In Defense of Trademark Dilution Surveys A Post-Moseley Proposal

File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM In Defense of Trademark Dilution Surveys: A Post-Moseley Proposal TODD ANTEN∗ In 2003, the Supreme Court ruled in Moseley v. V Secret Catalogue, Inc. that for a trademark holder to successfully recover under the Federal Trademark Dilution Act (FTDA), she must prove that “actual dilution” has occurred. While consumer surveys were commonly used pre-Moseley to prove a mere “likelihood of dilution” (by showing that consumers mentally associated the two marks at issue), post-Moseley trademark holders are currently at a loss as to how a consumer survey could ever prove “actual dilution.” Many commentators, including Judge Richard Posner, suggest that constructing a consumer survey that measures “actual dilution” is not even possible. This Note argues that consumer surveys still hold a significant place in proving dilution under the FTDA, so long as a survey is properly constructed to meet Moseley’s requirement of proving “actual dilution,” and not merely mental association. This Note also rises to Judge Posner’s challenge by recommending the use of “unique value surveys” — survey questions that directly measure the lessening of a senior mark’s “uniqueness,” which is the core injury of dilution. ∗ Head Writing and Research Editor, COLUM. J.L. SOC. PROBS., 2005-2006. The author thanks Lewis Anten and Professor Clarissa Long for their invaluable suggestions. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2 Columbia Journal of Law and Social Problems [39:1 [W]e doubt that dilution of the distinctiveness of a mark is something that can be measured on an empirical basis by even the most carefully constructed survey. — Eli Lilly & Co. v. Natural Answers, Inc.1 Attempt the impossible in order to improve your work. — Bette Davis2 I. INTRODUCTION Consumer surveys have always played prominent roles in trademark disputes, helping judges peek into the minds of consumers.3 For instance, federal judges routinely consider survey evidence to help them determine if a mark is confusing, generic, or has acquired secondary meaning.4 Dilution, however, appears to be the black sheep of the trademark survey family. The Federal Trademark Dilution Act of 1995 (“FTDA”)5 provides a holder of a famous trademark with a federal cause of action for “dilution,” defined as “the lessening of the capacity of a famous mark to identify and distinguish goods or services.”6 The FTDA, however, offers no guidance on how a trademark holder (a “senior user”) should prove that her trademark has been diluted by the mark of a defendant (a “junior user”). After the passage of the FTDA, courts immediately began to battle over whether a senior user could introduce a consumer survey as proof of dilu1. 233 F.3d 456, 468 (7th Cir. 2000). 2. Bette Davis Quotes, http://www.brainyquote.com/quotes/authors/b/ bette_davis.html (last visited Aug. 16, 2005). 3. One commentator asserts that “no other area of the law [other than trademark law] has relied to such a degree on the institutionalized use of statistical evidence.” Neal Miller, Facts, Expert Facts, and Statistics: Descriptive and Experimental Research Methods in Litigation, 40 RUTGERS L. REV. 101, 137 (1987). 4. See Schering Corp. v. Pfizer Inc., 189 F.3d 218, 225 (2d Cir. 1999) (“Surveys are, for example, routinely admitted in trademark . . . cases to show actual confusion, genericness of a name or secondary meaning, all of which depend on establishing that certain associations have been drawn in the public mind.”). See generally Gerald L. Ford, Lanham Act Surveys: 2002, 720 PLI/Pat 441 (2002) (discussing various trademark claims brought under the Lanham Act in 2002 where courts accepted consumer surveys as valid evidence). 5. 15 U.S.C. § 1125(c)(1) (2000). “The owner of a famous mark shall be entitled . . . to an injunction against another person’s commercial use in commerce of a mark or trade name, if such use begins after the mark has become famous and causes dilution of the distinctive quality of the mark. . . .” Id. 6. Id. § 1127. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 3 tion.7 The Supreme Court addressed this rising tension in Moseley v. V Secret Catalogue, Inc.,8 where the Court held that while the FTDA requires senior users to prove “actual dilution” above and beyond mere mental association between two marks, senior users may use consumer surveys as a method of proving actual dilution.9 Though the Court sympathized with the argument that “consumer surveys . . . are expensive and often unreliable,” it nonetheless asserted that these difficulties “are not an acceptable reason for dispensing with proof of an essential element of a statutory violation.”10 Despite the Supreme Court’s express approval of dilution surveys, not one post-Moseley court has yet accepted consumer survey data as proof of dilution under the FTDA.11 This uniform rejection is particularly disconcerting given that every federal circuit regularly accepts consumer survey evidence in other trademark contexts.12 While courts claim to reject dilution surveys because these surveys are expensive, unreliable, time-consuming, and subject to manipulation,13 such criticisms hold true for all trademark surveys.14 Courts have offered only one dilutionspecific reason for rejecting dilution surveys, a critique straightforwardly stated by Judge Richard Posner: “We are not sure what 7. Compare Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Div. of Travel Dev., 170 F.3d 449, 464–65 (4th Cir. 1999) (stating that the “[m]ost obviously relevant, and readily available [method of proving trademark dilution] is the skillfully constructed consumer survey”), with Eli Lilly & Co. v. Natural Answers, Inc., 233 F.3d 456, 468 (7th Cir. 2000) (doubting “that dilution of the distinctiveness of a mark is something that can be measured on an empirical basis by even the most carefully constructed survey”). 8. 537 U.S. 418 (2003). 9. Id. at 433–34. 10. Id. at 434. 11. See infra Part IV.B (discussing the few post-Moseley cases that directly address dilution surveys). 12. See Patrick M. Bible, Comment, Defining and Quantifying Dilution Under the Federal Trademark Dilution Act of 1995: Using Survey Evidence to Show Actual Dilution, 70 U. COLO. L. REV. 295, 317–18 n.112 (1998) (listing multiple examples from nearly every circuit where courts accepted properly conducted surveys as admissible evidence in other trademark contexts). 13. See Nabisco, Inc. v. PF Brands, Inc., 191 F.3d 208, 224 (2d Cir. 1999) (describing dilution surveys as “expensive, time-consuming and not immune to manipulation”); cf. Moseley, 537 U.S. at 434 (acknowledging arguments that dilution surveys might be “expensive and often unreliable”). 14. See Robert G. Bone, Enforcement Costs and Trademark Puzzles, 90 VA. L. REV. 2099, 2135–36 (2004) (describing trademark surveys generally as “expensive and difficult to design”). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 4 Columbia Journal of Law and Social Problems [39:1 question could be put to consumers that would elicit a meaningful answer [to prove actual dilution].”15 That is, courts question whether it is even possible to construct a survey question that measures whether a junior mark dilutes a senior mark. These critiques raise a fundamental question: is there a place for consumer surveys in federal trademark dilution claims? This Note argues that consumer surveys should still hold a significant place in proving dilution under the FTDA, particularly when the two marks in question are not identical, so long as a survey is properly constructed to meet Moseley’s requirement of proving actual dilution and not merely mental association between the two marks. This Note also rises to Judge Posner’s challenge of creating a survey question that can “elicit a meaningful answer” by proposing a method of constructing survey questions that directly measure the lessening of a senior mark’s “uniqueness,” the core injury of dilution. By employing these “unique value surveys,” which compare consumers’ ratings of a senior mark’s uniqueness in test and control populations, trademark holders will finally have the opportunity to prove through survey evidence that the mere presence of a dilutive junior mark weakens the senior mark’s uniqueness — a quality that companies spend extraordinary amounts of money, time, and effort to create and nurture, and that the FTDA was specifically designed to protect.16 Part II of this Note recounts the history of dilution law, culminating in the Supreme Court’s Moseley decision. Part III explores the advantages that consumer surveys provide over other methods of proof in FTDA cases, while Part IV examines why, despite these advantages, post-Moseley courts consistently reject dilution surveys. Finally, Part V proposes the use of unique value surveys as a valid method of measuring the lessening of a mark’s uniqueness, the cornerstone of dilution injury. 15. Ty Inc. v. Softbelly’s, Inc., 353 F.3d 528, 535 (7th Cir. 2003); see also Eli Lilly & Co. v. Natural Answers, Inc., 233 F.3d 456, 468 (7th Cir. 2000) (“[W]e doubt that dilution of the distinctiveness of a mark is something that can be measured on an empirical basis by even the most carefully constructed survey.”). 16. H.R. REP. NO. 104-374, at 2 (1995) (stating that the FTDA’s purpose is “to protect famous trademarks from subsequent uses that blur the distinctiveness of the mark . . . even in the absence of a likelihood of confusion”). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 5 II. THE EVOLUTION OF DILUTION LAW: FROM SCHECHTER TO MOSELEY Dilution law is not known for being particularly easy to understand. For instance, J. Thomas McCarthy recently lamented that no other concept in intellectual property law has created as much “doctrinal puzzlement and judicial incomprehension.”17 To avoid adding to this confusion, this Part first reviews the origin of dilution theory. This Part next describes the creation of the FTDA and the war that courts waged over defining “dilution,” and then discusses the Supreme Court’s attempt to settle this definitional controversy. A. THE GENESIS OF DILUTION: A NEW KIND OF TRADEMARK PROTECTION In the early days of trademark law, the central justification for recognizing trademark rights was to protect consumers from confusion or deception, so that consumers could feel confident that a product displaying a particular source’s mark truly originates from that source.18 Commonly known as “trademark infringement,” the creator of a trademark receives protection via an injunction when a court finds that another’s later use of a similar mark creates a “likelihood of confusion” which could plague con17. J. Thomas McCarthy’s frustration in explaining dilution is worth quoting in full: No part of trademark law that I have encountered in my forty years of teaching and practicing IP law has created so much doctrinal puzzlement and judicial incomprehension as the concept of “dilution” as a form of intrusion on a trademark. It is a daunting pedagogical challenge to explain even the basic theoretical concept of dilution to students, attorneys, and judges. I have tried mightily. I believe that few can successfully explain it without encountering blank stares of incredulity or worse, nods of understanding which mask and conceal puzzlement and misconceptions. J. Thomas McCarthy, Proving a Trademark Has Been Diluted: Theories or Facts?, 41 HOUS. L. REV. 713, 726 (2004). 18. See Hanover Star Milling Co. v. Metcalf, 240 U.S. 403, 412 (1916) (“The primary and proper function of a trade-mark is to identify the origin or ownership of the article to which it is affixed. Where a party has been in the habit of labeling his goods with a distinctive mark, so that purchasers recognize goods thus marked as being of his production, others are debarred from applying the same mark to goods of the same description . . . .”); Yale Electric Corp. v. Robertson, 26 F.2d 972, 973 (2d Cir. 1928) (“The law of unfair trade comes down very nearly to this — as judges have repeated again and again — that one merchant shall not divert customers from another by representing what he sells as emanating from the second.”). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 6 Columbia Journal of Law and Social Problems [39:1 sumers.19 This construction of trademark law is premised on the need to protect consumers, and not trademark holders.20 If a court finds that consumers are not likely to be confused between the sources of the two marks, then it will allow concurrent use, even if the marks are identical.21 Consumer confusion is all that matters. In the 1920s, an additional justification for trademark protection emerged. In a seminal article published in the Harvard Law Review, Frank I. Schechter suggested that “the use of trademarks on entirely non-related goods may of itself concretely injure the owner of the mark,” even in the absence of a likelihood of confusion.22 Under such circumstances, the injury to the trademark holder “is the gradual whittling away or dispersion of the [trademark’s] identity and hold upon the public mind of the mark or name by its use upon non-competing goods.”23 Viewing this “whittling away” as the real injury that trademark law must prevent, Schechter asserted that “the preservation of the uniqueness of a trademark should constitute the only rational basis for its protection.”24 Though Schechter did not use the word “dilution,”25 his words generally have been accepted as recognition that trademark protection can be used not only to aid consumers by helping them identify the source of a product, but also to protect producers’ in19. 15 U.S.C. § 1125(a)(1) (2000) (“Any person who, on or in connection with any goods or services, or any container for goods, uses in commerce any word, term, name, symbol, or device, or any combination thereof, . . . [which] is likely to cause confusion . . . shall be liable in a civil action by any person who believes that he or she is or is likely to be damaged by such act.”). 20. See Lynda J. Oswald, “Tarnishment” and “Blurring” Under the Federal Trademark Dilution Act of 1995, 36 AM. BUS. L.J. 255, 256 (1999) (“[T]he primary focus of traditional trademark law is protection of the consumer from deception, not protection of the value of the trademarks to their holders.”) (emphasis omitted). 21. See Melinda S. Giftos, Reinventing a Sensible View of Trademark Law in the Information Age, 2 CHI.-KENT J. INTELL. PROP. 2, ¶ 7 (2000) (“[T]rademark law allows concurrent use of the same mark, as long as the concurrent use does not confuse consumers. Thus, both Delta Airlines and Delta Faucets can use the same trademark without any problems.”) (footnote omitted). 22. Frank I. Schechter, The Rational Basis of Trademark Protection, 40 HARV. L. REV. 813, 825 (1927). 23. Id. 24. Id. at 831 (emphasis added). 25. JANE C. GINSBURG ET AL., TRADEMARK AND UNFAIR COMPETITION LAW 689 (3d ed. 2001) (“Schechter never actually referred to his theory as one protecting against ‘dilution.’”). