created a likelihood that the public would erroneously believe that defendants' dresses were advertised in or sponsored by the magazine and that the plaintiff's reputation and good will would thereby be injured."); Esquire, Inc. v. Esquire Bar, 37 F. Supp. 875, 876 (S.D. Fla. 1941) (finding the defendant's use of Esquire for a bar infringing of the magazine publisher's rights: "The Court finds from the evidence that the defendant's use of plaintiffs name 'Esquire' is calculated to, and does, cause the public (not otherwise fully informed) to believe there is some connection between the two, either that the plaintiff owns or controls the business of the defendant, or sponsors it, or has given leave to conduct the business under some contract, and that the defendant's business has the approval of plaintiff, or that the defendant's business is in some manner related to the plaintiff's business, Esquire, Inc., and thereby constitutes unfair competition in violation of plaintiffs rights."). Magazine cases remain an active part of the trademark docket. See, e.g., PRL USA Holdings, Inc. v. U.S. Polo Ass'n, Inc., 520 F.3d 109 (2d Cir. 2008); Entrepreneur Media, Inc. v. Smith, 279 F.3d 1135 (9th Cir. 2002); Westchester Media v. PRL USA Holdings, Inc., 214 F.3d 658 (5th Cir. 2000).
(55.) In Radio Co. of America v. R.CA. Rubber Corp., for example, the court conceded that, in a case involving noncompeting goods, "[s]trictly speaking, we are not dealing with unfair competition, in the usual sense of that term." 114 F. Supp. 162, 164 (N.D. Ohio 1953).
(56.) Most courts simply assumed that the harm from confusion regarding sponsorship or affiliation was the same as that caused by confusion regarding actual source. See, e.g., Triangle Publ'ns, Inc., 167 F.2d at 973 ("[Y]he wrong of the defendant consisted in imposing upon the plaintiff a risk that the defendant's goods would be associated by the public with the plaintiff, and it can make no difference whether that association is based upon attributing defendant's goods to plaintiff or to a sponsorship by the latter when it has been determined that plaintiff had a fight to protection of its trade name.").
(57.) Courts have made little attempt to give greater content to "sponsorship" or "affiliation." See Adolph Kastor & Bros. v. Fed. Trade Comm'n, 138 F.2d 824, 825 (2d Cir. 1943) ("At the outset we hold therefore that the word, 'Scout,' when applied to a boy's pocket knife, suggests, if indeed it does not actually indicate, that the knife is in some way sponsored by the Boy Scouts of America." (emphasis added)); Copacabana, Inc. v. Breslauer, 101 U.S.P.Q. (BNA) 467, 468 (Dec. Comm'r Pat. 1954) (rejecting application to register Copacabana for cosmetics in light of prior use of Copacabana for nightclub and restaurant despite finding that cosmetics are "entirely different" from the nightclub and restaurant because customers may assume that the cologne was "made by, sponsored by, or in some way connected with" Copacabana, Inc. (emphasis added)).
(58.) Indeed, that assumption is so strong that one commentator has argued that the trademark owner should bear responsibility for torts committed by the franchisee. Lynn M. LoPucki, Toward a Trademark-Based Liability System, 49 UCLA L. REV. 1099, 1099 (2002). LoPucki reasons that consumers attribute local franchisee behavior to the national chain, and since the chain intentionally benefits from that attribution, it should also bear the costs of that relationship. Id.
(59.) This was the conclusion of the court a century ago in denying the trademark owner relief. Borden Ice Cream Co. v. Borden's Condensed Milk Co., 201 F. 510, 514 (7th Cir. 1912). That case undoubtedly would come out differently today, however.
(60.) See 4 J. THOMAS MCCARTHY, MCCARTHY ON TRADEMARKS AND UNFAIR COMPETITION [section] 24:15 (4th ed. 2009) ("If, for example, the infringer's V-8 vitamin pills make the purchaser's child sick, she may well carry over an unfavorable reaction to plaintiff's V-8 vegetable juice."). A related argument focuses on the mark owner's control over its reputation, regardless of the current quality of the third party's goods or services. See, e.g., Wesley-Jessen Div. of Schering Corp. v. Bausch & Lomb Inc., 698 F.2d 862, 867 (7th Cir. 1983) ("Courts readily find irreparable harm in trademark infringement cases because of the victim's inability to control the nature and quality of the infringer's goods, not because the infringer's goods are necessarily inferior. Even if the infringer's goods are of high quality, the victim has the right to insist that its reputation not be imperiled by another's actions." (citations omitted)).
(61.) For a detailed discussion of the relevant marketing literature, see Mark P. McKenna, Testing Modern Trademark Law "s Theory of Harm, 95 IOWA L. REV. 63 (2009).
(62.) Brand extension refers to the practice of introducing new products under existing brands. Some researchers distinguish between brand extensions and line extensions. In this terminology, new products introduced in the same basic-level category as the parent brand would be line extensions, and new products in different basic-level categories would be brand extensions. Barbara Loken & Deborah Roedder John, Diluting Brand Beliefs: When Do Brand Extensions Have a Negative Impact?, 57 J. MARKETING 71, 74 n.3 (1993). The "basic" level is the one most easily recognized and discriminated by consumers. Id. at 74. Thus, the basic level category for Coca-Cola might be "soda." A line extension then would be a new type of soda offered under the Coca-Cola mark, such as Diet Coke. A brand extension would involve introduction of a new juice product under the Coca-Cola mark. Brand extensions, in this terminology, would thus be more remote from the original products than would line extensions. We use "brand extension" generally to refer to both practices.
