Citation: C. Scott Hemphill; Jeannie Suk, The Law, Culture, and Economics of Fashion, 61 Stan. L. Rev. 1147 (2009) Content downloaded/printed from HeinOnline Tue Feb 20 10:14:21 2018 -- Your use of this HeinOnline PDF indicates your acceptance of HeinOnline's Terms and Conditions of the license agreement available at http://heinonline.org/HOL/License -- The search text of this PDF is generated from uncorrected OCR text. -- To obtain permission to use this article beyond the scope of your HeinOnline license, please use: Copyright Information Use QR Code reader to send PDF to your smartphone or tablet device THE LAW, CULTURE, AND ECONOMICS OF FASHION C.Scott Hemphill* &Jeannie Suk** INTRODUCTION .................................................................................................... 1148 I. WHAT IS FASHION? ........................................................................................ 1155 A. Status......................................................................................................... 1156 B.Zeitgeist..................................................................................................... 1157 C. Copies Versus Trends................................................................................ 1159 D. Why PromoteInnovation in Fashion?...................................................... 1161 II. A MODEL OF TREND ADOPTION AND PRODUCTION ......................................... 1164 A. DifferentiationandFlocking.................................................................... 1164 B. TrendA doption ......................................................................................... 1166 C. Trend Production...................................................................................... 1168 III. How UNREGULATED COPYING THREATENS INNOVATION ............................. 1170 A. Fast-Fashion Copyists.............................................................................. 1170 B . The Threat to Innovation........................................................................... 1174 1.H armfulcopying ................................................................................. 1175 2. D istortinginnovation......................................................................... 1176 * Associate Professor of Law and Milton Handler Fellow, Columbia Law School. ** Assistant Professor of Law, Harvard Law School. The authors thank Robert Ahdieh, Yochai Benkler, Johanna Blakley, Rachel Brewster, Vernon Cassin, Glenn Cohen, Bob Cooter, Domenico De Sole, Hal Edgar, Liz Emens, Noah Feldman, Robert Ferguson, Amy Finkelstein, Terry Fisher, Jeanne Fromer, Bill Gentry, Jane Ginsburg, Victor Goldberg, Jack Goldsmith, Paul Goldstein, Jeff Gordon, Laura Hammond, Michael Heller, Lauren Howard, Olati Johnson, Avery Katz, Alon Klement, Meg Koster, Mark Lemley, Larry Lessig, Doug Lichtman, Clarisa Long, Ronald Mann, Sara Marcketti, Martha Minow, Henry Monaghan, Ed Morrison, Melissa Murray, Ben Olken, John Palfrey, Mitch Polinsky, Richard Posner, Alex Raskolnikov, Kal Raustiala, Susan Scafidi, David Schizer, Elizabeth Scott, Steve Shavell, Chris Sprigman, Matt Stephenson, Francine Summa, Cass Sunstein, John Witt, and audiences at Columbia Law School, Harvard Law School, Stanford Law School, the University of Tokyo, and the New Yorker magazine's 2008 annual conference for helpful discussions and comments. Bert Huang provided constant collaborative advice and support. We thank Sarah Bertozzi, Melanie Brown, Andrew Childers, Jon Cooper, Brittany Cvetanovich, Zeh Ekono, Joseph Fishman, Ilan Graff, Brett Hartman, Andrea Lee, Samantha Lipton, Ruchi Patel, Zoe Pershing-Foley, Miriam Weiler, and Ming Zhu for excellent research assistance, and the staff of the Columbia Law School, Fashion Institute of Technology, and Harvard Law School libraries for their efforts procuring difficult sources. Special thanks to the several dozen stakeholders-in fashion houses, government agencies, industry associations, and law firms-for interviews from which we gained valuable insights on fashion design and the fashion industry. Views and errors in this Article are those of the authors only. 1147 STANFORD LA W RE VIEW C.Is PiracyReally Beneficial?............................................1180 IV. TAILORED PROTECTION FOR ORIGINAL DESIGNS ............................1184 A. The Scope ofthe R ight ...............................................1185 B .ConsideringObjections...............................................1190 CONCLUSION ............................................................1195 APPEN DIX .............................................................1197 INDEX OF FIGURES AND TABLES Table 1.Selected U.S. Litigation Against Forever 21, 2007-2008 ...................1174 Table 2. Selected European Litigation, 2005-2008 .............................1191 Figure 1.Foley & Corinna and Forever 21 ..................................1197 Figure 2. Jonathan Saunders and Forever 21.................................1198 Figure 3. Yves Saint Laurent and Ralph Lauren ...............................1199 INTRODUCTION Fashion is one of the world's most important creative industries. It is the major output of a global business with annual U.S. sales of more than $200 billion-larger than those of books, movies, and music combined.' Everyone wears clothing and inevitably participates in fashion to some degree. Fashion is also a subject of periodically rediscovered fascination in virtually all the social sciences and the humanities. 2 It has provided economic thought with a canonical example in theorizing about consumption and conformity. 3 Social 1. U.S. apparel sales reached $196 billion in 2007. The U.S. Apparel Market 2007 Dresses Up... Way Up,Bus. WIRE, Mar. 18,2008 (reporting estimate by the NPD Group). Among fashion accessories, considering just one category, handbags, adds another $5 billion in sales. Tanya Krim, There's Nothing "Trivial" About the Purse-suitof the Perfect Bag, BRANDWEEK, Mar. 29, 2007 (reporting U.S. sales exceeding $5 billion in 2005). For comparison, U.S. publishers had net sales of $25 billion in 2007. Press Release, Ass'n of Am. Publishers, AAP Reports Book Sales Rose to $25 Billion in 2007 (Mar. 31, 2008), http://www.publishers.org/main/IndustryStats/indStats_02.htm. The motion picture and video industry had estimated revenues of $64 billion in 2003. U.S. CENSUS BUREAU, 2003 SERVICE ANNUAL SURVEY, INFORMATION SECTOR SERVICES (NAICS 51)-ESTIMATED REVENUE FOR EMPLOYER FIRMS: 1998 THROUGH 2003, at 1 tbl.3.0.1, available at http://www.census.gov/svsd/www/sas5l-l.pdf; see also MOTION PICTURE ASS'N OF AM., INC., ENTERTAINMENT INDUSTRY MARKET STATISTICS 2007, at 3, available at http://www.mpaa.org/USEntertainmentlndustryMarketStats.pdf (reporting U.S. box office sales ofnearly $10 billion in 2007). The music industry had U.S. revenue, measured at retail, of about $10 billion in 2007. RECORDING INDUS. ASS'N OF AM., 2007 YEAR-END SHIPMENT STATISTICS, availableat http://www.riaa.com/keystatistics.php. Thus fashion is comparable in importance to other core creative industries even if, as seems plausible, some apparel has a lower intellectual property content. 2. See, e.g., LARS SVENDSEN, FASHION: A PHILOSOPHY 7 (John Irons trans., Reaktion 2006) ("Fashion has been one of the most influential phenomena in Western civilization since the Renaissance."). 3. See,e.g., Harvey Leibenstein, Bandwagon,Snob,and Veblen Effects inthe Theory ofConsumers' Demand,64 Q.J.ECON. 183 (1950); see also, e.g.,Sushil Bikhchandani et al., A Theory ofFads,Fashion,Custom, and CulturalChangeasInformationalCascades,100 J. 1148 [Vol. 61:1147 March 2009] FASHION 1149 thinkers have long treated fashion as a window upon social class and social change. Cultural theorists have focused on fashion to reflect on symbolic meaning and social ideals. 5 Fashion has also been seen to embody representative characteristics of modernity, and even ofculture itself.6 Indeed, it is hard to imagine a locus of social life-whether in the arts, the sciences, politics, academia, entertainment, business, or even law or moralitythat does not exhibit fashion in some way. 7 People flock to ideas, styles, methods, and practices that seem new and exciting, and then eventually the intensity of that collective fascination subsides, when the newer and hence more exciting emerge on the scene. Participants of social practices that value innovation are driven to partake of what is "original," "cutting edge," "fresh," "leading," or "hot." But with time, those qualities are attributed to others, and another trend takes shape. This is fashion. The desire to be "in fashion"-most POL. ECON. 992 (1992); Philip R.P. Coelho & James E. McClure, Toward an Economic Theory of Fashion, 31 ECON. INQUIRY 595 (1993); Wolfgang Pesendorfer, Design Innovation and FashionCycles, 85 AM. ECON. REV. 771 (1995); Dwight E. Robinson, The Economics of FashionDemand, 75 Q.J. EcON. 376 (1961); George J. Stigler & Gary S. Becker, De GustibusNon Est Disputandum, 67 AM. ECON. REV. 76, 76 (1977). 