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 7 vestments of money and effort in creating strong trademarks.26 Under this conception, a trademark holder suffers injury when a junior mark’s mere presence in the marketplace makes the senior mark appear less unique, and thereby diminishes its value.27 Common hypothetical examples of junior marks that could dilute include BUICK aspirin and KODAK pianos.28 Despite Schechter’s insistence that trademark holders need protection from dilution, statutory recognition of dilution did not immediately follow, though judges sometimes applied dilutionlike principles to common law unfair competition claims.29 Eventually, by the mid-1950s, “[r]eacting to concern over the inequities of having hard-earned goodwill whittled away by use of similar marks,”30 courts and legislatures began to develop laws to pro26. See Bible, supra note 12, at 297 (“Schechter’s primary concern was how a mark used by a vendor of a non-competing product could devalue an original mark’s selling power.”); Oswald, supra note 20, at 260. (“[Dilution] focuses not on the protection of the consumer from confusion . . . but upon protection of the trademark asset in the hands of the trademark holder.”). 27. See Oswald, supra note 20, at 261–62 (“Dilution theory . . . confers a property-like right upon the trademark holder to enjoin uses that, while not confusing to the consumer, might diminish the identification power of the mark in the marketplace over time.”); Ameritech, Inc. v. Am. Info. Technologies Corp., 811 F.2d 960, 965 (6th Cir. 1987) (“Dilution occurs when the senior user possesses a distinctive mark, the junior use of which . . . might cause a gradual diminution in the mark’s distinctiveness, effectiveness and, hence, value.”). J. Thomas McCarthy describes the “assumption” behind dilution as being that “the unique and distinctive significance of the mark to identify and distinguish one source might be diluted and weakened,” ultimately reducing its selling power. McCarthy, supra note 17, at 720. 28. Oswald, supra note 20, at 260 (describing these as “standard hypothetical examples” of dilution). Dilution theory posits that when a consumer encounters BUICK aspirin, she will no longer think the word BUICK is as unique as she once did. Therefore, Buick’s ability to identify and distinguish its cars in the marketplace has been lessened by the very presence of the junior mark. Additionally, the junior mark need not be identical to the senior mark — BUIK aspirin or CODAK pianos might be dilutive marks if their presence “whittles away” each senior mark’s “identity and hold upon the public mind.” Schecter, supra note 22, at 825. 29. See Yale Elec. Corp. v. Robertson, 26 F.2d 972, 974 (2d Cir. 1928) (where Judge Learned Hand wrote that the unfair use of another’s trademark “is an injury, even though the borrower does not tarnish it, or divert any sales by its use; for a reputation, like a face, is the symbol of its possessor and creator”). 30. W. Whitaker Rayner, In Search of a Dilution Solution: Implementation of the Federal Trademark Dilution Act, 20 MISS. C. L. REV. 93, 95 (1999). Trademark law now recognizes two sub-categories of dilution: blurring and tarnishment. The difference between blurring and tarnishment is that [d]ilution by “blurring” may occur where the defendant uses or modifies the plaintiff’s trademark to identify the defendant’s goods and services, raising the possibility that the mark will lose its ability to serve as a unique identifier of the File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 8 Columbia Journal of Law and Social Problems [39:1 tect a mark’s distinctive characteristics and associated goodwill. Many states enacted anti-dilution statutes.31 B. THE FEDERAL TRADEMARK DILUTION ACT OF 1995 In 1996, Congress passed the Federal Trademark Dilution Act of 1995, finally providing dilution protection at the federal level.32 The FTDA specifies that “[t]he owner of a famous mark shall be entitled . . . to an injunction against another person’s commercial use in commerce of a mark or trade name, if such use begins after the mark has become famous and causes dilution of the distinctive quality of the mark.”33 The FTDA defines “dilution” as “the lessening of the capacity of a famous mark to identify and distinguish goods or services, regardless of the presence or absence of (1) competition between the owner of the famous mark and the other parties, or (2) likelihood of confusion, mistake or deception.”34 plaintiff’s product. “Tarnishment” generally arises when the plaintiff’s trademark is linked to products of shoddy quality, or is portrayed in an unwholesome or unsavory context likely to evoke unflattering thoughts about the owner’s product. Deere & Co. v. MTD Prods., Inc., 41 F.3d 39, 43 (2d Cir. 1994) (footnotes omitted). This Note only addresses dilution by blurring. While it is relatively straightforward to construct survey questions that measure a respondent’s opinion of whether a junior mark creates a negative impression, measuring blurring requires a survey to gauge how “unique” a mark is, a more difficult proposition. For scholarship that discusses proving dilution by tarnishment, see generally Lee Goldman, Proving Dilution, 58 U. MIAMI L. REV. 569 (2004); Martha Kelley, Is Liability Just a Link Away? Trademark Dilution by Tarnishment Under the Federal Trademark Dilution Act of 1995 and Hyperlinks on the World Wide Web, 9 J. INTELL. PROP. L. 361 (2002). Additionally, the scope of Moseley is arguably limited to dilution by blurring. Moseley, 537 U.S. at 432 (“Whether [tarnishment] is actually embraced by the statutory text, however, is another matter.”). 31. 4 J. THOMAS MCCARTHY, MCCARTHY ON TRADEMARKS AND UNFAIR COMPETITION § 24:80 n.2 (4th ed. 2005) (“The states with an anti-dilution statute as of 2004: Alabama, Alaska, Arkansas, California, Connecticut, Delaware, Florida, Georgia, Hawaii, Idaho, Illinois, Iowa, Kansas, Louisiana, Maine, Massachusetts, Minnesota, Mississippi, Missouri, Montana, Nebraska, New Hampshire, New Jersey, New Mexico, New York, Oregon, Pennsylvania, Rhode Island, South Carolina, Tennessee, Texas, Washington, West Virginia, Wyoming.”). 32. 15 U.S.C. § 1125 (2000) (Historical and statutory notes state that the FTDA was enacted on January 16, 1996). 33. Id. § 1125(c)(1). 34. Id. § 1127. For example, the holder of the mark STARBUCKS for coffee might claim that the junior mark CHARBUCKS for charcoal “causes dilution.” While coffee and charcoal products are not in competition with each other, nor are consumers confused over the sources of the products, these facts would not preclude Starbucks’s claim. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 9 Courts generally understand these provisions of the FTDA as requiring five elements to prove dilution: (1) the senior mark must be famous; (2) the senior mark must be distinctive; (3) the junior mark must be used in commerce; (4) the junior mark’s use must begin after the senior mark is already famous; and (5) the junior mark must “cause” dilution of the senior mark’s “distinctive quality.”35 While commentators disagree over the standards of proof for many of these elements,36 some of the most contentious battles surround the fifth element: what does it mean to “cause” dilution? C. “LIKELIHOOD OF DILUTION” VS. “ACTUAL DILUTION” Immediately after the passage of the FTDA, federal courts disagreed over the standards for proving whether a junior mark “causes” dilution of a senior mark. Some courts concluded that the FTDA requires a senior user to prove actual dilution has already occurred, while others concluded that a senior user need only prove that a junior mark will likely cause dilution in the future. The Fourth and Second Circuits held the strongest opposing positions in this battle. 1. The Fourth Circuit: Textual Analysis Dictates Actual Dilution In Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Division of Travel Development,37 the Fourth Circuit looked to the words of the FTDA to determine the proper standard for causation. Relying on textual analysis,38 the court decided that 35. Nabisco, Inc. v. PF Brands, Inc., 191 F.3d 208, 215 (2d Cir. 1999). 36. For example, discussions of how to prove when a mark is “famous” are plentiful. See, e.g., Heidi L. Belongia, Why Is Fame Still Confusing? Misuse of the “Niche Market Theory” Under the Federal Trademark Dilution Act, 51 DEPAUL L. REV. 1159 (2002); Daniel E. Mangis, Note, When Almost Famous Just Isn’t Famous Enough: Understanding Fame in the Federal Trademark Dilution Act as a Term of Art Requiring Minimal Distinctiveness, 21 REV. LITIG. 455 (2002); Adam Omar Shanti, Comment, Measuring Fame: The Use of Empirical Evidence in Dilution Actions, 5 MARQ. INTELL. PROP. L. REV. 177 (2001); Christopher R. Perry, Note, Trademarks as Commodities: The “Famous” Roadblock to Applying Trademark Dilution Law in Cyberspace, 32 CONN. L. REV. 1127 (2000). 37. 170 F.3d 449 (4th Cir. 1999). 38. The court noted that while many state anti-dilution statutes explicitly allowed for a “likelihood of dilution” standard in their texts, the FTDA does not. “[G]iven the critical provisions that expressly differentiate the federal Act on key points from the state statutes, we must assume that this was exactly what was intended by Congress.” Id. at 459. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 10 Columbia Journal of Law and Social Problems [39:1 the FTDA “provides remedy only for actual, consummated dilution and not for the mere ‘likelihood of dilution’ proscribed by the state statutes.”39 To dilute, according to the court, a junior mark must “cause[ ] actual economic harm to the famous mark’s economic value” by weakening its selling power.40 While the Ringling Bros. court admitted that “[p]roof [of actual economic harm] will be difficult,”41 it suggested three methods for proving actual economic harm: (1) proof of actual loss of revenue;42 (2) “the skillfully constructed consumer survey designed not just to demonstrate ‘mental association’ of the marks in isolation, but further consumer impressions from which actual harm and cause might rationally be inferred”;43 and (3) “relevant contextual factors,” such as “similarity of the marks.”44 2. The Second Circuit: Common Sense Requires a Likelihood of Dilution The Second Circuit resoundingly rejected the Ringling Bros. court’s logic in Nabisco, Inc. v. PF Brands, Inc.,45 where the court asserted that “[t]o read the statute as suggested by the Ringling opinion would subject the senior user to uncompensable injury.”46 Because the court simply could not believe that Congress intended for a trademark holder to sit and wait for an injury to occur before FTDA relief was available, the court claimed that the Ringling Bros. reading of the FTDA “depend[ed] on excessive literalism to defeat the intent of the statute.”47 To prove its point through common sense, the Second Circuit explained that requiring proof of actual loss of revenue is unsatisfactory because “[i]f the famous senior mark were being exploited with continually growing success, the senior user might never be 39. Id. at 458. 40. Id. at 461. 41. Id. at 464. 42. Id. at 465. 43. Id. The court described this factor as “[m]ost obviously relevant, and readily available.” Id. 44. Id. 45. 191 F.3d 208 (2d Cir. 1999). 46. Id. at 224. 47. Id. The court further opined that “it seems plausibly within Congress’s meaning to understand the statute as intending to provide for an injunction to prevent the harm before it occurs.” Id. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 11 able to show diminished revenues, no matter how obvious it was that the junior use diluted the distinctiveness of the senior.”48 Further, the court deemed surveys to be “expensive, timeconsuming and not immune to manipulation.”49 Rather, the court had faith that “persuasive circumstantial evidence of dilution . . . without being obligated to show lost revenue or engage in an expensive battle of surveys” would be sufficient for proving dilution under the FTDA.50 After Ringling Bros. and Nabisco, other circuit courts then chose sides. The Fifth Circuit agreed with the Fourth Circuit, deciding that the plain text of the FTDA requires a finding of “actual dilution.”51 However, the Third,52 Sixth,53 and Seventh Circuits54 sided with the Second Circuit, concluding that the FTDA only requires the court to find a likelihood of dilution before providing relief to the senior user. 3. The Supreme Court: Three Methods for Proving Actual Dilution In 2003, in Moseley v. V Secret Catalogue, Inc.,55 the Supreme Court resolved this conflict, holding that the FTDA’s “text unambiguously requires a showing of actual dilution, rather than a likelihood of dilution.”56 Appearing to adopt the Fourth Circuit’s 48. Id. at 223–24. 49. Id. at 224. 50. Id. 51. See Westchester Media v. PRL USA Holdings, Inc., 214 F.3d 658, 670 (5th Cir. 2000) (endorsing the Fourth Circuit’s holding “that the FTDA requires proof of actual harm since this standard best accords with the plain meaning of the statute”). 52. See Times Mirror Magazines, Inc. v. Las Vegas Sports News L.L.C., 212 F.3d 157, 179 n. 11 (3d Cir. 2000) (“The majority also holds . . . that irreparable injury may be shown even in the absence of actual economic harm, presumably siding with the Second Circuit . . . .”) (Barry, J., dissenting). 53. See V Secret Catalogue, Inc. v. Moseley, 259 F.3d 464, 475 (6th Cir. 2001) (concluding that the Second Circuit’s interpretation “both tracks the language of the statute and follows more closely Congress’s intent in enacting the FTDA”), rev’d, 537 U.S. 418 (2003). 54. See Eli Lilly & Co. v. Natural Answers, Inc., 233 F.3d 456, 468 (7th Cir. 2000) (“We . . . side with the Second Circuit and hold that proof of a mere ‘likelihood of dilution’ is sufficient to satisfy the ‘causes dilution’ element . . . .”). 55. 537 U.S. 418 (2003). 56. Id. at 433. Although a bill has been proposed in response to Moseley that would amend the FTDA so that it only requires a likelihood of dilution and not “actual dilution,” this bill is still in preliminary stages. Additionally, its passage would not affect the core of the proposal made in this Note. See infra note 199. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 12 Columbia Journal of Law and Social Problems [39:1 basic line of analysis, the Court held that under the FTDA “‘the owner of a famous mark’ is entitled to injunctive relief against another person’s commercial use of a mark . . . if that use ‘causes dilution of the distinctive quality’ of the famous mark,”57 but not if the use merely “is likely to cause dilution.” The Court suggested three possible methods of proving actual dilution: (1) demonstrating the loss of sales or profits;58 (2) using consumer surveys; and (3) “reliably” proving actual dilution “through circumstantial evidence — the obvious case is one where the junior and senior marks are identical.”59 Finally, the Court held that “at least where the marks at issue are not identical, the mere fact that consumers mentally associate the junior user’s mark with a famous mark is not sufficient to establish actionable dilution.”60 While the Court acknowledged that requiring actual dilution might result in “difficulties of proof,” the Court nonetheless found actual dilution to be “an essential element of a statutory violation.”61 III. CONSUMER SURVEYS MAY PROVIDE THE BEST METHOD FOR PROVING ACTUAL DILUTION In the period following the Supreme Court’s Moseley ruling, judges and scholars wondered how a court could ever uphold a finding of dilution under the FTDA with such a difficult proof requirement.62 Subsequent case law has demonstrated, however, 57. Id. at 432–33 (emphasis in original) (quoting 15 U.S.C. § 1125(c)(1) (2000)). 