(63.) In one study by Jean Romeo, for example, negative information about a brand extension had no significant negative effect on subjects' evaluations of the family brand as compared to their evaluations of the brand before learning about the extension. In fact, the only significant effect Romeo found was the positive effect negative information about an extension in a dissimilar product category (sherbert as an extension of Tropicana juice) had on the family brand. Jean B. Romeo, The Effect of Negative Information on the Evaluations of Brand Extensions and the Family Brand, 18 ADVANCES IN CONSUMER RES. 399, 404-05 (1991). Kevin Keller and David Aaker similarly failed to find any difference between the core brand evaluations of subjects who received negative information about an extension and a control group that had not received any extension information. See Kevin Lane Keller & David A. Aaker, The Effects of Sequential Introduction of Brand Extensions, 29 J. MARKETING RES. 35, 43, 46 (1992). In Keller and Aaker's terminology, a "core" brand refers to the original brand in its original context.
(64.) The situation is somewhat more complicated with respect to multiple or successive extensions, but the lesson is largely the same: extension information is unlikely to affect global assessments of a core brand. In Keller and Aaker's study, successful brand extensions increased evaluations of later extensions and of the core brand itself, at least when the core brand was of average quality. Keller & Aaker, supra note 63, at 46. Unsuccessful intervening extensions led to lower evaluations of later proposed extensions, but they did not affect evaluations of the core brand, regardless of the core brand's quality level. Id. Thus, the only apparent risk to a core brand from failed extension is that consumers will evaluate future extensions more negatively than they otherwise might have. Moreover, subjects tended to find the core brand owner equally credible even after receiving information about a brand extension they regarded as a bad fit. Id. at 46-47.
(65.) See Tulin Erdem, An Empirical Analysis of Umbrella Branding, 35 J. MARKETING RES. 339, 347 (1998) (finding that variance in the quality of toothbrushes given away as free samples from the owner of a known toothpaste brand had some cross-category effects (i.e., consumers updated their quality expectations of the toothpaste and bought less of it), but that those effects were "small in magnitude"). Erdem's study relied on purchase data after exposure to free toothbrush samples provided explicitly by the brand owner. Id. at 345. Thus, not only was there no doubt regarding the source of the toothbrushes, the brand owner aggressively tied the two products together. It is not clear whether the same results would have ensued if consumers had found the similarly branded toothbrushes on their own.
(66.) Cf. Thomas R. Lee, Glenn L. Christensen & Eric D. DeRosia, Trademarks, Consumer Psychology, and the Sophisticated Consumer, 57 EMORY L.J. 575 (2008) (finding that consumers are more sophisticated about purchasing decisions than trademark law generally assumes).
(67.) Joseph W. Chang, Will a Family Brand Image Be Diluted by an Unfavorable Brand Extension? A Brand Trial-Based Approach, 29 ADVANCES IN CONSUMER RES. 299, 302 (2002). Dish-washing detergent? Sprite? No, we don't know what they were thinking either.
(68.) Id. Positive evaluations of Sprite orangeades, on the other hand, enhanced consumers' attitudes towards Sprite lemonades.
(69.) Incongruity here refers to use of the brand for products that do not fit with one or more specific brand associations. Use of Neutrogena for sandpaper, for example, would be incongruous with the belief that Neutrogena is "mild." See Helge Thorbjornsen, Brand Extensions: Brand Concept Congruency and Feedback Effects Revisited, 14 J. PRODUCT & BRAND MGMT. 250, 250-51 (2005); see also Henrik Sjodin & Fredrik Torn, When Communication Challenges Brand Associations: A Framework for Understanding Consumer Responses to Brand Image Incongruity, 5 J. CONSUMER BEHAV. 32, 38 (2006).
(70.) Sjodin & Torn, supra note 69, at 38. The authors explain this somewhat counterintuitive result by suggesting consumers generally use a sub-typing strategy to resolve incongruous information. Id. That is, when a brand extension is atypical, consumers are likely to resolve the incongruity by storing the information about the extension in a separate cognitive category. When consumers create such sub-types, the parent brand is effectively insulated from feedback. Id. But whatever the explanation, the conclusion that incongruous information will not affect consumer brand beliefs is consistent with other research demonstrating that well-known brands are quite resistant to change. See Stephen J. Hoch, Product Experience Is Seductive, 29 J. CONSUMER RES. 448, 451 (2002) ("Using a simple associative learning procedure, [van Osselaer and Alba] showed that, in a few trials, people learn brand associations that later block the learning of new predictive attribute associations."); Maureen Morrin & Jacob Jacoby, Trademark Dilution: Empirical Measures for an Elusive Concept, 19 J. PUB. POL'Y & MARKETING 265, 274 (2000) ("It appears that very strong brands are immune to dilution because their memory connections are so strong that it is difficult for consumers to alter them or create new ones with the same brand name.").
(71.) See Girish N. Punj & Clayton L. Hillyer, A Cognitive Model of Customer-Based Brand Equity .]'or Frequently Purchased Products: Conceptual Framework and Empirical Results, 14 J. CONSUMER PSYCHOL. 124, 125 (2004); see also Laura R. Bradford, Emotion, Dilution, and the Trademark Consumer, 23 BERKELEY TECH. L.J. 1227, 1260 (2008).
(72.) A different argument that producers are harmed by uses of the same mark for noncompetitive goods might focus on the costs the producer would be required to incur in re-educating consumers about the quality of its goods if and when it entered the same market as the junior user. If, for example, Borden was unable to prevent another company from using the Borden name for ice cream and was forced to enter the ice cream market under a different name (since the other company would have established priority in the ice cream market), Borden would have to educate consumers about the quality of its new ice cream product instead of being able to rely on use of the Borden name to transfer quality messages to the new products. We think that our focus on uses that indicate control over quality covers those cases in which the quality information would transfer to the new product.