4. See, e.g., THORSTEIN VEBLEN, THE THEORY OF THE LEISURE CLASS (Dover Publ'n 1994) (1899); Georg Simmel, Fashion, 10 INT'L Q. 130 (1904), reprintedin 62 AM. J. Soc. 541 (1957); see also, e.g., QUENTIN BELL, ON HUMAN FINERY (Shocken Books 1976) (1949); PIERRE BOURDIEU, DISTINCTION: A SOCIAL CRITIQUE OF THE JUDGEMENT OF TASTE (Richard Nice trans., Harvard Univ. Press 1984) (1979); DIANA CRANE, FASHION AND ITS SOCIAL AGENDAS (2000); KURT LANG & GLADYS ENGEL LANG, COLLECTIVE DYNAMICS 465-88 (1961); PHILIPPE PERROT, FASHIONING THE BOURGEOISIE: A HISTORY OF CLOTHING IN THE NINETEENTH CENTURY (Richard Bienvenue trans., Princeton Univ. Press 1994) (1981); JOHN RAE, THE SOCIOLOGICAL THEORY OF CAPITAL 218-36, 245-76 (Charles Whitney Mixter ed., Macmillan Co. 1905) (1834); Bernard Barber & Lyle S. Lobel, "Fashion" in Women's ClothesandtheAmerican Social System, 31 Soc. FORCES 124 (1952). 5. See, e.g., ROLAND BARTHES, THE FASHION SYSTEM (Matthew Ward & Richard Howard trans., Farrar, Straus & Giroux 1983) (1967); JENNIFER CRAIK, THE FACE OF FASHION: CULTURAL STUDIES IN FASHION (1994); FRED DAVIS, FASHION, CULTURE, AND IDENTITY (1992); Edward Sapir, Fashion,in 6 ENCYCLOPAEDIA OF THE SOCIAL SCIENCES 139, 139-44 (Edwin R.A. Seligman ed., 1931). 6. See, e.g., JEAN BAUDRILLARD, FOR A CRITIQUE OF THE POLITICAL ECONOMY OF THE SIGN 78 (1981); FASHION AND MODERNITY (Christopher Breward & Caroline Evans eds., 2005); Herbert Blumer, Fashion: From Class Differentiation to Collective Selection, 10 SOC. Q. 275 (1969); A.L. Kroeber, On the Principleof Orderin Civilizationas Exemplified by ChangesofFashion,21 AM. ANTHROPOLOGIST 235 (1919). 7. See, e.g., ADAM SMITH, THE THEORY OF MORAL SENTIMENTS 283 (Augustus M. Kelley 1966) (1759) ("[T]he influence of custom and fashion over dress and furniture is not more absolute than over architecture, poetry, and music."); Jeff Biddle, A Bandwagon Effect in PersonalizedLicense Plates?, 29 ECON. INQUIRY 375 (1991); Bikhchandani et al., supra note 3, at 1010-14; John F. Burnum, Medical Practiced la Mode: How Medical Fashions DetermineMedical Care, 317 NEW ENG. J. MED. 1220 (1987); B. Peter Pashigian et al., Fashion,Styling, and the Within-Season Decline in Automobile Prices,38 J.L. & ECON. 281 (1995); Stigler & Becker, supra note 3, at 87; Cass R. Sunstein, Foreword: On Academic Fads and Fashions,99 MICH. L. REV. 1251 (2001); cf Lawrence v. Texas, 539 U.S. 558, 598 (2003) (Scalia, J., dissenting) ("[Tihis Court... should not impose foreign moods, fads, or fashions on Americans." (quoting Foster v. Florida, 537 U.S. 990 (2002) (Thomas, J., concurring))). STANFORD LAW REVIEW visibly manifested in the practice of dress-captures a significant aspect of social life, characterized by both the pull ofcontinuity with others and the push of innovation toward the new. In the legal realm, this social dynamic of innovation and continuity is most directly engaged by the law of intellectual property. At this moment, fashion itself has the attention of federal policymakers, as Congress considers whether to provide copyright protection for fashion design,8 a debate that is sure to continue in the face of fashion designers' many complaints of harm by design copyists.9 Despite being the core of fashion and legally protected in Europe, fashion design lacks protection against copying under U.S. intellectual property law. 1 Thus it has seemed sensible to posit that fashion design is relevantly different from literature, music, and art, where legal protection from copying is thought to be necessary to provide producers an incentive to create." Indeed, 8. See Design Piracy Prohibition Act, S. 1957, 110th Cong. § 2(a), (d) (2007); Design Piracy Prohibition Act, H.R. 2033, 110th Cong. § 2(a), (d) (2007); ABA Section of Intellectual Prop. Law, Proposed Resolution 2008 Council-lA (approved Aug. 9, 2008), available at http://www.abanet.org/intelprop/annual2008/business-session/ 2008Council 1A.pdf ("Resolved, that the Section of Intellectual Property Law, believing that there is sufficient need for greater intellectual property protection than is now available for fashion designs, supports, in principle, enactment of federal legislation to provide a new limited copyright-like protection for such designs; and now therefore, the Section supports enactment of H.R. 2033 ... or similar legislation."); see also Eric Wilson, When Imitation's Unflattering, N.Y. TIMES, Mar. 13, 2008, at G4 (describing designers' efforts to secure copyright protection). 9. For example, an industry-sponsored website collects quotations from designers Oscar de la Renta, Dayna Foley, Phillip Lim, Nicole Miller, Zac Posen, Narciso Rodriguez, and Diane von Furstenberg, and a video posted to the site quotes top executives at Armani, Chanel, Dior, Ferragamo, Hermes, and Marc Jacobs, among others. See Stop Fashion Piracy, The Industry Speaks Out, http://www.stopfashionpiracy.com/theindustryspeaks.php (last visited Jan. 31, 2009). 10. Garments are "useful articles" not protected by copyright, except to the extent that an article's expressive component is "separable" from its utility. See infra Part IV.A for an explanation and critique of the current copyright regime as applied to fashion. Trademark law protects fashion firms' logos against infringement and counterfeiting. For a discussion of trademarks and counterfeiting, see Jonathan M. Barnett, Shoppingfor Gucci on Canal Street: Reflections on Status Consumption, IntellectualProperty,and the Incentive Thesis, 91 VA. L. REv. 1381 (2005). Design patents provide protection in a few cases, but their demanding standards for protection and long lead time make them of limited use for most fashion articles. For a useful overview of the law and history of intellectual property protection and fashion design, see Susan Scafidi, IntellectualPropertyandFashionDesign, in 1INTELLECTUAL PROPERTY AND INFORMATION WEALTH: COPYRIGHT AND RELATED RIGHTS 115 (Peter K. Yu ed., 2006). For a comparative discussion of European copyright for fashion design, see Matthew S. Miller, Piracy in Our Backyard: A ComparativeAnalysis of the Implications of Fashion Copying in the United States for the International Copyright Community, 2 J.INT'L MEDIA & ENT. L. 133, 141-44 (2008). 11. See Kal Raustiala & Christopher Sprigman, The Piracy Paradox:Innovationand Intellectual Property in Fashion Design, 92 VA. L. REv. 1687 (2006); see also Design PiracyProhibitionAct: Hearingon H.R. 5055 Before the Subcomm. on Courts, theInternet, and IntellectualProperty of the H. Comm. on the Judiciary,109th Cong. (2006) (statement of David Wolfe, Creative Director, Doneger Group), 2006 WL 2127241; Sarah J. Kaufman, Note, Trend Forecast:Imitation is a Legal Form of Flattery-LouisVuitton Malletier v. 1150 [Vol. 61:1147 March 2009] some commentators even suggest that perhaps fashion design is so different from other arts that its vitality, or even survival, paradoxically depends on the existence of the opposite kind of regime-a culture of tolerated rampant copying. 12 This Article enters the debate about intellectual property protection and fashion design 13 -a debate in which the fashion industry finds itself divided1 4 -and argues for a limited right against design copying. We set the legal policy debate within a reflection on the cultural dynamics of innovation as a social practice. Fashion in the realm of dress is a version of a ubiquitous phenomenon, the ebb and flow of trends wherein the new ineluctably becomes old and then leads into the new. Fashion is commonly thought to express individuality, and simultaneously to exemplify conformity. The dynamics of fashion lend insight into the dynamics of innovation more broadly. Our motivation here is threefold. First, as the most immediate visible marker of self-presentation, fashion communicates meanings that have individual and social significance. Innovation in fashion creates vocabularies for self-expression that relate individuals to social worlds. As with other creative goods, intellectual property law plays a role in shaping the quantity and the direction of innovation produced by the fashion industry and made available for consumption by people who wear clothing-that is, everyone-a group larger than those who consume art, music, or books. Second, the fashion Dooney & Bourke, Inc., 23 CARDOZO ARTS &ENT. L.J. 531, 532-35 (2005). 12. See Kal Raustiala & Christopher Sprigman, Fashion Victims: How CopyrightLaw Could Kill the Fashion Industry, NEW REPUBLIC ONLINE, Aug. 14, 2007, available at http://www.law.ucla.edu/home/News/Detail.aspx?recordid= 1188; see also James Surowiecki, The PiracyParadox,NEW YORKER, Sept. 24, 2007, at 90. But see Julie P. Tsai, Comment, FashioningProtection: A Note on the Protection of Fashion Designs in the United States, 9 LEWIS & CLARK L. REv. 447 (2005); Diane von Furstenberg, Letter to the Editor, FashionPolice,NEW YORKER, Oct. 22, 2007, at 16. 