58. While the Court allowed proof of actual dilution through loss of sales or profits, it expressly rejected making such proof a requirement, stating that requiring actual dilution “does not mean that the consequences of dilution, such as an actual loss of sales or profits, must also be proved. To the extent that language in the Fourth Circuit’s opinion in the Ringling Bros. case suggests otherwise, we disagree.” Id. (internal citations omitted). 59. Id. at 434. 60. Id. at 433. While the plain language of this sentence indicates that proof of a mere mental association between two marks will not be sufficient, some commentators suggest it is possible for mere mental association to constitute sufficient proof in certain circumstances. See Dale M. Cendali et al., Moseley v. V Secret Catalogue, Inc.: One Answer, Many Questions, 93 TRADEMARK REP. 833, 837 (2003) (“One might read this language as suggesting that the FTDA requires more than a mental association to establish actual dilution. It would be a mistake to do so. . . . Thus, the better reading of this language is . . . a showing of actual dilution requires evidence that when consumers see the senior mark, they also think of the junior mark.”). 61. Moseley, 537 U.S. at 434. 62. See Linda B. Samuels, Is Dilution a Delusion?, 86 J. PAT. & TRADEMARK OFF. SOC’Y 325, 334 (2004) (“[W]hile the Court seems unconcerned about any difficulties associ- File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 13 that proving dilution, while difficult, is possible. Analysis of the three possible methods for proving actual dilution suggests that consumer surveys may be the best way for some trademark holders to satisfy the actual dilution requirement, especially when the marks in question are not identical. This Part first discusses the weaknesses of using the loss of sales or profits and circumstantial evidence as methods of proving dilution. This Part next explains why, in many cases, consumer surveys offer a superior method of proving dilution, and concludes by responding to possible critiques of the power of dilution surveys. A. PROBLEMS WITH LOSS OF SALES OR PROFITS AS EVIDENCE OF DILUTION One point upon which scholars and judges seem to agree is that it is impracticable to prove actual dilution by showing that a senior user suffered a loss of sales or profits as a result of a junior mark’s presence. This is because it is virtually impossible to trace exactly what caused the loss of a company’s revenue — perhaps it was dilution, but perhaps it was a competitor’s entry into the market, a reduction in advertising, or a complex interaction of many factors.63 Such suspicions are confirmed by the fact that no ated with requiring proof of actual dilution, it defies logic to believe that Congress would create a cause of action but, at the same time, render it virtually impossible to establish relief.”); Brian A. Jacobs, Note, Trademark Dilution on the Constitutional Edge, 104 COLUM. L. REV. 161, 193 (2004) (labeling Moseley’s requirement of actual dilution as “empty”). Jacobs proposes a new test for measuring actual dilution based on economic efficiency. “Courts should go on to ask . . . whether the junior mark’s reference to the senior mark provides useful information. Based on this evaluation, courts should find dilution only in the presence of a net efficiency loss.” Id. While Jacobs’ proposal is of theoretical interest, this Note focuses on guiding practitioners who must work within Moseley’s requirements and do not have the luxury of waiting for courts to reject a test mandated by the Supreme Court. William G. Barber provides an extremely direct explanation of why it is so difficult to measure actual dilution. Unlike cases of consumer confusion, “a customer rarely calls a trademark owner and says, ‘I think the capacity of your mark to identify and distinguish your products is being lessened by Company B’s use of your mark.’ It does not happen very often in the real world.” William G. Barber, How to Do a Trademark Dilution Survey (Or Perhaps How Not to Do One), 89 TRADEMARK REP. 616, 616–17 (1999). 63. See Amy E. Pulliam, Note, Raising the Bar Too High: Moseley v. V Secret Catalogue, Inc. and Relief Under the Federal Trademark Dilution Act, 53 CATH. U.L. REV. 887, 906 (2004) (“[A]s pointed out [in Nabisco], it may be impossible to prove that any loss in profits was the result of a diluting, junior mark.”) (footnote omitted); Jeffrey Enright, Note, Slow Death of a Salesman: The Watering Down of Dilution Viability by Demanding Proof of Actual Economic Loss, 77 CHI.-KENT. L. REV. 937, 950 (2002) (“Even if a senior File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 14 Columbia Journal of Law and Social Problems [39:1 court has ruled in favor of a senior user under the FTDA based on the loss of sales or profits.64 Second, a junior mark’s evocation of a senior mark might actually help the senior user’s sales. The Ringling Bros. court explained this paradox, stating that “common sense suggests that an occasional replicating use might even enhance a senior mark’s ‘magnetism’ — by drawing renewed attention to it as a mark of unshakable eminence worthy of emulation by an unthreatening non-competitor.”65 Under this argument, no trademark holder could confidently depend on a loss of sales or profits to prove actual dilution.66 This is particularly true if a trademark holder believes her mark is being diluted while company profits are increasing, but not as strongly as they would in the absence of the junior mark’s dilutive influence.67 Finally, the central premise of dilution does not require the loss of sales or profits. While such a consequence might be possible if the dilution is powerful enough, the true injury is the loss of uniqueness, not the loss of sales or profits.68 Moseley confirms mark holder could show loss of revenue, it would be exceedingly difficult to produce evidence that the loss resulted from dilution and not some other factor.”); see also Nabisco, Inc. v. PF Brands, Inc., 191 F.3d 208, 223–24 (2d Cir. 1999): To require proof of actual loss of revenue seems inappropriate. If the famous senior mark were being exploited with continually growing success, the senior user might never be able to show diminished revenues, no matter how obvious it was that the junior use diluted the distinctiveness of the senior. Even if diminished revenue could be shown, it would be extraordinarily speculative and difficult to prove that the loss was due to the dilution of the mark. 64. This has led some economic-minded commentators to doubt whether dilution even exists. See Kenneth L. Port, The “Unnatural” Expansion of Trademark Rights: Is a Federal Dilution Statute Necessary?, 18 SETON HALL LEGIS. J. 433, 447–48 (1994): No mark has ever actually, quantitatively been established to have been diluted. If it had, we should be able to point to a specific trademark that was worth X before the entry of a newcomer but worth X minus Y after the entry of a noncompeting and diluting newcomer. Because there is none leads me to doubt even the existence of the idea of dilution. (footnote omitted). 65. Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Div. of Travel Dev., 170 F.3d 449, 460 (4th Cir. 1999). 66. Cf. Thomas R. Lee, Demystifying Dilution, 84 B.U. L. REV. 859, 876 (2004) (noting that requiring proof by lost revenue would “arbitrarily foreclose some valid dilution claims,” such as when the senior mark is continually eroded on an ongoing basis). 67. Enright, supra note 63, at 950 (“In instances where a mark is becoming famous across the country, uniqueness of the mark could be diluted even as sales are increasing.”). 68. Lee, supra note 66, at 897 (emphasizing that “[t]he essence of dilution is not this anticipated consequence [of loss of sales or profits]; it is the interference with a famous File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 15 this sentiment by specifically rejecting any requirement to prove economic harm to the senior mark.69 B. PROBLEMS WITH CIRCUMSTANTIAL EVIDENCE At first glance, it might appear that presenting circumstantial evidence would be the most straightforward way for a senior user to prove actual dilution. This method, however, is also the most subjective and, as such, unreliable. The Supreme Court provided only one cryptic example of acceptable circumstantial evidence of actual dilution: when the junior and senior marks are identical.70 Without further explanation, however, courts are unsure what properly constitutes “circumstantial evidence.”71 The critical question is: do identical marks automatically present sufficient “circumstantial evidence” on their own to establish actual dilution, or is it that only identical marks qualify to present additional circumstantial evidence?72 mark’s distinctive connection to a single source”). But see Jacobs, supra note 62, at 188 (“The uniqueness of trademarks is not an interest in itself. Rather, the preservation of this uniqueness must serve someone’s interest. The mark holder surely benefits from the FTDA’s preservation of her mark’s uniqueness, but consumers also benefit, as they experience a more efficient market.”) (footnote omitted). However, the lack of a requirement to prove the loss of sales and revenues suggests that from an evidentiary perspective, the uniqueness of the trademark is a self-standing interest above and beyond any potential economic interests. The element of proving a junior mark “caused” dilution does not require economic analysis, regardless of the theoretical implications of an injunction’s ultimate economic impact. 69. Moseley v. V Secret Catalogue, Inc., 537 U.S. 418, 433 (2003) (explicitly disagreeing with the Ringling Bros. opinion by holding that proof of actual dilution “does not mean that the consequences of dilution, such as an actual loss of sales or profits, must also be proved”). 70. Id. at 434. 71. See, e.g., Ty Inc. v. Softbelly’s, Inc., 353 F.3d 528, 536 (7th Cir. 2003) (“The Court did not explain and no one seems to know what that ‘circumstantial evidence’ might be.”); Savin Corp. v. Savin Group, No. 02-9377(SAS), 2003 U.S. Dist. LEXIS 19220, at *43 (S.D.N.Y. Oct. 24, 2003) (“This sentence is not easy to interpret, as is apparent from the differing interpretations of lower courts.”), rev’d, 391 F.3d 439 (2d Cir. 2004). 72. One court posed the question as such: “Is the Court saying . . . that when the junior and senior marks are identical, that in itself is sufficient circumstantial evidence to prove actual dilution? Or, is the Court saying that circumstantial evidence of actual dilution, as opposed to direct evidence, is sufficient when the marks are identical?” Savin, No. 02-9377(SAS), 2003 U.S. Dist. LEXIS 19220, at *43 (internal citations omitted), rev’d, 391 F.3d 439 (2d Cir. 2004). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 16 Columbia Journal of Law and Social Problems [39:1 Most courts73 and commentators74 agree with the former, advocating that the mere existence of identical marks constitutes sufficient circumstantial evidence of actual dilution. A minority of courts, however, suggest that the Supreme Court’s example functions only as an entry pass into FTDA consideration. That is, when the marks are identical, the senior user must then provide further circumstantial evidence which proves actual dilution.75 Relying on circumstantial evidence, however, offers virtually no recourse for situations where the marks at issue are not identical. As one scholar notes, “[s]o far, courts and commentators have tended to read Moseley to foreclose circumstantial evidence, at least in cases of ‘non-identical’ trademarks.”76 If this interpretation of Moseley is accurate, then proving dilution through circumstantial evidence is completely unavailable when senior and 73. See, e.g., Am. Honda Motor Co. v. Pro-Line Protoform, 325 F. Supp. 2d 1081, 1085 (C.D. Cal. 2004) (“[W]hen identical marks are used on similar goods, dilution . . . obviously occurs.”); 7-Eleven, Inc. v. McEvoy, 300 F. Supp. 2d 352, 357 (D. Md. 2004) (“Though dilution claims require evidence of actual confusion, that requirement is satisfied when, as here, the defendant uses the plaintiff’s mark.”); GMC v. Autovation Techs., Inc., 317 F. Supp. 2d 756, 764 (D. Mich. 2004) (“GM’s evidence establishes actual dilution in that Defendant has used marks that are identical to the world famous GM Trademarks.”); see also Stacey L. Dogan, An Exclusive Right to Evoke, 44 B.C. L. REV. 291, 315–16 (2003) (stating that a senior user may prove dilution when “a defendant’s use by its very nature reduces the singularity of the famous mark (as when the defendant uses an identical mark)”). 74. For instance, J. Thomas McCarthy believes that “when a famous mark is used by another in identical form, this is ‘circumstantial evidence’ and no further evidence should be necessary to reach the conclusion that the famous mark has in fact been ‘diluted’ to some extent, no matter how relatively minor is the junior use.” MCCARTHY, supra note 31, at § 24:94.2. 75. See Savin, No. 02-9377(SAS), 2003 U.S. Dist. LEXIS 19220, at *44 (“In the instant case, plaintiff offers no circumstantial evidence of any kind tending to show actual dilution other than the fact that the marks are identical. This is not sufficient.”), rev’d, 391 F.3d 439 (2d Cir. 2004); Lee Middleton Original Dolls, Inc. v. Seymour Mann, Inc., 299 F. Supp. 2d 892 (D. Wis. 2004) (same); Nike, Inc. v. Circle Group Internet, Inc., 318 F. Supp. 2d 688, 695 (N.D. Ill. 2004) (same). In December of 2004, the Second Circuit reversed the district court’s Savin ruling. Savin, 391 F.3d 439 (2d Cir. 2004). The court noted that the district court’s ruling seemed to be the only ruling that independently held such a position, with a few other courts citing Savin as their only support. Id. at 452. Nonetheless, other cases, such as Lee Middleton Original Dolls, are still good law, and even if they were not, the presence of this uncertainty exemplifies the unreliability of relying on the circumstantial evidence prong to prove actual dilution. 76. Lee, supra note 66, at 907. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 17 junior marks are similar, perhaps even incredibly similar, but nonetheless non-identical.77 Additionally, the lack of consensus over the meaning of “circumstantial evidence” provides judges with vast discretion, allowing for wildly unpredictable results and creating an incentive for forum shopping. Until the Supreme Court clarifies what it considers to be proper “circumstantial evidence” (beyond one foggy example) or until district courts can agree on one clear definition, uncertainty will loom over this method of proof. As a result, senior users cannot necessarily rely on the “circumstantial evidence” method for proving actual dilution, even when the marks at issue are identical.78 C. PROPERLY CONSTRUCTED CONSUMER SURVEYS MAY BEST MEET THE “ACTUAL DILUTION” PROOF REQUIREMENT Consumer surveys have received a bad reputation in dilution cases. While courts freely rely on consumer survey evidence in other trademark contexts,79 the very same courts are less generous in accepting surveys as proof of actual dilution. As noted above, Judge Posner recently suggested that consumer surveys have little value in a dilution context,80 and other judges and commentators seem to agree with this critique.81 77. Courts are trending toward requiring absolute identicalness. See Savin, 391 F.3d at 453 (holding that “a mere similarity in the marks — even a close similarity — will not suffice to establish per se evidence of actual dilution”); Corbond Corp. v. Core Foam, Inc., 356 F. Supp. 2d 910, 920 (W.D. Wis. 2005) (“Proof of actionable dilution with circumstantial evidence is appropriate when the marks are identical, it does not seem to be sufficient where the marks are merely similar.”). 78. Until the Supreme Court clearly states otherwise, there is always a possibility that a court could arrive at the same conclusion as the district courts did in Savin and Lee Middleton Original Dolls. 79. See Bible, supra note 12, at 317 n.112 (where the author lists multiple examples from every circuit where courts accepted surveys as admissible evidence in other trademark contexts); see also Schering Corp. v. Pfizer Inc., 189 F.3d 218, 225 (2d Cir. 1999) (“Surveys are, for example, routinely admitted in trademark . . . cases to show actual confusion, genericness of a name or secondary meaning . . . .”). 80. Ty Inc. v. Softbelly’s, Inc., 353 F.3d 528, 535 (7th Cir. 2003) (where Judge Posner stated that he was “not sure what question could be put to consumers that would elicit a meaningful answer . . . .”). 81. See, e.g., Eli Lilly & Co. v. Natural Answers, Inc., 233 F.3d 456, 468 (7th Cir. 2000) (“[W]e doubt that dilution of the distinctiveness of a mark is something that can be measured on an empirical basis by even the most carefully constructed survey.”); Jonathan E. Moskin, Victoria’s Big Secret: Whither Dilution Under the Federal Dilution Act?, File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 18 Columbia Journal of Law and Social Problems [39:1 Such a pessimistic view of dilution surveys is unwarranted for several reasons. First, the complaint that dilution surveys are “expensive, time-consuming and not immune to manipulation”82 applies to any trademark survey. Every federal jurisdiction accepts consumer surveys as valid evidence in the context of trademark infringement under the Lanham Act, routinely finding that surveys provide proof of a likelihood of consumer confusion.83 In some trademark contexts, surveys are not only welcomed, but expected.84 Further, courts have not had problems rejecting consumer surveys based on dubious methodologies in trademark infringement cases.85 Even Judge Posner admitted in a trademark infringement case that, despite potential pitfalls in survey methodology, courts will still recognize consumer survey data as sufficiently reliable.86 While dilution surveys may be expensive and difficult to design, that does not make them inherently unreliable from an evidentiary perspective. Rather, the Federal Rules of Evidence clearly authorize the admission of dilution surveys as valid evidence when appropriate.87 The Solicitor General also 93 TRADEMARK REP. 842, 853 (2003) (“In the wake of Moseley, it remains to be seen if some other effective (or at least acceptable) methods can be designed of surveying actual effects on consumers caused by any given accused mark.”) (footnote omitted). 82. Nabisco, Inc. v. PF Brands, Inc., 191 F.3d 208, 224 (2d Cir. 1999). For further discussion responding to general critiques that survey research is expensive, unreliable, open to interpretation, and unable to produce useful results, see Matthew D. Bunker et al., Proving Dilution: Survey Evidence in Trademark Dilution Actions, 13 U. BALT. INTELL. PROP. L.J. 37, 49–52 (2004). 83. See supra note 79 and accompanying text. 84. Vincent N. Palladino, Surviving Secondary Meaning, 84 TRADEMARK REP. 155, 156 (1994) (explaining that “a party . . . may fairly be criticized for not introducing a survey when the issue is genericness”) (footnote omitted). 85. See, e.g., NFL Props., Inc. v. Prostyle, Inc., 57 F. Supp. 2d 665, 668 (D. Wis., 1999) (rejecting a survey because it “asks only one question . . . without further probing”); Am. Footwear Corp. v. Gen. Footwear Co., 609 F.2d 655, 660 (2d Cir. 1979) (rejecting a survey because of its “methodological defects”). 86. There, Judge Posner wrote: Many experts are willing for a generous (and sometimes for a modest) fee to bend their science in the direction from which their fee is coming. . . . The judicial constraints on tendentious expert testimony are inherently weak because judges . . . lack training or experience in the relevant fields of expert knowledge. But that is the system we have. It might be improved by asking each party’s hired expert to designate a third, a neutral expert who would be appointed by the court to conduct the necessary studies. . . . But with all this granted, we cannot say that the district judge committed a clear error. Indianapolis Colts, Inc. v. Metro. Balt. Football Club Ltd. Partnership, 34 F.3d 410, 415– 16 (7th Cir. 1994). 87. As Bible wrote: File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 19 supported such a position in his amicus curiae brief filed with the Supreme Court in Moseley.88 Therefore, dilution surveys offer the benefit of allowing judges to draw upon a large body of knowledge already relied upon by courts in other trademark survey contexts. Second, dilution surveys may offer valid data, if properly constructed to follow Moseley’s guidelines. Most criticisms of dilution surveys are based on flaws in survey construction or methodology.89 Many of these critiques correctly recognize that most “dilution” surveys are actually “consumer confusion” surveys, only measuring a mental association between two marks.90 However, as Moseley confirmed, “the mere fact that consumers mentally associate the junior user’s mark with a famous mark is not sufficient to establish actionable dilution.”91 Therefore, a valid dilution survey must prove not only that consumers mentally associate the junior and senior marks,92 but also that the junior mark “actually dilutes” the senior mark.93 While it is admittedly more Survey results are generally admitted after proper authentication of the testifying expert under Federal Rule of Evidence 901(b)(9), which requires a showing of the process by which the survey was conducted and proof that the process ‘produces an accurate result.’ In response to hearsay objections, courts universally have found that survey results are admissible as evidence of the state of mind of the person surveyed or are sufficiently reliable to fit within the catch-all exception to the hearsay rule. Bible, supra note 12, at 315–16 (footnotes omitted). 88. The Solicitor General’s brief argued: Courts must be cautious in evaluating the probative value of consumer surveys, just as they must be cautious in evaluating the probative value of many other kinds of evidence. But that is not a basis for categorically ruling out such evidence as a method for proving dilution. A properly designed and executed survey can have significant weight in a dilution case, just as it can have significant weight in a trademark infringement case. Brief for the United States as Amicus Curiae supporting Petitioners in part at 23–24, Moseley v. V Secret Catalogue, Inc., 537 U.S. 418 (2003) (No. 01-1015). 89. For several examples of this criticism, see cases discussed infra Part IV.A. 90. See, e.g., Gateway, Inc. v. Companion Prods., No. 01-4096-KES, 2003 U.S. Dist. LEXIS 21461, at *64 (D.S.D. Aug. 19, 2003) (“Although the consumer survey demonstrates that a significant portion of the public associates Cody Cow with Gateway, it does not demonstrate actionable dilution.”), aff’d, 384 F.3d 503 (8th Cir. 2004). 91. 537 U.S. at 433. 92. Mental association is a required precondition for proving actual dilution; after all, for dilution to occur, the two marks at issue must somehow be mentally connected. Actual dilution goes to the next level, where the marks are both mentally associated and the mere presence of one lessens the other’s capacity to identify and distinguish its goods or services. 93. See Bible, supra note 12, at 327–28 (“An effective survey, then, must establish not only that consumers associate the mark with both parties, but also that some quanta of the original mark’s identifying ability or selling power has been diminished.”); see also id. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 20 Columbia Journal of Law and Social Problems [39:1 difficult to construct a survey question that addresses dilution than one that addresses mental association, if a senior user can construct such a question, the survey’s construction must be deemed valid (assuming the survey is otherwise sound).94 Ultimately, senior users that construct valid dilution surveys will have more control over proving dilution than if they rely on a judge’s subjectivity in interpreting the scope of “circumstantial evidence.”95 Third, and most important, consumer surveys offer the only viable method for trademark holders to prove actual dilution in situations where senior and junior marks are not identical. While most courts do recognize identical marks as circumstantial evidence of dilution, these courts also require the marks to be identical.96 This limitation means that without the acceptance of a valid dilution survey, senior users who claim their marks are being diluted by non-identical (but extremely similar) junior marks can only prove actual dilution through the nearly impossible method of showing a loss of sales or profits.97 One commentator goes as far as stating that for non-identical marks, the need for survey evidence is “implicit.”98 These senior users should not be abandoned, especially when the text of the FTDA does not ex- at 314 (“[A] survey that clearly demonstrates actual dilution may be a plaintiff’s strongest and most persuasive evidence.”). 94. See infra Part V.B for recommendations on how to construct valid survey questions; see also infra note 170 and accompanying text, discussing various additional elements of valid surveys. 95. See Lee, supra note 66, at 921 (“Other forms of more clearly circumstantial evidence . . . require subjective judgments . . . . In this sense, survey evidence is more directly probative of dilution than are other forms of evidence . . . .”) (footnote omitted). 96. See supra note 77 and accompanying text. 97. Proving actual dilution through the loss of sales or profits would be the only available method because of the near impossibility of proof through circumstantial evidence. See J. Jarrod Thrash, Comment, Victoria’s Secret Is Not Safe with the Supreme Court: The Court Makes Its Foray into the Make-Believe World of the FTDA, 6 TUL. J. TECH. & INTELL. PROP. 199, 219 (2004) (asserting that for “cases where the marks are not exactly identical, but are similar, . . . [t]he Court’s conclusions would seem to indicate that . . . plaintiffs will face a much higher, if not impossible, burden of proving actual dilution with circumstantial evidence”). 98. Alan S. Cooper, Using and Excluding Surveys, Survey Experts, and Other Experts, SJ055 ALI-ABA 59, 63 (2004) (“[T]he Supreme Court in Moseley recognized that survey evidence is a means of demonstrating actual dilution . . . . Implicit in this conclusion is the Court’s suggestion that direct evidence in the form of a survey is a principal means of proving actual dilution where the marks are not identical.”). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 21 pressly exclude them from the universe of protected parties.99 In the absence of identical marks, these trademark holders will not have access to recourse under the FTDA, even when an injury is obvious and destructive, unless courts accept valid dilution surveys. D. COUNTERARGUMENTS AND RESPONSES Despite the significant benefits that dilution surveys may confer, especially when non-identical marks are at issue, not one post-Moseley federal court has accepted a consumer survey as constituting valid proof of actual dilution. Why are courts so unwilling to accept dilution surveys as proof of actual dilution? Most critiques fall into three categories: (1) concerns that apply to all trademark surveys; (2) concerns specific to judicial competence in gauging dilution under the FTDA; and (3) concerns specific to dilution survey construction. First, as noted above, many of the assaults on dilution surveys involve charges that such surveys are expensive, time-consuming, and subject to manipulability.100 Further, the raw act of conducting a survey involves a false reality which may not resemble how consumers actually make purchasing decisions.101 Of course, such criticisms also apply to all trademark surveys, such as infringement surveys, which courts universally accept. Second, courts may feel unprepared to gauge the reliability of dilution surveys because the FTDA is a relatively new creation.102 As such, perhaps federal courts feel unqualified to rule on dilution surveys, despite their relative comfort with infringement surveys.103 If such reticence is the true culprit for refusing to ac99. The FTDA does not require that the marks be identical — the junior mark only must “cause[ ] dilution.” 15 U.S.C. § 1125(c)(1) (2000). 100. Nabisco, Inc. v. PF Brands, Inc., 191 F.3d 208, 224 (2d Cir. 1999). See also Bone, supra note 14, at 2135–36 (stating that “[s]urveys are useful for [proving dilution], but they are expensive and difficult to design”). 101. Goldman, supra note 30, at 591. (“A survey, by inviting a respondent to think carefully about a mark, also may underestimate the amount of blurring that actually occurs in the marketplace, where consumers often are not quite so thoughtful.”). 102. While federal courts have extensive familiarity with Lanham Act claims of trademark infringement, the same courts did not encounter federal dilution claims until 1996, when Congress passed the FTDA. 103. Compare Ty Inc. v. Softbelly’s, Inc., 353 F.3d 528, 535 (7th Cir. 2003) (where the court could not conceive of a question on a dilution survey that would “elicit a meaningful File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 22 Columbia Journal of Law and Social Problems [39:1 cept dilution surveys, then this wariness will dissipate when a venturesome court realizes that dilution surveys are fundamentally no different than other trademark surveys. The prospect, however, of waiting until courts become more comfortable with the FTDA does little for trademark holders whose marks face dilution today. Third, federal courts may feel sufficiently competent to consider dilution surveys, despite their relative novelty, but nonetheless remain unconvinced that it is possible to construct a meaningful survey question which has the capability to prove actual dilution.104 That is, as Howard J. Shire and Michelle Mancino Marsh suggest, “perhaps dilution is such an elusive concept that it cannot be empirically tested in an accurate way.”105 A proper response to this criticism is to construct this “impossible” survey question and measure whether a sufficient percentage of respondents reveal that the presence of a junior mark lessens a senior mark’s capacity to distinguish and identify its products. Such a question must go beyond proving that a respondent mentally associates two marks — it must also prove that the junior mark directly lessened the senior mark’s “uniqueness.”106 Few trademark holders have ever attempted to construct such survey questions, most likely because many courts previously assumed that merely showing a mental association between two marks may constitute sufficient proof of actual dilution.107 Moseley, however, raises the bar — a new approach is needed. IV. CRITICISMS OF DILUTION SURVEYS MADE BY FEDERAL COURTS Before Moseley held that the FTDA requires proof of actual dilution, courts struggled to discern the evidentiary requirements answer”), with id. at 531 (where, in the very same case, the court analyzed the validity of a genericness survey, ruling that “the survey was worthless — 13- to 18-year-old girls being an arbitrary subset of consumers of beanbag stuffed animals”). 