(73.) On the organizational choices associated with franchise relationships, see, for example, James A. Brickley & Frederick H. Dark, The Choice of Organizational Form: The Case of Franchising, 18 J. FIN. ECON. 401, 403-07 (1987); Paul H. Rubin, The Theory of the Firm and the Structure of the Franchise Contract, 21 J.L. & ECON. 223 (1978); Mira Wilkins, The Neglected Intangible Asset: The Influence of the Trade Mark on the Rise of the Modern Corporation, 34 Bus. HIST. 66, 87-88 (1992).
(74.) See, e.g., Schlotzsky's, Ltd. v. Sterling Purchasing & Nat'l Distrib. Co., 520 F.3d 393 (5th Cir. 2008) (finding misrepresentations by distributor as to nature of relationship with franchisor sufficient to state a Lanham Act claim).
(75.) See, e.g., Bristol-Myers Squibb Co. v. McNeil-P.P.C., Inc., 973 F.2d 1033 (2d Cir. 1992). That case involved a lawsuit by makers of "Excedrin PM" nighttime pain reliever against makers of "Tylenol PM" nighttime pain reliever. The court separately analyzed the house brands ("Excedrin" vs. "Tylenol"), the mark "PM," the color of the lettering, and the color of the packaging; the pills themselves were also differently colored. Id. at 1039-46. The court found no confusion because of the prominence of and difference between "Excedrin" and "Tylenol," despite the similarity of the trade dress and the identity of the "PM" marks. Id. at 1047.
(76.) See Lisa P. Ramsey, Intellectual Property Rights in Advertising, 12 MICH. TELECOMM. & TECH. L. REV. 189, 251-52 (2006) (making this point in the context of slogans); see also George Miaoulis & Nancy D'Amato, Consumer Confusion and Trademark Infringement, 42 J. MARKETING 48, 54 (1978) (finding, in the context of competing goods, that the primary cue for association between two brands was not the name but the visual appearance).
(77.) See Arena Kirmani, Sanjay Sood & Sheri Bridges, The Ownership Effect in Consumer Responses to Brand Line Stretches, 63 J. MARKETING 88, 94-95 (1999) (finding that sub-branding was sufficient to insulate the BMW and Acura brands from any negative feedback); see also Sandra J. Milberg, C. Whan Park & Michael S. McCarthy, Managing Negative Feedback Effects Associated with Brand Extensions: The Impact of Alternative Branding Strategies, 6 J. CONSUMER PSYCHOL. 119 (1997) (finding that sub-branding may prevent negatively evaluated extensions from harming the parent brand).
(78.) Cf. Lillian R. BeVier, Competitor Suits for False Advertising Under Section 43(a) of the Lanham Act: A Puzzle in the Law of Deception, 78 VA. L. REV. 1, 2-3 (1992) (arguing that "competitors' incentives to sue are not correlated with the likelihood of consumer harm"); Ross D. Petty, Supplanting Government Regulation with Competitor Lawsuits: The Case of Controlling False Advertising, 25 IND. L. REV. 351 (1991).
(79.) But see Irene Calboli, The Sunset of "Quality Control" in Modern Trademark Licensing, 57 AM. U. L. REV. 341 (2007) (arguing that trademark owners should not be required to control the quality of their licensed products).
(80.) Compare Irene Calboli, Trademark Assignment "'With Goodwill": A Concept Whose Time Has Gone, 57 FLA. L. REV. 771 (2005), Irene Calboli, What If, After All, Trademarks Were "Traded in Gross"?, 2008 MICH. ST. L. REV. 345, and Allison Sell McDade, Trading in Trademarks: Why the Anti-Assignment in Gross Doctrine Should Be Abolished when Trademarks Are Used as Collateral, 77 TEX. L. REV. 465 (1998) (all arguing in favor of unrestricted licensing and sale of trademarks), with Mark A. Lemley, The Modern Lanham Act and the Death of Common Sense, 108 YALE L.J. 1687, 1709-10 (1999) (defending the rule against naked licensing).
(81.) See 15 U.S.C. [section] 1055 (2006) (providing that use of a mark by "related companies" inures to the benefit of the mark owner). A "related company" in this context is one "whose use of a mark is controlled by the owner of the mark with respect to the nature and quality of the goods or services on or in connection with which the mark is used." 15 U.S.C. [section] 1127 (2006).
(82.) Perhaps we're wrong about this and consumers, increasingly familiar with product placement and other complicated arrangements between advertisers and content creators, do in fact derive quality-related information from the presence of Coca-Cola cups. Economists have argued that the fact that a company is willing to spend money advertising is itself a signal of quality. See Phillip Nelson, Advertising as Information, 82 J. POL. ECON. 729, 730 (1974); I.P.L. Png & David Reitman, Why Are Some Products Branded and Others Not?, 38 J.L. & ECON. 207 (1995). One of us has elsewhere explained the self-limiting nature of this claim. Lemley, supra note 80, at 1692. But even if consumers assume that Coke guarantees the quality of American Idol, the general point holds: many of the kinds of uses that currently give rise to sponsorship or affiliation claims convey qualitatively different kinds of information than those that suggest responsibility for quality. And our formulation doesn't require guessing about whether a particular use conveys quality-related information--the plaintiff always has the opportunity to prove it.
(83.) Nonetheless, some courts have concluded that the fact of assumed permission is itself an actionable trademark harm. Pebble Beach Co. v. Tour 18 I Ltd., 155 F.3d 526, 544 (5th Cir. 1998) ("For a party to suggest to the public, through its use of another's mark or a similar mark, that it has received permission to use the mark on its goods or services suggests approval, and even endorsement, of the party's product or service and is a kind of confusion the Lanham Act prohibits."). In that case, the evidence of endorsement was merely based on similarity of the golf course layouts, and was not based on any affirmative representation by the defendant.