13. A recent efflorescence of law review commentary features debate on the merits and scope of copyright protection for fashion design, in view of the proposed Design Piracy Prohibition Act. See, e.g., Shelly C. Sackel, Art Is in the Eye of the Beholder: A Recommendationfor TailoringDesignPiracyLegislationto ProtectFashionDesign and the Public Domain, 35 AIPLA Q.J. 473 (2007); Lynsey Blackmon, Comment, The Devil Wears Prado: A Look at the Design Piracy Prohibition Act and the Extension of Copyright Protection to the World ofFashion,35 PEPP. L. REv. 107 (2007); Emily S. Day, Comment, Double-EdgedScissor: Legal Protectionfor FashionDesign, 86 N.C. L. REv. 237 (2007); Lisa J.Hedrick, Note, TearingFashionDesign ProtectionApart at the Seams, 65 WASH. & LEE L. REv. 215 (2008); Lauren Howard, Note, An Uningenious Paradox: Intellectual Property Protectionsfor Fashion Designs, 32 COLUM. J.L. & ARTS (forthcoming 2009); Elizabeth F. Johnson, Note, Interpretingthe Scope of the Design PiracyProhibitionAct, 73 BROOK. L. REv. 729 (2008); Laura C. Marshall, Note, Catwalk Copycats: Why Congress Should Adopt a Modified Version of the Design PiracyProhibitionAct, 14 J. INTELL. PROP. L. 305 (2007); Brandon Scruggs, Comment, Should Fashion Design Be Copyrightable?,6 Nw. J. TECH. & INTELL. PROP. 122 (2007); Megan Williams, Comment, Fashioninga New Idea: How the Design Piracy ProhibitionAct Is a Reasonable Solution to the Fashion Design Problem, 10 TUL. J. TECH. & INTELL. PROP. 303 (2007). 14. See, e.g., Wilson, supra note 8 (noting the "fashion industry's ongoing debate about knockoffs"). FASHION STANFORD LA WREVIEW industry has huge economic importance. 15 Getting the economics of this industry right is an important challenge that must inform an inquiry into its regulation by intellectual property law. Third, the debate over legal protection for fashion design connects to a larger debate about how much intellectual property protection we want to have. I The question of legal protection for fashion design poses the central question ofintellectual property: the optimal balance between, on the one hand, providing an incentive to create new works, and on the other hand, promoting the two goals of making existing works available to consumers and making material available for use by subsequent innovators. We treat fashion as a laboratory to ask this question anew. The fashion trend is a particularly vivid manifestation of a general innovation pattern wherein those engaged in innovation continually seek after the new and different while, at the same time, converging with others on similar ideas. Fashion conspicuously exhibits the challenge of providing incentives for individuals to innovate while preserving the benefits to innovation of moving in a direction with others. This Article offers a new model of consumer and producer behavior derived from cultural analysis in an area where consumptive choices are also expressive. In fashion we observe simultaneously the participation in collective trends and the expression of individuality. Consumers have a taste for trendsthat is, for goods that enable them to move in step with other people. But even in fulfilling that taste, they desire goods that differentiate them from other individuals. Fashion goods tend to share a trend component, and also to have features that differentiate them from other goods within the trend. Consumption and production of fashion must be understood with respect to both the trend features and the differentiating features. Formalizing these cultural observations, we call these two coexisting tastes "flocking" and "differentiation." Fashion puts into relief people's tendency to flock while also differentiating from each other. Individual differentiation within flocking is our account of fashion behavior. But we can observe versions of this dynamic too in other areas of innovation, for example, the production and consumption of books, music, film, and other arts. Where innovation is a site of both self-expression and social expression, we can see producers and consumers of creative goods 15. See the statistics cited supra note 1. Fashion is the third-largest employer in New York, after health care and finance. Rags andRiches, ECONOMIST, Mar. 6, 2004, at 75. 16. While other analysts have associated fashion with relatively marginal or exceptional forms of creativity, such as cuisine, magic, and stand-up comedy, see Raustiala & Sprigman, supra note 11, at 1765-74 (discussing fashion as a model for understanding the work of chefs and magicians); Daniel B. Smith, Creative Vigilantes: Magicians,Chefs, and Stand-Up Comics Protect Their Creations Without the Law, BOSTON GLOBE, Dec. 23, 2007, at IE (same, for chefs, magicians, and stand-up comics), we see the dynamics of fashion innovation as exemplifying those of more paradigmatic creative industries, such as art, literature, and music. [Vol. 61:11471152 flocking to themes in common, but differentiating themselves within that flocking activity. The model makes visible an important analytic distinction that is useful for thinking about creative goods-the distinction between close copying on one hand and participation in common trends on the other hand. Design copying must be distinguished from other forms ofrelation between two designs, which may go by any number of names including inspiration, adaptation, homage, referencing, or remixing. Our analysis resists elision of close copies and myriad other activities that produce, enable, and comprise trends. Goods that are part of the same trend are not necessarily close copies or substitutes. Rather, they may be efforts to meet the need of consumers for individual differentiation within flocking. The well-known fact that "borrowing" is common in fashion,17 and might be valuable to fashion innovation, does not itself provide support for the pennissibility of close copying in fashion design. Our theory leads us to favor a legal protection against close copying of fashion designs. The proliferation of close copies of a design is not innovation-it serves flocking but not differentiation. It is importantly distinct from the proliferation of on-trend designs that share common elements, inspirations, or references but are nevertheless saliently different from each other. With respect to close copies, there is no reason to reject the standard justification for intellectual property, that permissive copying reduces incentives to create. But this effect must be distinguished from the effects of other trend-joining activities, which enable differentiation within flocking. They foster and constitute innovation in ways that close copying does not. Thus we argue in favor of a legal right that would protect original fashion designs from close copies. Some readers will no doubt bristle at the implication that Prada, say, ought to enjoy better protection for its wares. That reaction misunderstands the project. Because the current legal regime denies design protection while providing trademark and trade dress protection, the primary threat to innovation currently is not to the major fashion conglomerates. As we explain, these luxury firms are already well protected by the. existing trademark and trade dress legal regime, brand investments, and the relatively small overlap between markets for the original and for the copy. The main threat posed by copyists is to innovation by smaller, less established, independent designers who are less protected along all of these dimensions. Affording design protection would level the playing field with respect to protection from copyists and allow more such designers to enter, create, and be profitable. Relative to the current regime, we would expect the resulting distribution of innovation to feature increased differentiation and range of expression. It would also push fashion producers toward investment in design innovation and 17. Venessa Lau, Can I Borrow That? When Designer "Inspiration" Jumps the Fence to Full-On Derivation,the Critics' Claws Pop Out, W MAG., Feb. 2008, at 100 (providing examples of derivation among top designers). March 2009] FASHION 1153 STANFORD LAW REVIEW away from proliferation of brand logos by established firms making use of what legal protection is available. Fashion highlights a social dynamic to which intellectual property law inevitably attends: the relation between the individual and the collective in the production and consumption of creative work. The interplay of individuality and commonality with others poses a constant tension in innovation and its regulation. The distinction we emphasize-essentially between copying and remixing-runs through intellectual property. 18 The idea that innovation-in the form of interpretation, adaptation, and remixing-is not harmed but benefited by legal protection against close copying suggests a need to attend to this often elided conceptual distinction in conducting the debate about how much intellectual property protection we want to have, not only in fashion, but elsewhere. This Article works between two modes of analysis: law and economics, and cultural theory. We use each set of lenses together. 19 Law engages culture through a system of regulation and distribution. Economic analysis of law, for its part, endeavors to design legal regulation that induces optimal private choices, given a set of criteria about what is desirable. 