104. See id. at 535 (“We are not sure what question could be put to consumers that would elicit a meaningful answer either in that case [Moseley] or this one.”). 105. Howard J. Shire & Michelle Mancino Marsh, Federal Dilution Claims After Moseley v. V Secret Catalogue, 94 TRADEMARK REP. 1066, 1084 (2004). 106. See infra Part V.B for recommendations on how to construct such questions. 107. See infra note 108 and accompanying text (presenting cases where courts recognized mere mental association as valid proof of actual dilution). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 23 of the FTDA, providing inconsistent opinions on dilution surveys. After Moseley, courts’ approaches to survey evidence embraced a more uniform position, rejecting trademark holders’ attempts to prove actual dilution when surveys only displayed mental associations between two marks. To convincingly prove dilution through survey evidence, however, senior users must understand why previous attempts at measuring “actual dilution” failed. This Part explores the reasons courts provide for rejecting dilution surveys. A. PRE-MOSELEY DILUTION SURVEYS In the early days of the FTDA, some courts were willing to consider survey evidence in dilution claims. In several cases, courts accepted surveys as proof of actual dilution, even though the surveys only measured mental association.108 Other courts, while not expressly rejecting that mere mental association could be sufficient to prove actual dilution, rejected dilution surveys on various other grounds.109 This inconsistency stemmed from the fact that “[t]here [was] no standard criteria for surveying for dilution.”110 Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Division of Travel Development111 tried to put an end to this 108. See, e.g., Wawa Inc. v. Haaf, No. 96-4313, 1996 U.S. Dist. LEXIS 11494, at *7–8 (E.D. Pa. Aug. 8, 1996) (mental association by 29% of respondents between HAHA 24 HR. markets and WAWA convenience stores constituted evidence of dilution), aff’d, 116 F.3d 471 (3d Cir. 1997); Kellogg Co. v. Exxon Corp., 192 F. Supp. 2d 790, 806 (W.D. Tenn. 2001) (mental association by 70% of respondents between Kellogg’s mascot Tony the Tiger and Exxon’s cartoon tiger mascot “would suffice as enough proof of dilution to preclude summary judgment”); Hershey Foods Corp. v. Mars, Inc. 998 F. Supp. 500 (M.D. Pa. 1998) (denying Hershey’s claim of trade dress dilution based an improper survey, but suggesting that the mental association of 51% of respondents between the packaging of Reese’s Pieces and Peanut Butter M&M’s could be sufficient for dilution). 109. See, e.g., Pharmacia Corp. v. Alcon Laboratories, Inc., 201 F. Supp. 2d 335, 381 (D.N.J. 2002) (holding that, in addition to multiple deficiencies in the survey, 14% was too low a finding of mental association between two marks, being “less than half of the lowest percentage of dilution in any reported decision basing a finding of dilution on a survey . . . [and that] [o]ther courts have demanded a much higher percentage [i.e., 70%]”). The first percentage refers to Wawa Inc., v. Haaf and the second percentage refers to Kellogg Co. v. Exxon Corp. See also National Football League Properties, Inc. v. Prostyle, Inc., 57 F. Supp. 2d 665, 668 (E.D. Wis. 1999) (mental association by 50% of respondents could not be respected because the survey did not use a control group). 110. Hershey Foods, 988 F. Supp. at 518. 111. 170 F.3d 449 (4th Cir. 1999). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 24 Columbia Journal of Law and Social Problems [39:1 confusion by providing clear criteria for surveying dilution. By offering a nuanced analysis of the difference between mere mental association and actual dilution, Ringling Bros. represents a turning point in courts’ conceptions of what surveys must accomplish to prove actual dilution.112 In this case, Ringling Bros., owner of the mark THE GREATEST SHOW ON EARTH, claimed that the Utah Division of Travel’s use of THE GREATEST SNOW ON EARTH in connection with its Utah tourism campaign diluted the circus’s mark.113 To prove actual dilution, Ringling Bros. conducted a survey of populations both inside and outside Utah, where respondents were asked to fill in the blank in the following statement: “THE GREATEST _____ ON EARTH.”114 Outside Utah, 41% responded only “show,” 0% responded only “snow,” and less than 0.5% responded both “show” and “snow.”115 Inside Utah, 25% responded only “show,” 24% responded only “snow,” and 21% responded both “show” and “snow.”116 Ringling Bros. asserted that these results proved actual dilution. The court disagreed, affirming the lower court’s finding that all respondents, both inside and outside of Utah, were able to hold both slogans in their minds without allowing one to bleed into the other, suggesting an absence of dilution.117 The court believed that, if anything, Utah’s slogan might increase awareness of the Ringling Bros. slogan.118 The court concluded that “the survey evidence does not show that use of Utah’s junior mark had caused any actual harm to Ringling’s [senior] mark in the form of a lessening of that mark’s former capacity to identify and distinguish Ringling’s circus as its subject.”119 This holding is central to dilution survey analysis because Ringling Bros. laid down the basic rule that Moseley later adopted: “If you seek to rely for proof of dilution only upon evi112. William G. Barber refers to the Ringling Bros. analysis as providing “the bestreasoned and most instructive guidance to date on how to properly design a dilution survey.” Barber, supra note 62, at 624. 113. Ringling Bros., 170 F.3d at 451–52. 114. Id. at 462. 115. Id. 116. Id. 117. Id. 118. Id. at 463 (finding that the survey indicates that “consumer familiarity with Ringling’s mark was greater in Utah (46%), where Utah’s mark was well-known, than in the rest of the country (41%), where Utah’s mark was virtually unknown”). 119. Id. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 25 dence of the mental impressions . . . then those impressions must go beyond mere recognition of a visual similarity of the two marks to allow a reasonable inference that the junior mark’s use has caused actual harm to the senior mark’s selling or advertising power.”120 The Ringling Bros. survey provides the backbone for post-Moseley dilution survey analysis. B. DILUTION SURVEYS AFTER MOSELEY Relatively few courts have directly addressed the validity of dilution surveys since the Supreme Court handed down its Moseley decision.121 Reasons for why so few dilution survey cases have arisen might include the following: (1) senior users, discouraged by courts’ negative comments about dilution surveys, believe that courts will not accept even the most properly constructed survey; (2) senior users are unsure how to create reliable dilution surveys under Moseley’s heightened requirements and are cautiously waiting for further guidance from other court decisions; or (3) senior users are willing to construct dilution surveys, but have not yet had the time to conduct new studies. The few postMoseley opinions available, however, have uniformly rejected dilution surveys. These cases are significant not only because they provide critical insight into what courts now expect from dilution surveys, but also because each case involves parties whose marks are similar but not identical — the parties in most need of dilution surveys. These cases showcase the evolving standards courts developed in assessing trademark dilution surveys. Immediately after the Supreme Court’s Moseley ruling, the Sixth Circuit faced the case Kellogg Co. v. Toucan Golf, Inc.122 Kellogg contended that the junior mark, an image of a lifelike 120. Id. 121. As of September 2005, over fifty federal court published opinions cite Moseley; however, the cases cited in this section are the only cases which directly address dilution surveys. For summaries of post-Moseley dilution cases, see generally Shire & Marsh, supra note 105 (providing a list and brief discussion of federal dilution cases by circuit). One ruling did address an infringement survey used in a dilution claim, but the court curtly dismissed the claim, holding that “[t]here is no evidence that Cody Cow, in actuality, lessens the strength of Gateway’s trademark. No evidence demonstrates that Cody Cow undermines the ability to identify and distinguish Gateway’s products.” Gateway, Inc. v. Companion Prods., No. 01-4096-KES, 2003 U.S. Dist. LEXIS 21461, at *64 (D.S.D. Aug. 19, 2003), aff’d, 384 F.3d 503 (8th Cir. 2004). 122. 337 F.3d 616 (6th Cir. 2003). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 26 Columbia Journal of Law and Social Problems [39:1 toucan which appeared on golf-related products, diluted Kellogg’s trademark of Toucan Sam, a cartoon toucan associated with Froot Loops cereal. Kellogg supported its claim with consumer surveys, which fell into the trap of only testing the mental association between the two marks. As a result, the court rejected Kellogg’s claim, stating that even though Kellogg’s surveys proved that Toucan Sam was highly recognizable, “Kellogg has failed to present evidence that any segment of the population recognizes Toucan Sam as the spokesbird only for Froot Loops in lesser numbers than it did before TGI started using its toucan marks.”123 Because there was no proof of a loss of uniqueness of the Toucan Sam mark, Kellogg’s FTDA claim failed. In 2004, a Second Circuit district court reached a similar conclusion in Starbucks Corp. v. Wolfe’s Borough Coffee, Inc.124 In this case, the district court rejected Starbucks’s request for summary judgment based on the claim that Wolfe’s Borough’s use of the term CHARBUCKS in association with one its own coffee blends diluted the STARBUCKS mark. As evidence, Starbucks provided survey results showing that “31% of the consumers surveyed associated the term ‘Charbucks’ with Starbucks, and that 43% of consumers surveyed would have a negative impression of a coffee called ‘Charbucks.’”125 In spite of this association, the court found no “direct evidence” that the presence of the CHARBUCKS mark in the market weakened the STARBUCKS mark’s ability to function as a unique identifier of Starbucks’s goods and services.126 While the Starbucks survey comes closer to providing proof of actual dilution than did the surveys offered in Kellogg (because Starbucks presented consumer survey evidence that went beyond mere mental association) the court still held 123. Id. at 628. According to Kellogg’s surveys, 94% of children and adults recognized Toucan Sam. Id. 124. No. 01-5981(LTS)(THK), 2004 U.S. Dist. LEXIS 19239 (S.D.N.Y. Sept. 28, 2004). Starbucks filed several additional claims as well, including trademark infringement, unfair competition, and false advertising. Id. at *2. 125. Id. at *26. 126. Id. at *26–27 (“There is no direct evidence . . . of diminution of Plaintiffs’ mark as a unique identifier of their goods and services . . . by reason of the presence of Defendant’s coffee in the market.”). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 27 that the evidence was not direct enough to support summary judgment.127 Two months later, in Playtex Products v. Georgia-Pacific Corp.,128 the Second Circuit rejected Playtex’s claim that GeorgiaPacific’s use of the label QUILTED NORTHERN MOIST-ONES on the packaging of pre-moistened baby wipes diluted its WET ONES mark on a virtually identical product. Unlike the lower court, which dismissed Playtex’s dilution claim based on a lack of sufficient similarity between the two marks,129 the Second Circuit addressed Playtex’s dilution survey evidence. The court found two major flaws in the survey’s construction: (1) the survey “did not use any Georgia-Pacific packaging, but rather used index cards with the words ‘Moist-Ones’ and ‘Wet Ones’ placed out of context”;130 and (2) “the index cards did not contain the words ‘Quilted Northern.’”131 Because of these deficiencies, the court concluded that “[a]lthough evidence in the record may establish that consumers associate ‘Moist-Ones’ with ‘Wet Ones,’ association is not sufficient to establish a dilution claim.”132 The critical lesson of Playtex is that a court will be more likely to accept a dilution survey when the stimulus closely approximates what consumers actually encounter in the market.133 Additionally, the stimulus should present the entire mark, not just a portion. Malletier v. Dooney & Bourke, Inc.,134 however, offers the most comprehensive insight into what a court desires from a dilution study, as it is the only published post-Moseley decision which discusses deficiencies in surveys specifically designed to measure dilution, and not just mental association. In this case, Malletier 127. Id. at *27. However, the court did allow for a finder of fact to determine the “weight and significance of the survey,” leaving open the possibility that a trial court could find actual dilution. Id. 128. 390 F.3d 158 (2d Cir. 2004). 129. Playtex Prods. v. Georgia-Pacific Corp., No. 02-7848(HB), 2003 U.S. Dist. LEXIS 13981, at *26–27 (S.D.N.Y. Aug. 12, 2003). 130. Playtex Prods., 390 F.3d at 168. 131. Id. 132. Id. 133. Commentators reaffirm the importance of replicating market conditions in dilution surveys, suggesting that surveys should “not vary or obscure the second comer’s goods or services. Rather, specimens that actually depict the second comer’s products or services are likely to be viewed as more credible by a fact finder.” Steven B. Pokotilow & Stephen A. Fefferman, FTDA Survey Evidence: Does Existing Case Law Provide Any Guidance for Constructing a Survey?, 91 TRADEMARK REP. 1150, 1168 (2001). 134. 340 F. Supp. 2d 415 (S.D.N.Y. 2004). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 28 Columbia Journal of Law and Social Problems [39:1 (“Louis Vuitton”) claimed that the Dooney & Bourke’s It-Bags, featuring a “DB” monogram in a multicolored array, diluted the trade dress of Louis Vuitton’s recently unveiled Multicolore bags, featuring the “LV” monogram and various geometric shapes in a multicolored array.135 While the court did not dispute that the “Monogram Multicolore trademarks are inherently distinctive and famous,”136 the court ultimately ruled against Louis Vuitton on the question of “whether [Louis Vuitton] has adequately demonstrated that it is likely to be able to prove actual dilution.”137 The court described Louis Vuitton’s dilution survey goals and results as follows: The objective of Dr. Jacoby’s dilution study is to test Louis Vuitton’s belief that Dooney & Bourke’s “use of a monogram with multiple bright colors in a repeating pattern similar to the Louis Vuitton Multicolore design would cause dilution,” defined as “either a lessening of the purchase intentions toward and/or perceived distinctiveness, exclusivity, value and desirability of the Louis Vuitton Multicolore bags, or an increase in purchase intentions toward the Dooney It-Bags.” . . . Dr. Jacoby found that Dooney & Bourke’s It-Bags “cause dilution by blurring at [twenty-three] percent and that the Dooney lookalikes had a tendency to decrease the distinctiveness, value and exclusivity of the Monogram Multicolore.”138 Dr. Jacoby concluded that many respondents, after viewing the It-Bags, would be less likely to purchase or desire the Louis Vuitton Multicolore bag, an injury he attributed to dilution. After considering the survey’s construction and methodology, the court rejected Dr. Jacoby’s study as proof of actual dilution139 for three reasons: (1) the screening questions were overinclu- 135. Id. at 448. Louis Vuitton’s anger stemmed from the fact that while their bags sold for up to $3,950 each, the Dooney & Bourke bags only cost between $125 and $350 each. Rob Walker, The Acceptable Knockoff, N.Y. TIMES, Dec. 12, 2004, § 6 (Magazine), at 46. 136. Malletier, 340 F. Supp. 2d at 448. 137. Id. 138. Id. at 449–50 (footnotes omitted, brackets in the original). 139. Id. at 450. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 29 sive;140 (2) the study was not objective because Dr. Jacoby restarted the study after it was halfway done, raising questions of validity;141 and (3) the scope of “dilution” wrongly expanded “beyond ‘distinctiveness,’ ‘exclusivity,’ and ‘value,’ by asking questions relating to desirability,” because “desirability is unrelated to a mark’s ability to identify and distinguish Louis Vuitton’s handbags.”142 Therefore, the court concluded, “the Jacoby Dilution Study lends little, if any, support to the assertion that the Dooney & Bourke It multicolored bags actually dilute the Monogram Multicolore trademarks.”143 Malletier provides trademark holders the clearest indication of what courts now demand from dilution surveys. For a survey question to measure actual dilution, it cannot measure “distinctiveness,” “exclusivity,” or “value” by using “desirability” as a proxy.144 While the Malletier court does not provide examples of questions that would be acceptable, the clear implication is that survey constructors should be careful to ask questions addressing distinctiveness only, and not desirability. These four cases — Kellogg, Starbucks, Playtex, and Malletier — provide a solid foundation for thinking about how future courts will approach dilution surveys in a post-Moseley world. Surveys which merely prove mental association between two marks will not be sufficient. Rather, a proper dilution survey must measure a senior mark’s loss of distinctiveness or uniqueness, using stimuli which replicate market conditions as closely as possible. 140. Id. (holding that “screening questions used in this study rely on respondents’ understanding of the ambiguous word ‘purse’ which . . . results in a potentially overinclusive universe”). 141. Id. (“Having failed to obtain satisfactory responses (i.e., no dilution) from his first fifty-eight respondents, Dr. Jacoby terminated the original survey and started over with a new pool of fewer than one hundred respondents, but used substantially the same questions.”). 142. Id. at 450–51. 143. Id. at 451. 144. While the court does not present many examples of questions asked in Dr. Jacoby’s survey, one includes: “I would like you to suppose you actually owned one or both of these [Louis Vuitton] bags. If you actually owned one or both, how would knowing [these Dooney & Bourke bags were] being sold make you feel?” Id. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 30 Columbia Journal of Law and Social Problems [39:1 V. PROPOSAL: USING “UNIQUE VALUE SURVEYS” TO TARGET UNIQUENESS Constructing a dilution survey that comports with Moseley is admittedly a tricky task.145 There are very few cases in which a trademark holder presented survey evidence specifically designed to accommodate the Moseley standard, and the Supreme Court offered no guidance in its opinion on what a valid survey would look like.146 Further, although several scholars have offered recommendations for constructing valid dilution surveys,147 none of these suggestions address Moseley’s instruction that mere mental association of two marks is not valid proof of actual dilution.148 This Part discusses prior suggestions of how to measure actual dilution through consumer surveys and why they are inadequate, and then proposes the use of unique value surveys to address those infirmities. A. PREVIOUS SUGGESTIONS 1. Brand-Equity Surveys One possibility, recommended by Patrick M. Bible, is the use of brand-equity surveys. Such surveys ask questions that gauge if a junior user has weakened a senior user’s “brand equity” — that is, the “set of assets and liabilities that a trademark adds to or subtracts from the value provided by a product or service to a 145. This Note focuses on the challenges specific to constructing dilution survey questions. Other issues which address general survey construction (e.g., sample size, statistical analysis) are beyond the scope of this Note. See infra note 170 and accompanying text. 146. See Chris Brown, A Dilution Delusion: The Unjustifiable Protection of Similar Marks, 72 U. CIN. L. REV. 1023, 1042 (2004) (“[The Court] required a showing of actual dilution but provided virtually no guidance as to how such a showing could be made. It endorsed surveys but did not indicate how the surveys should define dilution or how they should verify its existence.”). 147. Several of the methods discussed in Part V.A were also discussed, albeit briefly, in an amicus brief submitted by the Solicitor General to the Supreme Court for Moseley. Brief for the United States as Amicus Curiae supporting Petitioners in part at 22–24, Moseley v. V Secret Catalogue, Inc., 537 U.S. 418 (2003) (No. 01-1015). 148. Such a failure is understandable, as these suggestions were offered before the Supreme Court’s Moseley ruling. The only post-Moseley commentary that discusses dilution surveys recommends measuring “distinctiveness” but offers no advice on how to do so. Bunker et al., supra note 82, at 53–54. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 31 business or its customers.”149 A brand-equity survey may ask respondents to rank on a nine-point scale certain attributes about the senior mark, such as whether its products are durable or not, if the brand is low-class or high-class, or if the consumer associates the brand with home and family or not.150 For example, if a junior user begins to market BLACK & BECKER baby food, then “the BLACK & DECKER [power tools] mark’s equity would begin to blur from tough to tender and from durable to delicate.”151 Extending this example, a brand-equity survey would survey two populations: one which has been exposed to both BLACK & DECKER power tools and BLACK & BECKER baby food (a test population), and one which has only been exposed to BLACK & DECKER power tools (a control population).152 After asking respondents from both populations, “On a scale of 1–9, how durable are BLACK & DECKER brand power tools?”, if the test population’s rankings are lower than those of the control population to a statistically significant degree, this would constitute proof that the presence of the BLACK & BECKER baby food mark dilutes the senior mark, weakening the “clear, distinctive personality which is part of what the consumer pays for each time he or she purchases the product.”153 Provided that the survey is not otherwise flawed, Bible believes courts should clearly accept such evidence.154 While Bible’s analysis offered useful guidance to courts in the early days of the FTDA,155 it fails to meet Moseley’s current re149. Bible, supra note 12, at 328 (quoting Steve Hartman, Brand Equity Impairment — The Meaning of Dilution, 87 TRADEMARK REP. 418, 419 n.3 (1997)). 150. Id. at 331. 151. Id. at 328. While Bible’s example actually uses identical marks of BLACK & DECKER, this Note deliberately explores situations where marks are not identical, because it is in such circumstances that dilution surveys will be most useful. 152. Id. at 322–23. 153. Id. at 328. Bible provides several other examples of question construction, including asking how much the respondent agrees with a particular value statement about the senior user’s products. Id. at 329. 154. See Pulliam, supra note 63, at 911 n.109 (“Bible guides the practitioner, stating that the court will likely question the survey evidence in the following manner, considering: (1) was the population carefully selected and defined; (2) does the sample relied upon accurately represent the population; (3) was the data accurately acquired and reported; (4) was the data analyzed using accepted statistical analysis; (5) were the questions asked clearly; (6) was the survey conducted by qualified personnel; and (7) was objectivity carefully protected while conducting the survey.”) (internal citations omitted). 155. Several court opinions have cited Bible’s article, including Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Div. of Travel Dev., 170 F.3d 449, 461 n.6 (4th Cir. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 32 Columbia Journal of Law and Social Problems [39:1 quirement of measuring actual dilution. Bible’s brand-equity surveys measure a range of values associated with a senior mark, but they do not strike at the “whittling away” of a mark’s “uniqueness,” which is central to a dilution claim.156 Rather, brand-equity questions measure mental associations based on other qualities, such as desirability. As Malletier instructs, a proper dilution survey must directly measure a senior mark’s ability to distinguish its products without offering other associations as a proxy.157 Asking questions about desirability, dependability, quality or classiness cannot prove actual dilution. 2. Typicality Surveys Bible alternatively suggests using typicality surveys, which test the respondent’s “ability to conjure up a particular product category.”158 In a typicality survey, respondents from both test and control populations are asked to name all the products they associate with the senior mark (e.g., BLACK & DECKER). If respondents in the test population mention the junior user’s products (e.g., baby food) with greater frequency than do respondents in the control population, this may constitute proof that the BLACK & DECKER mark for power tools has been diluted.159 Analysts may also examine other trends, such as how often respondents in the test population mention baby food, the amount of time that passes before power tools are mentioned, and the number of other products named before respondents mention power tools.160 This survey targets how “typical” the senior user’s products are in the mind of the public. If respondents from the 1999). Several amicus briefs submitted to the Supreme Court for Moseley also cited the article. See Brief for the United States as Amicus Curiae supporting Petitioners in part at 22, Moseley v. V Secret Catalogue, Inc., 537 U.S. 418 (2003) (No. 01-1015); Amicus Curiae Brief of the Intellectual Property Law Association in support of Respondents at 8, Moseley, 537 U.S. 418 (No. 01-1015). 156. See Schechter, supra note 22, at 825 (describing the injury inherent in dilution as “the gradual whittling away or dispersion of the [trademark’s] identity”). 157. See Malletier v. Dooney & Bourke, Inc., 340 F. Supp. 2d 415, 450–51 (S.D.N.Y. 2004) (“[D]esirability is unrelated to a mark’s ability to identify and distinguish Louis Vuitton’s handbags . . . .”). 158. Bible, supra note 12, at 329 (quoting Alexander F. Simonson, How and When Do Trademarks Dilute: A Behavioral Framework to Judge “Likelihood” of Dilution, 83 TRADEMARK REP. 149, 152–53 (1993)). 159. Id. at 328–30. 160. Id. at 330. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 33 test population mention baby food before power tools, then the presence of BLACK & BECKER baby food lessens BLACK & DECKER’s ability to identify and distinguish its power tools. This typicality survey method, however, is also flawed. Typicality surveys fundamentally ask the same questions as association surveys (e.g., “What do you think of when you see X?”), which Moseley rejects as insufficient proof of actual dilution. Even though typicality surveys attempt to equate typicality with distinctiveness, the mere mental association of the two marks cannot create a presumption of actual dilution.161 In Ringling Bros., for instance, the court did not inquire as to the order or timing when respondents filled in “GREATEST ____ ON EARTH” with “show” or “snow”; rather, the court ruled that for surveys such as these, respondents’ “impressions must go beyond mere recognition of a visual similarity of the two marks.”162 The only question is whether consumers can keep both marks separate in their minds, not which mark consumers think of first. 3. “Expanded” Typicality Surveys A third possibility, offered by William G. Barber, is similar to Bible’s typicality survey. Barber believes that a valid dilution survey should compare “the strength or association of the plaintiff’s mark absent the alleged diluting use with its strength/association in the presence of the alleged diluting use. If such strength/association has been adversely affected by the alleged diluting use, then dilution is shown.”163 Structurally, Barber recommends the same basic format as used in typicality surveys, but advocates broadening the questions to measure any associations between the senior and junior marks, not just the naming of products.164 161. Moseley v. V Secret Catalogue, 537 U.S. 418, 433 (2003). 162. Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Div. of Travel Dev., 170 F.3d 449, 462–63 (4th Cir. 1999). 163. Barber, supra note 62, at 629–30. 