(84.) Nicole L. Votolato & H. Rao Unnava, Spillover of Negative Information on Brand Alliances, 16 J. CONSUMER PSYCHOL. 196, 196 (2006).
(85.) Id. at 198.
(87.) Previous research suggested to the authors that consumers might react differently to different types of negative information--information about competence, on the one hand, and moral misdeeds on the other. See id. at 197; see also Tom J. Brown & Peter A. Dacin, The Company and the Product: Corporate Associations and Consumer Product Responses, 61 J. MARKETING 68 (1997); Bogdan Wojciszke, Hanna Brycz & Peter Borkenau, Effects of Information Content and Evaluative Extremity on Positivity and Negativity Biases, 64 J. PERSONALITY & SOC. PSYCHOL. 327 (1993). Specifically, this earlier research suggested that consumers react more negatively to competence-based information than moral failures when the target of the information is a company; just the reverse is true when the target of the information is a person. See Votolato & Unnava, supra note 84, at 197.
(88.) Votolato & Unnava, supra note 84, at 201 (emphasis added). These findings, as the authors also note, may help explain why spillover effects are not frequently reported in practice. Id.
(89.) See 4 MCCARTHY, supra note 60, [section] 24:9 (discussing the circularity problem that occurs when consumer perception drives licensing law but licensing law drives consumer perception); James Gibson, Risk Aversion and Rights Accretion in intellectual Property Law, 116 YALE L.J. 882, 907-27 (2007). For criticism of property-based market allocation claims in IP more generally, see, for example, Yochai Benkler, Some Economics of Wireless Communications, 16 HARV. J.L. & TECH. 25 (2002); Mark A. Lemley, Property, Intellectual Property, and Free Riding, 83 TEX. L. REV. 1031 (2005).
(90.) We evaluate such claims in more detail in a separate paper. See Mark McKenna & Mark A. Lemley, Owning Markets?: Trademark Law and Market Foreclosure (working paper 2010).
(91.) See ROBERT P, MERGES ET AL., INTELLECTUAL PROPERTY IN THE NEW TECHNOLOGICAL AGE 754-58 (rev. 4th ed. 2007) (detailing these cases and analyzing them under trademark principles); Stacey L. Dogan & Mark A. Lemley, The Merchandising Right: Fragile Theory or Fait Accompli?, 54 EMORY L.J. 461, 471-89 (2005) (same).
(92.) Bantam Dell, the book publisher, was established in 1945. Bantam Dell Publishing Group, Publishers' Catalogues, http://www.lights.ca/publisher/db/8/1628.html (last visited Sept. 6, 2009). Michael Dell founded PC's Limited in 1984 and the company changed its name to Dell Computer Corporation in 1988. Dell, Wikipedia, http://en.wikipedia.org/wiki/Dell (last visited Sept. 6, 2009).
(93.) Dilution law raises similar problems, but dilution protection is, at least in theory, available only to truly famous marks. 15 U.S.C. [section] 1125(c)(1) (2006).
(94.) For a detailed discussion of the evidence on all sides, see Lee et al., supra note 66; cf. Barton Beebe, Search and Persuasion in Trademark Law, 103 MICH. L. REV. 2020 (2005) (criticizing trademark theorists for making inconsistent assumptions about the sophistication of consumers).
(95.) Several studies have determined that use of a sub-brand or other mechanism for differentiating an extension from the parent brand effectively insulates the parent brand from any feedback effects. See, e.g., Kirmani et al., supra note 77 (finding sub-brands effective in preventing negative feedback from extension to parent brand); Milberg et al., supra note 77 (stating that sub-branding may prevent negatively evaluated extensions from harming the parent brand). This research suggests that consumers are relatively adept at recognizing attempts to differentiate, and that they are able to categorize brand attitudes finely when they have reason to differentiate. Thus, any risk of confusion is attenuated to the extent the uses entail signals of differentiation.
(96.) The media often contributes to this problem. See, e.g., Cieply, supra note 34 (accepting without question that "approval [from MLB] is crucial in a baseball film that intends to use protected trademarks").
(97.) See Katya Assaf, The Dilution of Culture and the Law of Trademarks, 49 IDEA 1, 41-42 (2008).
(98.) This paragraph and the paragraph that follows are adapted from Stacey L. Dogan & Mark A. Lemley, Grounding Trademark Law Through Trademark Use, 92 IOWA L. REV. 1669, 1694-95 (2007).
(99.) For a more detailed analysis of the retailer issue and an argument that Internet intermediaries play the same role as retailers for trademark purposes, see Eric Goldman, Brand Spillovers, 22 HARV. J.L. & TECH. 381 (2009).
(100.) Not all courts have held similarly--some even seemed to reject the claim on the ground the alleged confusion was not material. See, e.g., Univ. of Pittsburgh v. Champion Prods., Inc., 566 F. Supp. 711, 721 (W.D. Pa. 1983) ("There is no evidence that the consumer cares who has made the soft goods or whether they were made under license.").
(101.) See Dogan & Lemley, supra note 91, at 472-75 (detailing this history).
(102.) Id. at 472-78.
(103.) See, e.g., AMF Inc. v. Sleekcraft Boats, 599 F.2d 341, 348-49 (9th Cir. 1979); Polaroid Corp. v. Polarad Elecs. Corp., 287 F.2d 492, 495 (2d Cir. 1961).