20 This instrumental project can benefit from a cultural account that identifies a set of features to be optimized. The ambition here is to generate insights that deepen understanding of both culture and economics while blurring their boundaries, to clarify the goals and consequences of legal regulation. Culture-oriented readers may perceive the cultural insights here to subsume economic ones, while at the same time, economically oriented readers may perceive the economic insights to subsume culture. This is a not altogether unintended result of an approach that we might call "cultural law and economics," and on which we hope to elaborate in the future.2 1 Though our own fuller excursus on the approach is beyond the scope here, it is arguably both a new method of boundary-crossing 18. See LAWRENCE LESSIG, REMIX: MAKING ART AND COMMERCE THRIVE IN THE HYBRID ECONOMY (2008); cf Jeannie Suk, Note, Originality,115 HARV. L. REV. 1988, 1993 (2002) (exploring literary rewritings, which "revise texts that are part of our shared cultural vocabulary," and observing that "[w]hen certain texts have shaped our means of talking and thinking about important ideas, riffing on those texts in new literary works is a powerful way to refashion our language, worldview, and aesthetic"). 19. By way of comparison, the field of cultural economics applies economics to "the production, distribution and consumption of all cultural goods and services." RUTH TOWSE, Introduction to A HANDBOOK OF CULTURAL ECONOMICS I (Ruth Towse ed., 2003); cf BRUNO S. FREY, ARTS AND ECONOMICS: ANALYSIS AND CULTURAL POLICY (2000); 1 HANDBOOK OF THE ECONOMICS OF ART AND CULTURE (Victor A. Ginsburgh & David Throsby eds., 2006); JAMES HEILBRUN & CHARLES M. GRAY, THE ECONOMICS OF ART AND CULTURE (2001); RECENT DEVELOPMENTS INCULTURAL ECONOMICS (Ruth Towse ed., 2007); DAVID THROSBY, ECONOMICS AND CULTURE (2001). 20. See, e.g., LOUIS KAPLOW & STEVEN SHAVELL, FAIRNESS VERSUS WELFARE (2002). 21. Future work may offer a programmatic treatment. Cf Christine Jolls et al., A BehavioralApproach to Law andEconomics, 50 STAN. L. REV. 1471 (1998). This Article is satisfied to develop the approach through application. 1154 [Vol. 61:1147 March 2009] that demands development, and one that nuanced scholars of law, culture, and economics have engaged all along. The Article proceeds as follows: Part I begins by discussing two major theories of fashion based on status and zeitgeist, which will become important to our ensuing analysis. It then offers the key distinction between copying and trends, which we argue is necessary for accurate understanding of fashion innovation. Finally, this Part briefly discusses the normative question whether fashion is a desirable site of innovation. Part II theorizes the culture of fashion as the simultaneous operation of two phenomena that we call "differentiation" and "flocking." It models fashion consumption as the simultaneous adoption of a trend feature combined with differentiating features of a good, and explains how designers come to offer products that appeal to both differentiation and flocking at once. Part III explains the threat to innovation posed by a recent, important change in industry structure-namely, new "fast-fashion" manufacturers and retailers that engage in unregulated copying on a large scale. This Part shows how fast-fashion copyists both reduce innovation and affect its direction. In response, Part IV proposes a new intellectual property right that grows out of our analysis. The new right would protect original designs, but only from close copies. Our proposal takes an intermediate stand between permitting free copying of fashion designs and creating a broad right of exclusion. The Conclusion underscores the broad implications of the social dynamics of innovation explored here for the field ofintellectual property generally. I. WHAT ISFASHION? Fashions change. Styles emerge, become fashionable, and are eventually replaced by new fashionable styles.22 What is obvious is that the demand for new fashions is not reducible simply to material or physical needs. Though one may need a replacement pair ofjeans when an old pair gets holes from wear, or a warmer coat when the weather gets cold, for most people across the socioeconomic spectrum, the purchase of clothing is far from limited to these kinds of situations. Nearly all of us inevitably participate in fashion, even if we do not try to follow it. Fashion change is an elusive phenomenon, in need of cultural explanation. Thinkers in a range of fields have reflected on what fashion is, and in particular what accounts for fashion, the movement from introduction to adoption to decline of particular styles. We begin by discussing two principal theories of fashion that will become important in our ensuing analysis. 22. See, e.g., George B. Sproles, Analyzing Fashion Life Cycles: Principles and Perspectives,45 J. MARKETING 116, 116 (1981). FASHION STANFORD LA WREVIEW A. Status The most influential and widely held theory posits fashion as a site of struggle over social status. This is a view most concretely articulated in terms of social class at the turn of the century by Georg Simmel, the German sociologist, who was in turn influenced by Thorstein Veblen's classic work, The Theory ofthe Leisure Class.2 3 According to this view, fashion is adopted by social elites for the purpose of demarcating themselves as a group from the lower classes. The lower classes inevitably admire and emulate the upper classes. Thereupon, the upper classes flee in favor of a new fashion in a new attempt to set themselves apart collectively. This trickle-down process, moving from the highest to the lowest class, is characterized by the desire for group distinction on the part of the higher classes, and the attempt to efface external class markers through imitation on the part ofthe lower classes.24 Change in fashion is thus endlessly propelled by the drive to social stratification on the one hand and to social mobility on the other. When the magazine Vogue was founded in 1892, its first published pages presented the editorial goal as the representation of the lifestyle of New York high society, "the establishment of a dignified authentic journal of society, fashion and the ceremonial side of life.'" 25 According to a recent history of the magazine, at the turn of the century, the social context of Vogue's origin was one in which the most privileged families of New York "felt invaded by parvenus who, with little lineage but plenty of money, attempted to join in its aristocratic activities."'26 From the beginning, Vogue's representations of the fashions of the upper class were accompanied by those of the homes and parties ofprominent families, as well as articles on social etiquette.27 This feature has stayed constant throughout the last century, as Vogue has been the most visible and important U.S. publication devoted to fashion.28 The magazine exerts tremendous influence on consumers and the fashion industry,29 and continues today to feature prominently the link between 23. See VEBLEN, supranote 4; Simmel, supranote 4. 24. See GRANT MCCRACKEN, CULTURE AND CONSUMPTION 94 (Indiana Univ. Press 1990) (1988) (characterizing fashion as an upward "chase and flight" pattern rather than a trickle-down process). 25. Arthur B. Tumure, Statement, reprintedin VOGUE VOLUME I NOS. 1-28, at 16, 16 (N.Y., The Fashion Co. 1893). 26. NORBERTO ANGELET7I & ALBERTO OLIVA, IN VOGUE 2 (2006). 27. Id. 28. Id. 29. See, e.g., Xazmin Garza, The Making of Style, LAS VEGAS REV.-J., June 13, 2008, at 13CC (citing "the fashion equivalent of the bible, Vogue magazine"); Karen Thomas, "Men's Vogue" Goes for the Sophisticated Guy, USA TODAY, Aug. 24, 2005, at 2D (describing Vogue as "a 100-year-old women's fashion bible"); Emily Wax, For India's "Brand Freaks," Gucci Trumps Gandhi, WASH. POST, Feb. 11, 2008, at Al0 (reporting launch of Indian edition of "Vogue magazine, the bible ofhigh-end fashion"). 1156 [Vol. 61:1147 fashion, high society, and wealth. It functions as an arbiter of taste and style, representing fashion trends and contributing to their creation. The images of the lifestyles presented are unabashedly those of elites-wealthy socialites, celebrities, and occasionally people associated with high culture. But these images are not intended only for the wealthy. The dominant reach of Vogue depends on circulation outside of the social elite and among the many other readers. It aims at aspiring middle-class consumers as well as affluent uppermiddle-class and upper-class women. 30 Though the social class account has been criticized as too simplistic and one-dimensional,3 1 the broad influence of status is still in abundant evidence today. Fashion trends reach many consumers via observation of the ways ofthe wealthy and other high-status people. Within that project of cultural dissemination there is self-conscious openness about the trickle-down aspect of fashion trends. Fashion magazines, for example, sometimes juxtapose images of new high-priced fashion items, unaffordable by a long stretch for most ofthe readership, with pictures of similar, lower-priced items and information about where to obtain them.32 The drive of the ordinary consumer to emulate those who can afford the most expensive fashion is assumed and indeed promoted in the popular discourse of fashion. B. Zeitgeist The other major theory of fashion sometimes goes by the term "collective selection," associated with the sociologist Herbert Blumer.33 On this theory, fashion emerges from a collective process wherein many people, through their individual choices among many competing styles, come to form collective tastes that are expressed in fashion trends. The process of trend formation begins vaguely and then sharpens until a particular fashion is established.3 4 The themes of the trend reflect the spirit of the times in which we are living. 