164. See id. at 630. Each group would be asked what (if anything) they associate the plaintiff’s mark with, or what goods or services (if any) they identify it with, or similar questions. If the test group identifies the plaintiff’s mark with the defendant or defendant’s products to a greater extent than the control group, or if the test group associates the plaintiff’s mark with negative images associated with the defendant or File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 34 Columbia Journal of Law and Social Problems [39:1 Barber’s surveys suffer from the same weakness as Bible’s typicality surveys: a difference in mere mental association between a test group and a control group does not prove that a mark’s power to identify and distinguish has been lessened. Barber correctly notes that “[o]nly if the defendant’s use reduces or damages the selling power (i.e., distinctiveness) of the plaintiff’s mark has dilution really occurred.”165 However, Barber then exposes his proposal’s flaw by noting that his design “attempts to measure this reduction in distinctiveness by comparing the strength or association of the plaintiff’s mark with and without the alleged diluting use.”166 Barber equates distinctiveness with associations, which Moseley prohibits. While other commentators offer their own suggestions167 and evaluate those of others,168 these recommendations consistently suffer from the same weaknesses that plague Bible and Barber’s proposals, either conflating mental association with distinctiveness or measuring values that are not synonymous with distinctiveness. Such confusion is understandable, as all of the proposals discussed above, including those of Bible and Barber, were suggested before the Supreme Court’s Moseley ruling. However, as Malletier indicates, a valid post-Moseley dilution survey must di- defendant’s products to a greater extent than the control group, then this would show dilution. 165. Id. at 631. 166. Id. 167. See, e.g., Bunker et al., supra note 82, at 53–54 (recommending measuring a mark’s “distinctiveness,” but offering no advice on how to construct such a survey); Sandra Edelman & Bruce R. Ewing, The Federal Trademark Dilution Act of 1995: A Litigation Perspective, 86 TRADEMARK REP. 485, 502–03 (1996) (discussing brand equity impairment and studies that time respondent’s identification of senior user’s products after being exposed to junior user’s mark); Hartman, supra note 149 (recommending testing brand equity impairment); Jonathan E. Moskin, Dilution or Delusion: The Rational Limits of Trademark Protection, 83 TRADEMARK REP. 122, 138 (1993) (suggesting that surveys measure the reduced likelihood of a respondent to purchase senior user’s product, though the author laments that “[r]ealistically, however, there may be no way to conduct such a survey for purposes of a lawsuit.”); Eric A. Prager, The Federal Trademark Dilution Act of 1995: Substantial Likelihood of Confusion, 7 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. 121, 132–33 (1996) (recommending the use of likelihood of confusion surveys as surrogates for testing the strength of a senior user’s mark); Elliot B. Staffin, The Dilution Doctrine: Towards a Reconciliation with the Lanham Act, 6 FORDHAM INTELL. PROP. MEDIA & ENT. L.J. 105 (1995) (proposing survey questions that gauge mental association). See also Barber, supra note 62, at 631 n.62 (listing additional proposals). 168. See Pokotilow & Fefferman, supra note 133 (reviewing court reactions to preMoseley dilution surveys). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 35 rectly measure the distinctiveness of the senior mark, and not measure other characteristics as proxies.169 B. PROPOSAL: UNIQUE VALUE SURVEYS For a dilution survey to meet the rigorous Moseley standard, it must go beyond testing consumers’ mental associations between two marks, and instead offer proof that the capacity of the senior mark to identify and distinguish its products has been lessened, as required by the FTDA.170 This Note proposes that the best method for achieving this goal is through “unique value surveys” that directly measure a senior mark’s “uniqueness.”171 This Section first describes how to construct unique value surveys, how to prove survey results are valid, and the minimum standards courts should establish before recognizing that such surveys supply proof of actual dilution. This Section then presents examples of unique value surveys before finally identifying and responding to possible criticisms of this proposal. 1. Question Construction in Unique Value Surveys Because post-Moseley courts consistently reject dilution surveys172 that only provide evidence of mental association and not “the lessening of the capacity of a famous mark to identify and 169. Malletier v. Dooney & Bourke, Inc., 340 F. Supp. 2d 415, 450–51 (S.D.N.Y. 2004) (rejecting “desirability” as a proxy for distinctiveness, exclusivity, or value). 170. Moseley v. V Secret Catalogue, 537 U.S. 418, 433 (2003). This requirement is one of many that a proper dilution survey must meet. This Note focuses only on how to construct a question which may elicit a meaningful response. Related issues, such as sampling, establishing statistical significance, and interviewer training, are generally beyond the scope of this Note. It is not unprecedented for an article addressing survey questions to focus on question construction. See Palladino, supra note 84, at 155 (explaining that his article “focuses on designing questions,” not “the many steps involved in conducting a survey, which range from selecting the universe of respondents to tabulating their answers.”). For discussion of the standards any trademark survey should meet, see generally Michael Rappeport, Litigation Surveys — Social “Science” as Evidence, 92 TRADEMARK REP. 957 (2002); Shari Seidman Diamond, Reference Guide on Survey Research, in REFERENCE MANUAL ON SCIENTIFIC EVIDENCE 229, 229–76 (Federal Judicial Center 2d ed., 2000). 171. The term “unique value survey” is an original term proposed by this Note, referring only to the novel method of question construction proposed in this section. 172. See supra Part IV.B (discussing post-Moseley cases which rejected dilution surveys). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 36 Columbia Journal of Law and Social Problems [39:1 distinguish goods or services,”173 a proper dilution survey must ask respondents questions that directly measure this “lessening.” To do so, a survey should follow the general form of brand-equity and typicality surveys by conducting a survey in both a control population (where only the senior mark is present) and a test population (where both the senior and junior marks are present). The survey should then ask respondents, “On a scale of 1–9, how unique is [the senior mark]?”174 If scores from the test population provide lower ratings of uniqueness than do scores from the control population to a statistically significant degree, this would constitute proof of injury toward the senior user; lower rankings of the senior mark’s uniqueness directly prove that the capacity of the senior mark to distinguish its products has been lessened due to the dilutive power of the junior mark. This method of proving dilution respects Schechter’s argument that the preservation of a trademark’s uniqueness should be actionable.175 Surveyors may choose to ask other questions to support the inquiry of uniqueness, such as measuring mental association or desirability, or gathering additional contextual information. Nevertheless, the nucleus of proving whether a senior mark has been diluted by a junior mark’s presence lies first in asking respondents for their rankings of the senior mark’s uniqueness. This is the purest way to identify a mark’s reduced ability to identify and distinguish its goods and services — the essence of dilution.176 15 U.S.C. § 1127 (2000). The definition of “unique” in this context is “unusual” or “notable.” WEBSTER’S THIRD NEW INTERNATIONAL DICTIONARY 2500 (1993). While the traditional definition of “unique” suggests it cannot be used in comparative contexts (an item is either one-of-akind or is not), “[i]n modern use both comparison and modification are widespread and standard.” Merriam-Webster OnLine, http://www.m-w.com/cgi-bin/dictionary?book= Dictionary&va=unique (last visited Aug. 16, 2005). 175. Schechter, supra note 22, at 831 (emphasis added). A survey should not ask control and test populations “How much would you pay for this item?”, because this question measures desirability of an item, not the uniqueness of a mark. See supra Part V.A.1. 176. The surveyor must ask about uniqueness before asking other questions (such as mental association questions) to avoid tainting the population, accidentally giving an indication of what a “correct” answer would be. However, a survey could help respondents feel comfortable with the ranking method by first asking respondents to gauge the uniqueness of other marks that are unrelated to the study. 173. 174. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 37 2. Proving Validity of Unique Value Surveys An important element for establishing the validity of a unique value survey is to make sure that test and control populations are as similar as possible in every respect except for the presence of the junior mark. As Barber suggests, “the most practical way to do this is to use as the control group a sample of consumers in geographic areas remote from where the defendant is using the accused mark. The test group then would be taken from geographic areas where both parties are using their marks.”177 Reliability can be increased by securing multiple control and test populations from various locations; if scores from test populations consistently rank a senior mark with lower scores of uniqueness than do control populations, this could constitute valid proof of actual dilution.178 To avoid concerns that a particular respondent in the control group is familiar with the junior mark, or that a respondent in either population is not familiar with the senior mark, post-survey screening questions may confirm that the respondent is a representative member of the appropriate population, with scores from non-representative members being excluded. Analysts then may control for as many variables as possible. If the score differential between the two populations is still statistically significant, such a survey could offer compelling proof of actual dilution.179 177. Barber, supra note 62, at 630. However, less expensive alternatives exist. “For example, the surveyor could ask all of the respondents post-interview questions to determine which ones had been previously exposed to the defendant’s mark. Those who had not would be categorized in the control group, those who had in the test group.” Id. at 631. 178. Id. at 630–31. 179. Addressing trademark dilution surveys in general, some commentators suggest testing a specific population — one which already has been exposed to the senior mark — repeatedly over various times (e.g., annually). If a junior mark then arises in that market, the same respondents could then be treated as both the test and control populations. For example, [a] regular and continuing series of surveys could track the mark’s place in the minds of consumers. This survey data would establish a baseline that would make it easier to argue the diluting effects of a junior mark. It would also be easier to show the potential for dilution by a junior mark before it enters the market. Bunker et al., supra note 82, at 54. While such a method sounds attractive, it is ultimately inadequate. First, this strategy requires senior users to commit to a never-ending cycle of surveys (and their accompanying financial costs), even if no danger of dilution ever arises. Second, this strategy compromises the validity of control groups. Only by comparing two different groups at the same time, rather than comparing the same group at dif- File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 38 Columbia Journal of Law and Social Problems [39:1 3. Establishing Standards for Unique Value Surveys Imposing a permanent injunction upon a junior user under the auspices of the FTDA is a severe remedy, considering that the junior mark need not confuse consumers, nor even be in competition with the senior user’s products.180 As such, in instances where unique value surveys display statistically significant levels of dilution, courts must decide whether the amount of actual dilution present is severe enough to warrant relief under the FTDA. Setting a minimum bar for gauging dilution is thoroughly within a court’s purview. For example, courts already recognize minimal requirements for “likelihood of confusion” surveys, usually in the 15% to 20% range.181 Setting a fixed number, however, is also dangerous. Considering the extreme nature of relief under the FTDA, Bible suggests that courts “may require a showing slightly higher than the typical [15% to 20%] necessary for finding infringement based on likelihood of confusion.”182 While Bible erroneously believed that high levels of mental association could constitute valid proof of actual dilution, his suggestion does reveal the natural intuition that courts should set a stricter standard of proof for dilution than for confusion. Courts should use their discretion in good faith by only acknowledging actual dilution when results from unique value surveys reach statistically significant levels that leave little doubt actual dilution is taking place. While different circuits may reach different standards, a spread where at least half of the respondents in the test population rank the senior mark as being less unique than do respondents in the control population certainly ferent times, can a senior user adequately control for external factors. Otherwise, the element of temporality will always raise questions over a survey’s validity. Third, how would a senior user even know where a junior user’s mark will arise in advance? While such a method may prove useful to notify a senior user that its mark is somehow losing uniqueness, it offers no evidence that a particular junior mark caused the dilution at issue. 180. 15 U.S.C. § 1127 (2000) (allowing for FTDA relief “regardless of the . . . absence of (1) competition between the owner of the famous mark and the other parties, or (2) likelihood of confusion”). 181. Bible, supra note 12, at 335. See also Xuan-Thao N. Nguyen, The New Wild West: Measuring and Proving Fame and Dilution Under the Federal Trademark Dilution Act, 63 ALB. L. REV. 201, 239 n.270 (1999) (listing many cases where courts considered confusion rates which ranged from 15% to 20%, and one case where a 14% confusion rate was given weight). 182. Bible, supra note 12, at 335. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 39 suggests dilution in action. To display this dilution, analysts could first determine the average ranking from the control population and then determine the percentage of respondents from the test population who provided significantly lower rankings.183 Additionally, courts might heed Judge Posner’s advice by appointing a neutral third-party analyst to assist in interpreting submitted surveys, thereby avoiding the problem of “battling experts.”184 Ultimately, courts must balance the extremity of the relief under the FTDA with the need to protect senior users from dilution. Before a court awards injunctive relief to a senior user when the junior mark is both non-identical and used on a noncompeting good, the court must have full confidence that actual harm is taking place. If, however, a unique value survey indicates that actual dilution is undoubtedly lessening a senior mark’s ability to distinguish its products to a significant degree, a court should not hesitate to award relief under the FTDA, especially because the FTDA’s other requirements help to limit illegitimate dilution claims.185 4. Examples: Applications of Unique Value Surveys Applying unique value surveys to previous cases where courts rejected surveys as proof of actual dilution reveals the potential power of this proposal, particularly when the marks are not identical.186 For example, in Kellogg Co. v. Toucan Golf, Inc.,187 Kellogg could have surveyed control and test populations, asking “On 183. As a control measure, analysts could also offer a large difference in the populations’ average ratings as evidence of actual dilution. 