(104.) See, e.g., King of the Mountain Sports, Inc. v. Chrysler Corp., 185 F.3d 1084, 1090 (10th Cir. 1999) ("[I]n the rare, pure sponsorship action, other factors [than similarity]--such as the relation in use and the manner of marketing between the goods or services and the degree of care likely to be exercised by purchasers--have little importance.").
(105.) Barton Beebe's analysis of the case law suggests that we may have this problem even with true source confusion cases because outcomes tend to be driven by the court's focus on intent. Barton Beebe, An Empirical Study of the Multifactor Tests for Trademark Infringement, 94 CAL. L. REV. 1581 (2006). If causation in fact runs in the direction this suggests, all of trademark infringement analysis may be fundamentally flawed.
(106.) Anheuser-Busch, Inc. v. Balducci Publ'ns, 28 F.3d 769, 775 (8th Cir. 1994) (bolstering its finding that confusion was likely by citing the results of a survey showing over half of the respondents thought Balducci needed Anheuser-Busch's approval to publish the ad). Unfortunately, that case is not the only one in which a court accepted this sort of evidence. See also Anheuser-Busch, Inc. v. VIP Prods., LLC, No. 4:08cv0358 TCM, 2008 WL 4619702, at *2 (E.D. Mo. Oct. 16, 2008) (enjoining a dog toy labeled "Buttwiper" because a survey showed that 30.3% of consumers "mistakenly believed that VIP's 'Buttwiper' [was] manufactured and marketed by, or with the approval of, [Anheuser-Busch] or that there [was] some affiliation between 'Buttwiper' and [Anheuser-Busch]"). But see NFL Props., Inc. v. ProStyle, Inc., 16 F. Supp. 2d 1012, 1019 (E.D. Wis. 1998) (rejecting surveys of this sort as calling for a legal conclusion); Novo Nordisk of N. Am., Inc. v. Eli Lilly & Co., No. 96 Civ. 5787 (BSJ), 1996 WL 497018, at *6 n.24 (S.D.N.Y. Aug. 30, 1996) (discounting some of the survey's conclusions because of "serious flaws in the questions").
(107.) Beebe, supra note 105.
(108.) See, e.g., Ty Inc. v. Perryman, 306 F.3d 509, 512 (7th Cir. 2002) (Posner, J.) (rejecting sponsorship dilution claim because "in that attenuated sense of free riding, almost everyone in business is free riding").
(109.) See Lemley, supra note 80, at 1711-13 (documenting numerous examples in addition to the ones we discuss here).
(110.) 6 MCCARTHY, supra note 60, [section] 31:146; see also L.L. Bean, Inc. v. Drake Publishers, Inc., 811 F.2d 26, 30 (1st Cir. 1987) ("'[T]rademarks offer a particularly powerful means of conjuring up the image of their owners, and thus become an important, perhaps at times indispensable, part of the public vocabulary.'" (quoting Robert C. Denicola, Trademarks As Speech: Constitutional Implications of the Emerging Rationales for the Protection of Trade Symbols, 1982 WIS. L. REV. 158, 195-96)).
(111.) And some trademark owners undoubtedly deserve to be made fun of.
(112.) See Keith Aoki, How the Worm Dreams Itself to Be American: Reflections on the Relationship Between the Expanding Scope of Trademark Protection and Free Speech Norms, 3 LOY. L.A. ENT. L.J. 523, 528 (1997); Denicola, supra note 110, at 195-96; Rochelle Cooper Dreyfuss, Expressive Genericity: Trademarks as Language in the Pepsi Generation, 65 NOTRE DAME L. REV. 397, 397-98 (1990); Wilkins, supra note 73, at 87-88; Steven M. Cordero, Note, Cocaine-Cola, the Velvet Elvis, and Anti-Barbie: Defending the Trademark and Publicity Rights to Cultural Icons, 8 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. 599, 601-03 (1998); Tara J. Goldsmith, Note, What's Wrong with This Picture? When the Lanham Act Clashes with Artistic Expression, 7 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. 821, 877-78 (1997). There is significant sociological literature on the process by which consumers "recode" products and brands, imbuing them with independent social significance in a way frequently not intended by the trademark owner. For a discussion, see Michael Madow, Private Ownership of Public Image: Popular Culture and Publicity Rights, 81 CAL. L. REV. 125, 140 (1993).
(113.) In MasterCard International Inc. v. Nader 2000 Primary Committee Inc., 70 U.S.P.Q.2d (BNA) 1046, 1048 (S.D.N.Y. 2004), for example, Ralph Nader used a parody of the MasterCard "priceless" advertising campaign to attract viewer attention and make a political point.
(114.) For constitutional arguments, see generally Denicola, supra note 110; Wendy J. Gordon, A Property Right in Self-Expression: Equality and Individualism in the Natural Law of Intellectual Property, 102 YALE L.J. 1533, 1537 (1993); Mark A. Lemley & Eugene Volokh, Freedom of Speech and Injunctions in Intellectual Property Cases, 48 DUKE L.J. 147, 171-78 (1998); Lisa P. Ramsey, Increasing First Amendment Scrutiny of Trademark Law, 61 SMU L. REV. 381,417, 424 (2008).
(115.) Trademark law incorporates a variety of "defenses" that attempt to soften its impact on First Amendment interests, but those defenses are somewhat notorious for their indeterminacy. See, e.g., William McGeveran, Rethinking Trademark Fair Use, 94 IOWA L. REV. 49, 49 (2008) (arguing that while most cases raising these free speech issues have been decided in favor of speech, "the procedural structure of trademark law's various 'fair use' doctrines.., generate excessive ambiguity and prolong litigation," thereby "discourag[ing] speakers from using trademarks expressively in the first place, creating a classic chilling effect").