30. See MEDIAMARK RESEARCH & INTELLIGENCE GROUP, 2008 SURVEY OF THE AMERICAN CONSUMER (2008). Vogue has a circulation of 1.2 million and a total audience of 10.6 million people, and median household income of readers is $64,640. Id. Its mission statement describes the magazine as: America's cultural barometer, putting fashion in the context of the larger world we live inhow we dress, live, socialize; what we eat, listen to, watch; who leads and inspires us.... Vogue's story is the story of ... what's worth knowing and seeing, of individuality and grace, and of the steady power of earned influence. Vogue Mission Statement, reprinted in Cond6 Nast Media Kit, http://condenastmediakit.com/vog (last visited Feb. 18, 2009). 31. See, e.g., DAVIS, supranote 5; Blumer, supranote 6. 32. See, e.g., Raustiala & Sprigman, supra note 11, at 1705-11 (describing the "Splurge vs. Steal" feature ofMarie Clairemagazine). 33. Blumer, supranote 6; see also ORRIN E. KLAPP, COLLECTIVE SEARCH FOR IDENTITY (1969); LANG & LANG, supra note 4; Dwight E. Robinson, Style Changes: Cyclical, Inexorable,andForeseeable,53 HARV. Bus. REV. 121 (1975). 34. Blumer, supranote 6, at 282. March 2009] FASHION 1157 STANFORD LA W REVIEW This theory arises as a direct critique of the trickle-down theory. The driver of fashion is not necessarily imitation of high-status people per se. Rather, people follow fashion because they desire to be infashion. That is, people want to associate themselves with things that are new, innovative, and state of the art. They want to keep pace with change. If a particular fashion starts in a certain group, then other people join, not simply out of desire to emulate that group, but because being in fashion is desirable. 35 As a means of signaling and communicating about oneself, and of perceiving messages about others, 36 dress has a symbolic function and is even considered by some social theorists to be a code or a language that provides visual cues and signifiers of identity, personality, values, or other social meanings.3 7 Consumers choose among many possible options that are available in the market, and select the styles that they will wear, not merely based on their size and physical needs. They often think of their fashion choices as expressions of individuality and personal style. At the same time that the selections so operate at the individual level, they also aggregate into collective tastes. 38 Through the process of selection and aggregation of tastes, the fashion trend that emerges reflects the zeitgeist. This movement happens through individual choices, but it has a collective character that implicates society. For example, September 11 was widely thought to have affected fashion. 39 A fashion for military looks may arise when the country is at war.40 Styles-not just sales-may refer to an economic downturn.4 1 A style sported by a particular public figure may capture the zeitgeist or inspire a trend.42 35. Id. 36. See Morris B. Holbrook & Glenn Dixon, Mapping the Market for Fashion: Complementarity in ConsumerPreferences, in THE PSYCHOLOGY OF FASHION 109 (Michael R. Solomon ed., 1985); see also ERVING GOFFMAN, THE PRESENTATION OF SELF IN EVERYDAY LIFE 24 (1959). 37. See, e.g., BARTHES, supra note 5, at 59; CRANE, supranote 4. 38. Blumer, supranote 6, at 282. 39. See, e.g., Amy M. Spindler, Best of the Collections; Clothes of Quiet Inspiration, N.Y. TIMES, Jan. 20, 2002, at E37 (interpreting some designers' collections after September 11 as suggesting American iconography); Guy Trebay, Waitingfor Takeoff: Designers Offer a Peek of Spring, N.Y. TIMES, Sept. 10, 2002, at Bi 1 ("Many American designers, in the season shown after 9/11 ...were moved to express.., the anxiety that had crept into most comers of American life."). 40. See, e.g., Cathy Horyn, Macho America Storms Europe's Runways, N.Y. TIMES, July 3, 2003, at Al (detailing the prevalence of such Iraq War-inspired fashion as "an image that symbolized the virile Texas cowboy in boots and broad hat" and "battle jackets and cartridge belts fashioned from banker's broadcloth" on the runways ofMilan). 41. See, e.g., David Colman, When Fashion Goes for Broke, N.Y. TIMES, Sept. 3, 2008, at G6 ("'Whenever the economy gets tough, fashion responds by playing it safe,' said Jim Moore, the creative director of GQ ....); Eric Wilson, Combatingthe Gloom? Child's Play, N.Y. TIMES, Oct. 23, 2008, at E4 (interpreting a particular trend in 2008 as designers' efforts to "cope with the consumer gloom in the only way they know-that is, by channeling the mind-set of their inner children. It may be just a coincidence, but children's books and a [Vol. 61:11471158 March 2009] The symbolic function of fashion depends on the interplay of individual and social meanings. Fashion features the tension between the desire to be distinct as an individual and the desire to connect with a collectivity. Another way of saying this is that the fashion process imposes social constraints and parameters within which individual choices of communication and expression are shaped and directed. Fashion is then driven forward as a combination of individual differentiation and collective identification, and of the personal and the social impulses. Without necessarily denying the importance of status or imitation in the explanation of fashion trends, what we are calling the zeitgeist theory is in effect a critique of a status account in which fashion trends essentially consist of imitation of high-status people. The zeitgeist theory views trends as the collective aggregation of individual choices throughout society. These choices, which are both expressive and consumptive, converge on themes that reflect the milieu and social context of the times. C. Copies Versus Trends In each ofthese theories, consumers desire, and producers provide, articles that are on trend. Some observers assume that the trendy articles are copies: either the exact same article purchased from the same producer, or else a close copy of most elements of the original's design. But such copies play only a limited role in the rise and fall of trends. Participation in a trend-by a consumer or a designer-does not necessarily or usually entail copying. First, one individual may seek to imitate another-as the status theory suggests-but without necessarily copying her dress. One can imitate another's style by consciously or unconsciously being influenced to wear clothes in that style. Copying is a more literal and direct process in which one targets the original for replication. For example, a consumer can imitate the length of a color palette by Crayola have emerged as a pop cultural theme in art and fashion with surprising alacrity, as if in anticipation of a need for more simplistic comforts"); see also Suzy Menkes, Bulls, Bears and the Bellwether Hemline, N.Y. TIMES, July 17, 2008, http://www.nytimes.conV2008/07/17/fashion/] 5jolie.html (published online) (discussing the history of fashion's response to recession, focusing on plummeting hemlines). 42. See, e.g., Teri Agins, Over-40 Finds a Muse: Designersfor the Middle-Aged Pin Hopes on Mrs. Obama, WALL ST. J., Dec. 6,2008, at W4 (reporting on Michelle Obama's influence on fashion and quoting a magazine editor describing her as "represent[ing] the post-feminist generation-a woman who can wear a sheath dress and show her arms-and women are responding to her ability to be feminine, sexy and still powerful."); Ray A. Smith, Pulling Off the ObamaLook, WALL ST. J.,June 9, 2007, at P1 ("With the suit-andno-tie look gaining prominence lately-presidential hopeful Barack Obama has drawn attention for sporting a version of the approach, and Microsoft's Steve Ballmer and Boeing CEO Jim McNerney have done it, too-more men are trying it out themselves."); Eric Wilson, Merrily They Dress, N.Y. TIMES, Nov. 20, 2008, at El ("Ever since the Obamas appeared on election night as a coordinated fashion tableau, as if they had just stepped out of a holiday greeting card portrait, sales of red dresses have been terrific, said Kay Unger, who makes party frocks."). FASHION 1159 STANFORD LA WREVIEW skirt without necessarily purchasing a copy of that skirt. Copying, in other words, is only a subset of a wide range of imitative practices. Second, consumers may join trends without an imitative motive. The zeitgeist theory emphasizes not imitation, but rather an individual's distinct desire to be in fashion. People can want to be in fashion without necessarily having as their object the emulation of the lifestyle, values, or status associated with a particular group that first sported the style. They may instead-or alsoseek to join a collective moment. Such convergence does not require a copy of what others are wearing. Third, designers may furnish on-trend articles without closely copying one another. Instead, they may engage in interpretation, or "referencing. 