184. Indianapolis Colts, Inc. v. Metro. Baltimore Football Club Ltd. P’ship, 34 F.3d 410, 415 (7th Cir. 1994) (discussing the difficulties judges face when survey experts disagree, suggesting that such a system “might be improved by asking each party’s hired expert to designate a third, a neutral expert who would be appointed by the court to conduct the necessary studies”). But see Michael Rappeport, The Role of the Survey “Expert” — A Response to Judge Posner, 85 TRADEMARK REP. 211, 220 (1995) (suggesting that a neutral third party is unnecessary because “over and over judges have surely demonstrated that they are capable of using their fundamental legal training to make very competent evaluations of survey testimony. In essence, when it comes to survey research evidence, there already is a neutral expert in court. We call that person the judge.”). 185. For instance, only “famous” marks are eligible for relief. 15 U.S.C. § 1125(c)(1). 186. If the marks are identical, then the senior user need not even present a survey — identical marks are generally treated as constituting “circumstantial evidence” of actual dilution. See supra note 73. 187. 337 F.3d 616 (6th Cir. 2003). See supra Part IV.B for discussion of the facts of this case. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 40 Columbia Journal of Law and Social Problems [39:1 a scale from 1–9, how unique is this mascot?”, while showing respondents a Froot Loops cereal box with Toucan Sam prominently displayed.188 If the two populations offered similar rankings of uniqueness, then the survey would not provide valid proof of actual dilution. If the test population’s rankings of Toucan Sam’s uniqueness were consistently and reliably lower than those of the control population, the survey would provide valid proof of actual dilution and Kellogg would deserve relief under the FTDA. Even though the cartoon toucan and the lifelike toucan are not identical, a unique value survey at least would provide Kellogg with an opportunity to prove actual dilution is at work.189 Unique value surveys also can provide proof of actual dilution under more complex fact patterns, such as when the junior user claims its mark may increase the senior user’s sales. For example, in Ringling Bros.-Barnum & Bailey Combined Shows, Inc. v. Utah Division of Travel Development,190 the circus could have surveyed respondents both inside and outside Utah, asking, “On a scale from 1–9, how unique is the slogan ‘THE GREATEST SHOW ON EARTH?’” If respondents in Utah provided lower rankings of uniqueness than respondents outside of Utah, Ringling Bros. could argue that the Utah Division of Travel’s slogan dilutes its mark. Such evidence would suggest that, all else being equal, the Ringling Bros. slogan is no longer as unique to Utahans. If, however, the populations provided similar rankings of the slogan’s uniqueness, the court should not find dilution. Under this method, the potential counterargument that Utah’s mark might increase Ringling Bros.’ sales through “magnetism” is irrelevant191 — the injury is the loss of uniqueness, not the loss of sales.192 188. In this hypothetical, a surveyor must present the entire box of Froot Loops and not a decontextualized image of Toucan Sam, to adhere to the Playtex court’s recommendation that the stimulus mimic the mark’s actual appearance in the market as closely as possible. Playtex Prods. v. Georgia-Pacific Corp., 390 F.3d 158, 168 (2d Cir. 2004). 189. See supra Part III.B for discussion of the difficulties facing non-identical marks under the FTDA. 190. 170 F.3d 449 (4th Cir. 1999). See supra Part IV.A for discussion of the facts of this case. 191. Id. at 460 (“[C]ommon sense suggests that an occasional replicating use might even enhance a senior mark’s ‘magnetism’ — by drawing renewed attention to it . . . .”). 192. But see note 68 and accompanying text (suggesting that loss of uniqueness is not an injury in itself). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 41 5. Potential Criticisms of Unique Value Surveys a. The Word “Unique” Is Too Vague One possible weakness of unique value surveys concerns the definition of “unique,” and whether this word has different meanings to different populations. What the word “unique” means to one person may differ from what it means to another. If respondents in the test population have a different understanding of “unique” than do respondents in the control population, then any difference between the two groups could be based on reactions to the word “unique,” and not a true lessening of the mark’s distinctiveness. In other words, so goes this critique, measuring “uniqueness” does not actually measure anything. This concern, however, is based on the assumption that different populations have consistently idiosyncratic understandings of the meaning of the word “unique,” an unlikely proposition. While it is true that any individual respondent may have a slightly independent conception of the definition of “unique,” there is no reason to believe that a randomly sampled test population would have a consistently different definitional understanding of “unique” when compared to a randomly sampled control population (particularly to such extreme levels as to suggest actual dilution).193 While it might seem arbitrary to have subjects assign a “score” to Toucan Sam’s “uniqueness,” unique value surveys do not measure a particular score’s significance. Instead, unique value surveys measure the test and control populations’ average scores relative to each other. Assuming that the test and control populations will, on average, have similar understandings of the word “unique,” a statistically significant difference in the average rank of uniqueness, after controlling for external variables, may be attributed to actual dilution. Additionally, courts can mitigate the effects of vagueness by setting a strict minimum level of cognizable dilution, refusing to recognize measurements of actual 193. A possible solution to this worry is to first ask the unique value question, and then ask the respondent to pick the definition of “unique” they used from a list. The analyst could then screen out respondents whose definitions of “unique” are inconsistent with “unusual” or “notable.” See supra note 174 (discussing the appropriate definition of “unique”). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 42 Columbia Journal of Law and Social Problems [39:1 dilution that are statistically significant but weak.194 Finally, adequate sample sizes and comparability of test and control populations, including the use of multiple populations in various geographic locations, abate the power of this counterargument. b. Measurements of Uniqueness Are Actually Measurements of Mental Association. Second, it might be possible that all respondents uniformly understand the word “unique” to mean “mental association.” If this is so, then unique value surveys would only serve to measure a mental association between the marks, and not the true “uniqueness” of the senior mark. For example, asking “How unique is the slogan ‘THE GREATEST SHOW ON EARTH?’” might be understood as asking “Can you think of other slogans similar to the slogan ‘THE GREATEST SHOW ON EARTH?’” If this is accurate, then unique value surveys would fall victim to the same weaknesses as the surveys discussed by Bible, Barber, and Malletier. Analysts may avoid this pitfall by asking control and test populations questions designed to measure mental association.195 Since mental association of two marks is a necessary precondition for dilution,196 results in the test population that show a mental association between the two marks should be higher than results from a unique value survey. If respondents are first asked to rank the senior mark’s uniqueness, and later asked questions addressing mental association, analysts may compare the two scores. If scores measuring uniqueness are different from those measuring mental association, this would suggest that respondents treat the two concepts differently. 194. See supra Part V.B.3 (discussing possible minimum standards for recognizing unique value surveys). 195. Such questions would gauge, for instance, if consumers associate THE GREATEST SNOW ON EARTH with the Ringling Bros. circus. 196. One mark cannot weaken another unless consumers somehow connect the two. See Jacobs, supra note 62, at 193 (stating that “courts should continue to ask initially whether consumers mentally associate the defendant’s mark with the plaintiff’s mark” as a prerequisite). But see McCarthy, supra note 17, at 737–38 (“[The] assumption of mental connection is not accurate with marks consisting of words that have a meaning other than as a source signifier.”). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 43 c. It Will Still Be Too Difficult to Prove Actual Dilution A third possible criticism of unique value surveys is that despite their supposed benefits for allowing trademark holders to prove dilution of their famous marks, such surveys will still rarely provide results proving that a junior mark strongly dilutes a senior mark. While such a counterargument cannot be proven until plaintiffs test this assumption, it is possible that results displaying significant levels of actual dilution will be uncommon. Advocates of this counterargument should consider that relief under the FTDA should likewise be uncommon, reserved for cases where a junior mark clearly and strongly lessens a senior mark’s uniqueness.197 Surveys will continue to be “expensive, timeconsuming and not immune to manipulation.”198 FTDA claims will continue to be reserved for famous marks. These restrictions help confine relief under the FTDA to the extraordinary cases where dilution is obvious and destructive. More important than the ultimate results of a survey, however, is that senior users have some method for proving dilution. Currently, a senior user who believes her mark is being diluted by a non-identical junior mark has no practical recourse. Providing these senior users the option of presenting unique value surveys as proof of actual dilution will give them a chance to gain FTDA relief. While actual dilution still may be difficult to prove, unique value surveys crack the door open slightly wider for trademark holders who currently have nowhere else to turn. 197. McCarthy, supra note 17, at 747 (stating that FTDA relief should only apply to “extraordinary cases”). 198. Nabisco, Inc. v. PF Brands, Inc., 191 F.3d 208, 224 (2d Cir. 1999). File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 44 Columbia Journal of Law and Social Problems [39:1 VI. CONCLUSION Congress may eventually amend the FTDA, lowering the requirement from “causes dilution” to “likely to cause dilution.”199 Such an amendment would no doubt benefit senior users, because they would only need to prove a likelihood of future dilution rather than waiting for a junior user to place her goods or services in the market and become powerful enough to cause actual dilution at judicially cognizable levels. Until the FTDA is amended, however, senior users must prove actual dilution to receive relief. This Note has argued that while various methods of proving actual dilution are available, consumer surveys may offer the best opportunity for senior users to gain relief under the FTDA, particularly when the two marks in question are not identical. 199. Various organizations have lobbied Congress to amend the FTDA since its passage in 1996. Most recently, Congress addressed this request through the proposed Trademark Dilution Revision Act of 2005 (“TDRA”). H.R. 683, 109th Cong. (2005). The TDRA would, inter alia, create a federal cause of action when a junior user’s mark is “likely to cause dilution by blurring,” thus overturning Moseley’s requirement of actual dilution. Id. § 2. This legislation is still in preliminary stages — a version has been passed by the House of Representatives and is currently being reviewed by the Senate. The TDRA, however, may not pass (an outcome that the American Civil Liberties Union supports), or it may eventually pass but in a substantially different form. See Susan Progoff, The State of Trademark Dilution Law, http://www.comml-iba.org/ modules.php?name=News&file=article&sid=65 (June 23, 2004) (“Because the current bill is in the early stages of the legislative process, however, it remains to be seen whether any amendment will be enacted and whether the bill will survive in its current form.”). Until such legislation is passed in a way that directly overturns Moseley, senior users will continue to face extreme burdens unless they have the option of presenting consumer survey data. Even if the TDRA eventually passes in its current form, overturning Moseley, the probative value of unique value surveys remains strong. While the new “likely to dilute” standard would provide much assistance to senior users who do not want to wait for an injury to occur before qualifying for relief (such as the plaintiff in Nabisco), unique value surveys still provide a powerful way for senior users who have already been harmed to unequivocally prove their marks have been diluted. Because senior users could choose (and often have chosen) to introduce survey evidence to prove mental association under the “likely to dilute” standard, there would be little cost to adding an additional question to a survey which asks for measurements of uniqueness. If responses come even close to suggesting that actual dilution is present, such a result would, by definition, establish that the junior mark is “likely to dilute.” Finally, much like the “circumstantial evidence” prong of Moseley, the “likely to dilute” standard is subject to a judge’s idiosyncratic discretion. A unique value survey, however, may provide such compelling proof of dilution that no reasonable judge could ever suggest that the junior mark is not “likely” to dilute — the survey shows that it already has diluted. File: Anten-10-15 Created on: 9/29/2005 2:21 PM Last Printed: 11/9/2005 1:59 PM 2005] In Defense of Trademark Dilution Surveys 45 Though courts consistently reject consumer surveys as providing proof of actual dilution, primarily because most surveys only measure the mental association between two marks and not the loss of a mark’s uniqueness that characterizes dilution, unique value surveys solve this problem by directly measuring the lessening of the senior mark’s capacity to identify and distinguish its products by directly gauging a loss of “uniqueness.” Unique value surveys will likely be as expensive, time-consuming, and unreliable as any other trademark survey, but they also respond to the challenge of Judge Posner and others who doubt whether a “question could be put to consumers that would elicit a meaningful answer.”200 Unique value surveys finally can reveal that answer. 200. Ty Inc. v. Softbelly’s, Inc., 353 F.3d 528, 535 (7th Cir. 2003).