(116.) See, e.g., Assaf, supra note 97, at 77-79.
(117.) See, e.g., "Slumdog Millionaire" Throws Product Placement into Reverse, VARIETY, Oct. 30, 2008, http://weblogs.variety.com/hal/2008/10/slumdog-million.html (documenting the efforts to which trademark owners went to have their marks removed from the background of Slumdog Millionaire).
(118.) New Kids on the Block v. News Am. Publ'g, Inc., 971 F.2d 302, 307 (9th Cir. 1992); see also Yankee Publ'g Inc. v. News Am. Publ'g Inc., 809 F. Supp. 267, 275-76 (S.D.N.Y. 1992) (holding that the First Amendment is implicated by expressive, rather than commercial, uses of a trademark); Denicola, supra note 110, at 194-96.
(119.) See Jessica Litman, Breakfast with Batman: The Public Interest in the Advertising Age, 108 YALEL.J. 1717, 1729-30 (1999).
(120.) 15 U.S.C. [section] 1125(a)(1)(B) (2006); Southland Sod Farms v. Stover Seed Co., 108 F.3d 1134 (9th Cir. 1997). The definition of falsity is itself surprisingly malleable. For examples, see Rebecca Tushnet, It Depends on What the Meaning of "False" Is: Falsity and Misleadingness in Commercial Speech Doctrine, 41 LOY. L.A.L. REV. 227, 231-48 (2007).
(121.) One significant limitation on the scope of false advertising claims is the requirement that a plaintiff be in competition with the defendant to have standing to assert a claim. See, e.g., Barrus v. Sylvania, 55 F.3d 468, 470 (9th Cir. 1995); Serbin v. Ziebart Int'l Corp., 11 F.3d 1163, 1177-79 (3d Cir. 1993). For criticism of this rule, see BeVier, supra note 78.
(122.) 15 U.S.C. [section] 1116 (2006). The Eleventh Circuit has read the Supreme Court decision in eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388 (2006), to require proof of irreparable harm and consideration of the balance of hardships before a court can enjoin infringement. N. Am. Med. Corp. v. Axiom Worldwide, Inc., 522 F.3d 1211, 1228 (11th Cir. 2008). But we do not think it makes sense, where a plaintiff has proven that the defendant's use is likely to confuse consumers about actual source or responsibility for quality, to allow the defendant to keep confusing consumers and only pay damages. Injunctions should almost always be appropriate in trademark infringement cases as we have defined them.
(123.) See, e.g., Cashmere & Camel Hair Mfrs. Inst. v. Saks Fifth Ave., 284 F.3d 302, 315 (1st Cir. 2002) (noting the requirement that the plaintiff prove materiality and that "one method of establishing materiality involves showing that the false or misleading statement relates to an inherent quality or characteristic of the product"); Southland Sod Farms, 108 F.3d at 1139 (articulating the elements of a false advertising claim, including the requirement that "the deception is material, in that it is likely to influence the purchasing decision"); Johnson & Johnson * Merck Consumer Pharms. Co. v. Smithkline Beecham Corp., 960 F.2d 294, 299 (2d Cir. 1992) (requiring proof of materiality); Johnson & Johnson v. GAC Int'l, Inc., 862 F.2d 975, 977 (2d Cir. 1988) (holding materiality is presumed if statements are literally false but must otherwise be proven).
(124.) See also Forest Group, Inc. v. Bon Tool Co., No. H-05-4127, 2008 WL 2962206, at * 7 (S.D. Tex. July 29, 2008) (holding that the marking of products with patent numbers that did not actually cover the products was false but not actionable because it was not "likely to influence purchasing decisions" of consumers); cf In re Century 21-RE/MAX Real Estate Adver. Claims Litig., 882 F. Supp. 915, 928 (C.D. Cal. 1994) (holding that an overstatement of 25,308 real estate transactions completed in 1992 (out of more than 600,000 transactions) was not likely to influence the purchasing decision of any consumer).
(125.) We think the doctrines of initial-interest and post-sale confusion can be criticized on similar grounds: one may reasonably doubt whether either type of confusion, to the extent it even exists, actually affects purchasing decisions. Cf. Glynn S. Lunney, Jr., Trademark Monopolies, 48 EMORY L.J. 367, 404-08 (1999) (arguing that post-sale confusion, at least in the context of prestige goods, is irrelevant). But full analysis of those doctrines is outside the scope of this Article.
(126.) For suggestions along somewhat similar lines, see, for example, Gibson, supra note 89, at 949 (suggesting that courts should impose a materiality requirement in sponsorship or affiliation confusion cases); Michael Grynberg, Trademark Litigation as Consumer Conflict, 83 N.Y.U.L. REV. 60 (2008). Grynberg would go further than we do, making materiality an element of every trademark case. While he is undoubtedly correct that enforcing trademarks can harm non-confused consumers, we think the law strikes the right balance in presuming materiality from actual confusion as to source or quality.
(127.) Sometimes the plaintiff's purported injury is neither a lost sale of their own goods nor confusion about the defendant's goods that materially affects the decision to purchase the defendant's goods, but instead a claim of injury to the plaintiff's mark based on some harm other than confusion. These claims fit under the dilution rubric, and we think they fit neither within the trademark infringement framework nor the false advertising framework, but instead in their own category. See 15 U.S.C. [section] 1125(c) (2006). Commentators can and do disagree over how much, if anything, belongs in that category. See Dogan & Lemley, supra note 91, at 493-95; Kenneth L. Port, The "'Unnatural" Expansion of Trademark Rights." Is a Federal Dilution Statute Necessary?, 85 TRADEMARK REP. 525 (1995); Jerre B. Swann& Theodore H. Davis, Jr., Dilution, an Idea Whose Time Has Gone," Brand Equity as Protectable Property, The New/Old Paradigm, 1 J. INTELL. PROP. L. 219 (1994); Rebecca Tushnet, Gone in Sixty Milliseconds: Trademark Law and Cognitive Science, 86 TEX. L. REV. 507 (2008). But analytically the category exists.