4 3 They may quote, comment upon, and refer to prior work. 4 Unlike much close copying, such interpretation does not pass off the work as the work that is being copied. Instead, it marks awareness of the difference between the two works as it looks to the prior work as a source of influence, or even a precursor. Even where the influence is not completely conscious or direct, the latter work draws on the meaning of the earlier work, rather than being simply a copy of it. For example, the look of a Chanel knit jacket has been interpreted repeatedly in other designers' styles, so that it has become a classic style drawing on the spirit of the look without purporting to be a Chanel product. Another Chanel classic, the quilted handbag, has been similarly reinterpreted. This practice, by which designers draw freely upon ideas, themes, and styles available in the general culture, and refer back to others' prior designs, has led to the widespread but incorrect view that there is no real originality in fashion design.4 5 This view is no more correct than the analogous complaint about music: that homage and pastiche somehow deny any claim of originality to new works. The important point is that interpretations are different from copies in their goals and effects. Close copies can substitute for and reduce the value of the original, thereby reducing the incentive to create, to a greater extent. Rather than being substitutes, interpretations may even be complements for other on-trend articles. 46 A status theory of fashion might lend to the view that trend-joining is essentially copying. Accordingly, the fashion trend rises as a form of 43. See Raustiala & Sprigman, supra note 11, at 1700 ("reference"); id. at 1728 ("referencing"). 44. For example, Proenza Schouler's spring 2008 collection was widely understood to draw upon the previous work of Balenciaga designer Nicolas Ghesquiere. Lau, supra note 17. There are many such examples every season. Id. 45. See, e.g., Amy Kover, That Looks Familiar.Didn'tI Design It?, N.Y. TIMES, June 19, 2005, § 3, at 34 ("Mr. Schwartz of A.B.S. has some advice for newcomers: Stop whining. 'When you are talking about fashion, lose the word original,' he said. 'Ask the small designers where they got their inspiration. They pull their inspiration from others. It's in the air. You don't sit by the window and wait for it to materialize."'). 46. For further discussion of complementarity, see infra Part II.B. For further discussion ofsubstitution, see infra Part III.B. 1160 [Vol. 61:1147 March 2009] emulation, and then declines when elites or early adopters feel the need to distinguish themselves from the copying masses and adopt a new style as a means to do so. If one thus equates trend-joining with copying, then one might reasonably conclude that fashion is driven by copying. 47 But it is important to see that status does not exhaust the motivations for fashion. Under a zeitgeist theory, fashion is not just imitation of elites or early adopters, and is not reducible to copying. Fashion choices are expressions of individuality that combine into collective tastes. Fashion reflects the desire for the new, for movement with the collectivity, for contact with the spirit of the times. This theory leads us to disaggregate fashion trends from copying, and see that fashion moves not necessarily as the result of a market's saturation with copies. Copies may play a role in fashion change, but they are not the engine without which innovation in fashion would slow and stagnate. D. Why PromoteInnovationin Fashion? Before further developing and applying these distinctions between copying and trends, we first pause with readers who may wonder whether fashion is worth promoting. After all, one might well agree with our account of the features of fashion, but consider fashion innovation to be undesirable. Everyone takes part in apparel fashion on some level. Everyone inevitably expresses themselves through the clothes they wear (even if to communicate that they are too serious to care about fashion). But some may consider fashion frivolous or wasteful. They may believe that we would be better off if fashion did not exist and if clothing were used only for the literal purpose of covering the body or keeping warm. This set ofintuitions lies behind the Anglo-American and European history of sumptuary laws, which, until the eighteenth century, purported to limit the expenditures people could make on clothing, to protect against the vice of wasteful spending for personal appearance and ostentatious display, including for purposes of following fashions. 48 Moral disapproval of expenditure on fashion is traditional. Normative regulation of fashion goes back to the Greeks and the Bible.49 The moral stance found, albeit incompletely enforced, in many 47. See Raustiala & Sprigman, supranote 11. 48. See ALAN HUNT, GOVERNANCE OF THE CONSUMING PASSIONS: A HISTORY OF SUMPTUARY LAW (1996). 49. Solon, the legendary lawgiver of ancient Athens, created some of the first sumptuary laws, regulating conspicuous consumption at funerals-including how many shawls a widow could wear. See Anne Theodore Briggs, Hung Out To Dry:Clothing Design ProtectionPitfallsin United States Law, 24 HASTINGS COMM. & ENT. L.J. 169, 204 (2002). Deuteronomy 22:5 says that "[tihe woman shall not wear that which pertaineth unto a man, neither shall a man put on a woman's garment: for all that do so are abomination unto the Lord thy God." 1 Corinthians11 sets out guidelines about head covering while praying. FASHION STANFORD LAW REVIEW religious traditions from Christianity to Buddhism, rejects luxury spending on garments and promotes plain garb.5 Another reason for looking askance at fashion may be concern about visible markers of status hierarchy. Many historical sumptuary laws actually imposed hierarchical dress codes, granting privileges to wear certain garments to the upper class or prohibiting the lower class from wearing certain garments.5 1 Perhaps fashion is normatively undesirable because it is a way in which class and wealth disparities can easily be shown. Chairman Mao, in the pursuit of egalitarianism and Marxist rejection of surplus value, dictated that a billion people should wear an identical unadorned outfit, and for some decades they did so, 52 notwithstanding China's rich history of fashion and its contemporary unabashed re-embrace of consumer capitalism. 53 With respect to the morality of expenditures or the issue of wastefulness, for the purposes of this Article, we treat fashion consumption the same way we would ordinarily treat the consumption of other nonharmful goods that have creative and expressive components, such as books, music, films, and art. (To varying degrees, fashion is present in those areas as well. For example, there may be a trend of memoirs about addiction, films about Iraq, biographies of presidents, or novels about ancient biblical secrets.) It is difficult to see how the argument about wastefulness or immorality of spending on a coveted suit or dress would be different in kind from paying a sum for a work by a highly regarded painter. We assume that if consumers are prepared to pay for fashion in its various forms, regulation ought to be set to promote innovation and allow consumers a variety ofoptions. 54 Some readers may resist this set of assumptions in various ways. First, the idea that the measure of the value of fashion is akin to the measure of the value of books, music, and art may strike some as absurd.55 Even though fashion is not widely regarded as one of the "fine arts," it is undeniably a creative good that has expressive features. It is no more logical to denigrate the value fashion choices confer upon consumers than to denigrate the value of the best-selling thriller many are reading or the hit song many are listening to. We may of 50. Well-known examples include the highly regulated attire among the Puritans, the Amish, Catholic nuns, Buddhist monks, and Ultra-Orthodox Jews. 51. See HUNT, supranote 48, at 172. 52. See, e.g., PATRICIA BUCKLEY EBREY, THE CAMBRIDGE ILLUSTRATED HISTORY OF CHINA 294 (1996) (noting the Communist Party's early efforts to rid Chinese cities "of what they saw as decadence-flashy clothes and provocative hairstyles"). 53. For detailed discussion of China's ancient and complex history with issues of intellectual property, see generally WILLIAM P. ALFORD, To STEAL A BOOK IS AN ELEGANT OFFENSE: INTELLECTUAL PROPERTY LAW INCHINESE CIVILIZATION (1995). 54. This is a common assumption in economic models about fashion. Gene M. Grossman & Carl Shapiro, Foreign Counterfeitingof Status Goods, 103 Q.J. ECON. 79, 89 (1988). 