(128.) See Verizon Commc'ns Inc. v. Law Offices of Curtis V. Trinko, LLP, 540 U.S. 398, 408 (2004) (identifying collusion for purposes such as price fixing as the "supreme evil of antitrust'); N. Pac. Ry. Co. v. United States, 356 U.S. I, 5 (1958) (identifying price fixing, division of markets, group boycotts, and tying arrangements as unlawful activities "in and of themselves").
(129.) N. Pac. Ry. Co., 356 U.S. at 5 ("[T]here are certain agreements or practices which because of their pernicious effect on competition and lack of any redeeming virtue are conclusively presumed to be unreasonable and therefore illegal without elaborate inquiry as to the precise harm they have caused or the business excuse for their use." (emphasis added)).
(130.) See, e.g., Mark A. Lemley & Christopher R. Leslie, Categorical Analysis in Antitrust Jurisprudence, 93 IOWA L. REV. 1207, 1214-15 (2008).
(131.) For a similar proposal for differentiating between different types of derivative works in the copyright context and treating some as presumptively harmful while requiring proof of harmfulness for others, see Christopher Sprigman, Copyright and the Rule of Reason, 7 J. TELECOMM. & HIGH TECH. L. 317 (2009); see also Christina Bohannan & Herbert Hovenkamp, IP and Antitrust: Errands into the Wilderness (Univ. of Iowa Coll. of Law, Working Paper No. 09-16, 2009), available at http://ssrn.com/abstract=1377382 (proposing to import the "antitrust injury" doctrine into IP law, though not specifically trademark law).
(132.) See, e.g., Richard Craswell, Taking Information Seriously: Misrepresentation and Nondisclosure in Contract Law and Elsewhere, 92 VA. L. REV. 565,594-601 (2006).
(133.) Cf. Grynberg, supra note 126 (calling for greater attention to the interests of nonconfused consumers as well as confused ones).
(134.) White Sox Slurp on Promotion, Will Start Games at 7:11, ESPN, Oct. 11, 2006, http://sports.espn.go.com/mlb/news/story?id=2621231.
(135.) Though in light of the research on brand alliances we discussed above, this type of negative feedback seems unlikely.
(136.) The people responsible for writing section headings have now been sacked.
(137.) See, e.g., Barrus v. Sylvania, 55 F.3d 468, 470 (9th Cir. 1995); Serbin v. Ziebart Int'l Corp., 11 F.3d 1163, 1177-79 (3d Cir. 1993).
(138.) See Wal-Mart Stores, Inc. v. Samara Bros., Inc., 529 U.S. 205, 215 (2000) (distinguishing product design and product packaging trade dress and holding that only product packaging is capable of being inherently distinctive; product configuration can be protected only with evidence of secondary meaning); cf. Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763 (1992) (holding that inherently distinctive, non-functional trade dress is protectable without evidence of secondary meaning).
(139.) In antitrust, for instance, courts have long distinguished between reasonable and unreasonable restraints of trade when applying the statutory prohibition against any restraint of trade. 15 U.S.C. [section] 1 (2006) ("Every contract, combination in the form of trust or otherwise, or conspiracy, in restraint of trade or commerce among the several States, or with foreign nations, is declared to be illegal."); see, e.g., N. Pac. Ry. Co. v. United States, 356 U.S. 1, 5 (1958).
(140.) 15 U.S.C. [section] 1114(1)(a) (2006) (making liable "[a]ny person who shall, without the consent of the registrant ... use in commerce any reproduction, counterfeit, copy, or colorable imitation of a registered mark in connection with the sale, offering for sale, distribution, or advertising of any goods or services on or in connection with which such use is likely to cause confusion, or to cause mistake, or to deceive").
(141.) 15 U.S.C. [section] 1125(a)(l)(A) (2006).
(142.) As a matter of fact, courts used to make a distinction between trademark and trade dress and to require evidence of materiality in product design cases. See, e.g., Crescent Tool Co. v. Kilborn & Bishop Co., 247 F. 299, 300 (2d Cir. 1917) ("[I]t is an absolute condition to any relief whatever that the plaintiff.., show that the appearance of his wares has in fact come to mean that some particular person.., makes them, and that the public cares who does make them, and not merely for their appearance and structure." (emphasis added)); see also Lunney, supra note 125, at 376-78 (discussing this approach in more detail).
(143.) 539 U.S. 23 (2003).
(144.) Id. at 33; cf. Pharmacommunications Holdings Inc. v. Avencia Int'l Inc.,  F.C. 144 (Can.) (requiring proof of harm in Canadian law).
(145.) See, e.g., Riverbank, Inc. v. River Bank, 625 F. Supp. 2d 65, 74 (D. Mass. 2009). We think the definition of material confusion the judge used in Riverbank is likely too broad, but the fact that materiality was important at all shows a path to the goal we would like to reach in existing law.
(146.) See, e.g., Jean Wegman Bums, Confused Jurisprudence: False Advertising Under the Lanham Act, 79 B.U.L. REV. 807, 871-73 (1999) (parsing the weak legislative support for a materiality requirement in false advertising).