55. This has been a strong intuition of some colleagues with whom we have discussed this project. [Vol. 61:11471162 course engage in value judgments about, say, the artistic value of Grisham relative to Proust, of pop music relative to Bach-and of fashion relative to literature and music. But that kind of hierarchical value distinction among cultural products is not to be confused with the notion of value on which we rely here. The choice to purchase these goods is, on our welfare account, evidence of value, and that is unrelated to the quality or merits of particular cultural products or genres of cultural production. Indeed it is the only evidence that can be measured, short of a separate normative assessment of whether people are wise to desire the things they do. Here we assume the desirability of investments in creative goods and in fashion as a creative good. Second, some may view fashion consumption as a product of social pressure (i.e., to look "cool" or at least not to look like a "dork"), 56 and therefore unable to confer meaningful welfare gains on its consumers. Participation in fashion seems to be freely chosen by consumers. For the purposes of this Article, we assume that, especially when it comes to economic choices that are not necessary for human survival, adults' decisions may be construed as voluntary and therefore as a desirable pursuit of their life plans. Finally, there may be concerns about negative positional externalities of fashion. These concerns pertain to status signals generated by fashion as a means of displaying wealth or other markers of status. For example, if fashion serves to distinguish some from others,57 the satisfaction some people receive from signaling their high status through fashion may be offset by the disutility of others. On this view, participation in fashion trends is spending to reduce that disutility. This expenditure is wasteful. It would be better if nobody spent in this way. Accordingly, if fashion were eliminated (A la Mao, or school uniforms), social welfare would improve; increasing fashion innovation cannot be seen as a gain in welfare. This is a plausible view of negative externalities that corresponds to a theory of fashion as driven by status. But if the centrality of such status seeking is displaced with what we have been calling the zeitgeist theory, the status signal is not the dominant aspect of fashion. As we have explained above, the desire to be "in fashion" involves more than signals about status. It is a means of individual expression through which people partake in collective movement and the spirit of the times. Fashion enables this expressive process, and as such has benefits much like those associated with other consumptive goods that are also expressive. Signals of status are undeniably present in all these goods (just think of the high-end art market, high-brow literary fiction, or opera performance), but so too-and more importantly-are means of expression. Our view that innovation in fashion is socially desirable rests on assumptions 56. Cf., e.g., Vanessa O'Connell, FashionBulliesAttack-In Middle School, WALL ST. J., Oct. 25, 2007, at DI ("Teen and adolescent girls have long used fashion as a social weapon.... But today, guidance counselors and psychologists say, fashion bullying is reaching a new level of intensity as more designers launch collections targeted at kids."). 57. See supra Part I.A. March 2009] FASHION 1163 STANFORD LAW REVIEW that are shared with the assumption that in general the creation of new novels and new songs is socially desirable. II. A MODEL OF TREND ADOPTION AND PRODUCTION This Part reflects on fashion as a cultural phenomenon and identifies features of fashion that may be engaged by legal regulation. The aim is to distill the features we have discussed above under the rubric of the zeitgeist theory, which points us to two conditions that exist simultaneously and in relation to one another. We call them "differentiation" and "flocking." A. DifferentiationandFlocking Through fashion, people communicate and express themselves. Fashionable individuals' personal style is often described as "unique" or "inimitable." 58 If consumers use fashion to express themselves as distinctive individuals, then it is valuable to have available a large range of different identifiers. Fashion goods provide a vocabulary. What consumers might value in fashion then is the availability of a variety of goods to choose from, a proliferation of the number of meanings that can be made. The availability of a variety of different goods enlarges the vocabulary and the meanings that can be communicated. If consumers have a taste for differentiation of identity through fashion, then individual differentiation becomes an identifiable desired feature, for the purpose of intellectual property regulation of fashion. We posit that "differentiation" is a key feature ofthe consumption and production offashion. But fashion would not be fashion were it not for its basically collective character. Even as individuals strive to differentiate themselves through fashion choices, fashion is a means of participating in group movement. We call this "flocking." Consumers tend to engage in flocking in buying new clothes, not because they need them, but because their existing clothes seem outdated. They want to be "in fashion." Flocking among consumers is again not necessarily a function of imitation or copying of any particular groups or individuals, though it may be. It can be a manifestation of a desire to partake of the collective moment, to be in step with society, or to be in touch with the present. It may be pleasurable for people to move in a collective direction, joined by others in expressive 58. See, e.g., Arienne Thompson & Erin O'Neill, Brotherly Style Sense, USA TODAY, Aug. 12, 2008, at B14 ("The Jonas Brothers may be burning up the music charts, but their unique sense of style is also getting them noticed in the fashion world."); Bruce Weber, DianeKeaton Reflects on Keeping 'Em Laughing,N.Y. TIMES, Mar. 17, 2004, at El ("Her famously unique wardrobe (for the interview she wore a black business suit, jacket and skirt, over a pair of blue jeans) is the fashion equivalent ofphilosophical Berra one-liners."). [Vol. 61:11471164 endeavor. There may be pleasure in a convergence, in participation in similar themes and ideas that reflect the times that all are experiencing. Fashion is simultaneously characterized by differentiation and flockingtwo phenomena that might appear to be in tension. On the one hand, the expressive and communicative aspects of fashion choices seem to benefit from a distribution of innovation that produces goods that are differentiated from each other. Thus we identify differentiation as a desired goal in fashion. On the other hand, we also notice benefits of moving in a common direction and partaking of the same trend. Thus we also identify flocking as desirable. The idea is well captured by Anna Wintour, editor of Vogue, who noted that what is laudable in fashionable People is at once "looking on-trend and beyond trend and totally themselves." Our theory then is that in fashion we observe the interaction of the tastes for differentiation and for flocking, or differentiation within flocking. The relation between differentiation and flocking is the key dynamic. People want to engage in flocking in a way that allows individual differentiation within it. They want to be part of a trend, but not be a replica of others who also join the trend. It would not be fashion if only flocking behavior were present. A world in which exactly one design of suit exists, due to demand or fiat, could be said to have apparel but not fashion. Nor would it be fashion if only individual differentiation were present. A world in which no collective patterns could be discerned could not be said to have fashion either. Fashion consists of both human desires, to flock and to differentiate, in relation to each other. It might be feasible to posit a more exact relationship between differentiation and flocking-for example, to specify a utility function that captures the relationship between the two preferences. One source of complexity is that tastes for differentiation and flocking will vary across consumers. One could in theory posit a person at one extreme who overwhelmingly values differentiation and thus avoids trendiness or any similarity to what others are doing. One could also posit a person at the other extreme who wishes to appear exactly the same as others. But the key point of the differentiation-flocking model is that the tastes of consumers are not at these particular extremes but rather express measures of both differentiation and flocking. The precise relationship between the two varies with the consumer, or even for the same consumer under different circumstances. For example, the same person might favor conservative suits (flocking) and extreme neckties (differentiation). The relationship between differentiation and flocking can also vary with the particular fashion trend or the particular item of fashion. 60 Furthermore, a consumer's utility from a particular configuration of differentiation and flocking may depend on how much differentiation versus flocking others are engaging in. Much complexity accompanies the attempt to 59. Anna Wintour, Editor'sLetter,VOGUE, Aug. 2008, at 70. 