(147.) In our view, this substantiality requirement ought to track the requirement that an "appreciable number of ordinarily prudent consumers" be confused. McGregor-Doniger Inc. v. Drizzle Inc., 599 F.2d 1126, 1130 (2d Cir. 1979) ("[A]n appreciable number of ordinarily prudent purchasers are likely to be misled, or indeed simply confused, as to the source of the goods."); see also Int'l Ass'n of Machinists & Aerospace Workers v. Winship Green Nursing Ctr., 103 F.3d 196, 201 (1st Cir. 1996) ("[T]he law has long demanded a showing that the allegedly infringing conduct carries with it a likelihood of confounding an appreciable number of reasonably prudent purchasers exercising ordinary care."). We believe, however, that the thresholds courts have established in the confusion context are likely too low since research suggests a significant level of background confusion among consumers. See Thomas R. Lee et al., An Empirical and Consumer Psychology Analysis of Trademark Distinctiveness, 41 ARIZ. ST. L.J. 1033, 1098 (2010) (reporting that over 26% of respondents in an empirical study thought that the word "wonderful" indicated the source of the chocolate coconut macaroons depicted in the stimulus, even when the authors placed the word in small font in the bottom right-hand comer of the package and regarded the use as "non-trademark use").
(148.) We make the case against such market-ownership claims in Lemley & McKenna, supra note 90.
(149.) On the dubious legal provenance of the merchandising tight, see, for example, Dogan & Lemley, supra note 91, at 472-78.
(150.) Compare Au-Tomotive Gold, Inc. v. Volkswagen of Am., Inc., 457 F.3d 1062, 1078 (9th Cir. 2006) (assuming confusion from similarity of marks in case in which defendant sold car logo key chains, notwithstanding clear disclaimer), and Boston Athletic Ass'n v. Sullivan, 867 F.2d 22, 34 (1st Cir. 1989) (dispensing with proof of likely confusion in case involving Boston Marathon T-shirts), with WCVB-TV v. Boston Athletic Ass'n, 926 F.2d 42, 46 (1st Cir. 1991) (requiting proof of likely confusion in case involving unauthorized TV broadcast of the Boston Marathon). Tom McCarthy refers to the question as a circular one because if consumers think a particular use requires authorization then it will require authorization. 4 MCCARTHY, supra note 60, [section] 24:9. And if courts take the Jacoby view that we need only ask consumers what they think the law ought to be, see Jacob Jacoby, Sense and Nonsense in Measuring Sponsorship Confusion, 24 CARDOZO ARTS & ENT. L.J. 63 (2006), confusion analysis will be circular indeed. But this is a problem largely endemic to trademark law. Requiring proof that a representation is likely to affect consumer purchasing decisions would reduce the circularity problem since it is not just expressed consumer belief about the law, but actual consumer behavior, that plaintiffs would have to show.
(151.) See, e.g., Abdul-Jabbar v. Gen. Motors Corp., 85 F.3d 407 (9th Cir. 1996); Waits v. Frito-Lay, Inc., 978 F.2d 1093 (9th Cir. 1992).
(152.) See Ben Quarmby, Pirates Among the Second Life Islands: Why You Should Monitor the Misuse of Your Intellectual Property in Online Virtual Worlds, 26 CARDOZO ARTS & ENT. L.J. 667, 678-80 (2009) (documenting cases in which virtual world vendors sell branded "goods" without authorization).
(153.) See, e.g., Brendan James Gilbert, The Second Life of Intellectual Property 6-8 (June 7, 2009) (unpublished manuscript), available at http://ssrn.com/abstract=1409091 (documenting trademark dispute between two virtual retailers on Second Life, SLART Gallery and SLART Garden).
(154.) See, e.g., Johnson & Johnson * Merck Consumer Pharms. Co. v. Smithkline Beecham Corp., 960 F.2d 294, 299 (2d Cir. 1992); Johnson & Johnson v. GAC Int'l, Inc., 862 F.2d 975, 977 (2d Cir. 1988) (holding that materiality is presumed if statements are literally false but must otherwise be proven).
(155.) An example from a false advertising case in which the court found the false representation to be material is instructive. In Pore Wonderful LLC v. Purely Juice, Inc., plaintiff Pom Wonderful presented credible evidence that consumers sent questions to both Pom Wonderful and Purely Juice which made clear that they were "extremely concerned about product ingredients (including sugar) and how those ingredients would either improve or hinder their health." No. CV-07-02633, 2008 WL 4222045, at *9 (C.D. Cal. July 17, 2008). Further, Pom Wonderful's internal market research showed that the "primary reason its customers drank POM Wonderful [was] for the health benefits," and Purely Juice's president and founder testified that Purely Juice would lose its position in the marketplace if it could not label or advertise its product as "100% pomegranate juice" with "no sugar added." Id. In light of this evidence, the court found that consumers were likely to base their purchasing decisions on precisely the kinds of false statements contained in Purely Juice's marketing, particularly the untrue statements that its products were "100% pomegranate juice" and had "no sugar added." Id.
Mark A. Lemley, William H. Neukom Professor, Stanford Law School; partner, Durie Tangri LLP.
Mark McKenna, Associate Professor, Notre Dame Law School. We thank Dan Burk, Dick Craswell, Stacey Dogan, Jim Gibson, Rose Hagan, Laura Heymann, Tom Lee, Paul Levy, David McGowan, Joe Miller, Masaharu Miyawaki, Lisa Ramsey, Chris Sprigman, Rebecca Tushnet, and participants at a workshop at Stanford Law School, the Workshop on Trademarks and Trademark Data at the OECD (Paris, France), and the Ninth Annual Intellectual Property Scholars Conference for helpful conversations on this topic or comments on an earlier draft. We also thank Zaiba Baig, Erin Czerney, and Brandon Marsh for research assistance.