60. See, for example, our comparison of handbags and apparel in Part I.B. March 2009] FASHION 1165 STANFORD LAW REVIEW pin down the exact relationship. A key element of our theory is that the tastes for differentiation and flocking exist together in a dynamic relationship. Finally, notice that the relation between flocking and differentiation maps on to the relation between copying and innovation. Just as direct reproduction ofan existing novel is not innovation, if fashion were all about producing exact copies of existing articles, it would not be a practice of innovation.The impulse to flock in fashion is expressed in the aspects of fashion that draw on and sometimes copy existing works, but what makes the field a creative endeavor is the drive to differentiate-to reinterpret, change, remix, and transform, and as such, resist the sheer replication of existing works even while incorporating them. That is the creative impulse. In other words, differentiation constitutes innovation in fashion. Without the differentiation component, fashion would not be a form of innovation. Our favoring of differentiation in fashion then is an outgrowth of our assumption of the theory of incentives underlying intellectual property law about the effects ofcopying on creators' incentives. B. TrendAdoption The process of trend adoption reflects differentiation and flocking. Think of a fashion item as having two kinds of attributes, a trend feature (around which consumers flock) and various differentiating features. The trend feature is some shared, recognizable design element such as a wrap dress, a fitted fringed jacket, a driving shoe, or a floral print.6 1 The differentiating features are all design elements other than the trend feature that make the items within the trend nevertheless different from each other. Consumers are able to identify a trend feature, factoring it out from the other features. Their recognition process may be simple-seeing many items with the trend feature in stores or on the street-or it may be enhanced by advertising or magazine articles that identify the trend feature. Many consumers prefer new items that are part of a trend. A consumer does not care solely about the presence of a trend, however. In addition, the consumer has a taste for differentiation in the article's other features, and preferences that vary according to body shape, aesthetics, or personal style. Fashion-conscious people generally do not seek to wear precisely the same outfit as someone else.62 Rather the consumer seeks goods that contain the trend feature but are differentiated. 61. Bright florals were a trend for spring 2008. Hilary Alexander, ParisRound-up, DAILY TELEGRAPH, Oct. 8, 2007, at 20 (noting floral theme across many shows, with the specific implementation varying greatly). 62. See, e.g., Amy Odell, Internet Saves Inaugural-BallAttendees from Wearing the Same Dress,N.Y. MAG., Jan. 2, 2009 (describing a new website, DressRegistry.com, "that allows women to register the dresses they're wearing to big events like the inaugural balls so they don't end up wearing the same thing as someone else"). The social anxiety that attaches to this phenomenon has, for decades, been a recurring target of popular parody. See, e.g., I Love Lucy: Lucy andEthelBuy the Same Dress (CBS television broadcast Oct. 19, 1953). [Vol. 61:11471166 How does a trend catch on? Suppose designers in one season produce, say, an unusually large number of designs with floral prints. Consumers recognize the floral print as the feature that is part of the potential trend, by seeing the prints in stores and on other consumers. The trend takes off, provided that enough consumers conclude two things-first, that enough other people are buying items with the trend feature that a trend will occur; second, that the consumer's idiosyncratic preferences are well-enough served by a particular item that the consumer buys it. To take off, the trend must offer something sufficiently new. After all, new clothing is not an essential good in this context, and the new trend is competing with a closet full of existing clothes. Put differently, a new trend exhibits a network effect in consumption: individuals buy if enough others are buying or can be expected to buy-for example, because articles with the same trend feature appear in many shops at the same time. If multiple vendors offer the same new trend element at the same time, together with the differentiating details also necessary to satisfy consumer demand for differentiation, this is more likely to produce a successful new trend. Consumers, ever on the lookout for something new, identify a new trend feature, not much present in the previous season's items, as a fresh basis for asserting commonality. The feature could be as simple as the introduction of a loose fit in jeans after a period when skinny jeans were everywhere. But among the looser jeans available there can be a nearly infinite variety of combinations of cut, color, fabric, texture, wash, and rise. Our flocking-differentiation model is distinct from some status models of trend adoption in which a fashion good is a repository of status, and individuals who purchase goods convey their status by displaying the item.63 A high-end "it" bag is the paradigmatic case. As a particular handbag obtains "it" status, for example, there might develop a long waitlist for the desired bags, which are sparingly doled out by stores, with priority given to customers of high status.64 Even outside of the narrow band of "it" bags, high-end designer handbags often have status-conveying functions. When a high-end designer bag becomes trendy, many want precisely the same bag, making it a particularly good exemplar ofthe status model. If the status model applies best to a subset of designer handbags, the present flocking-differentiation model better captures consumers' attitudes toward apparel, where consumers seek to be on trend but also have a taste for differentiation. Thus, arguments made in favor of permitting counterfeit bags, 63. See, e.g., Raustiala & Sprigman, supra note 11, at 1718 (basing the "induced obsolescence" model on the proposition that "[c]lothing is a status-conferring good"). 64. See, e.g., MICHAEL TONELLO, BRINGING HOME THE BIRKIN: My LIFE IN HOT PURSUIT OF THE WORLD'S MOST COVETED HANDBAG (2008) (describing one man's effort to circumvent the legendary waiting list for a Birkin bag); cf. Sex and the City: Coulda, Woulda, Shoulda (HBO television broadcast Aug. 5, 2001) (showcasing a New York fashionista's desperate attempt to secure a Birkin bag of her own-and the comic humiliation that ensued). March 2009] FASHION 1167 STANFORD LAW REVIEW so as to thwart the ability of wealthy consumers to convey status through them, do not apply in precisely the same way to apparel. The foregoing suggests three preconditions for the success of a trend. First, the new trend feature must be sufficiently uncommon among previously available articles. Second, the new trend feature must be sufficiently prevalent. And third, there must be a sufficient differentiation of items that contain the trend feature so as to satisfy demand for differentiation and help to achieve a critical mass of consumers. C. Trend Production Designers, too, engage in a process of differentiation and flocking. In any given season, they flock to similar hemlines, dress shapes, and tailoring. They converge on similar or related styles and themes. Yet the precise result reached by each producer is different. 65 Flocking results, in part, from shared influences. If images of war fill the news, military-inspired styles may enter multiple collections. 66 If a celebrity or a new film gains acclaim for a distinctive style, that style may be incorporated into the work of several different designers. 67 Forecasting services furnish a common input to some designers, particularly the followers. 68 Designers and other personnel move from fashion house to fashion house, making their imprint on multiple brands.69 Common pressures in the real world-women's entry into the professional workplace in unprecedented numbers, for example-can lead to a "convergent evolution" of independently derived, parallel innovation. 70 New technological possibilities, such as a novel fabric, can produce commonalities in collections as well.71 65. This is shown in the "runway reports" offered by fashion magazines. See generally, e.g., Runway Report: Fall'sNew Looks, HARPER'S BAZAAR, June 1,2008, at 182 (assembling trends from fall collection in a special edition of magazine). 66. See, e.g., Horyn, supranote 40. 67. See, e.g., Ruth La Ferla, Forget Gossip, Girl, the Buzz Is About the Clothes,N.Y. TIMES, July 8, 2008, at Al (describing the "'Gossip Girl' influence" on designer collections). 68. See, e.g., Vanessa O'Connell, How FashionMakes Its Wayfrom the Runway to the Rack, WALL ST. J., Feb. 8, 2007, at D1 (describing use of such services by J.C. Penney and others). 69. See, e.g., Lau, supra note 17 (collecting examples of designers and consultants whose moves-between Helmut Lang and Calvin Klein, Mami and Chlo6, and Tom Ford and Burberry Prorsum-contributed to a shared style at each pair of firms). 70. Convergent evolution is "the recurrent tendency of biological organization to arrive at the same 'solution' to a particular 'need."' SIMON CONWAY MORRIS, LIFE'S SOLUTION: INEVITABLE HUMANS IN A LONELY UNIVERSE, at xii (2003). 71. See, e.g., Michele Loyer, Brave New World of "Techno" Fabric,INT'L HERALD TRIB., Oct. 11, 1996, at 24 ("Two years ago, fashion designers like Calvin Klein, Donna Karan and Giorgio Armani started using technical fabrics, until then restricted to industrial use (fire-proofing) or motorcycling, in their sportswear lines."); Heesun Wee, Spandex Market Expected to Stretch, GLOBE & MAIL (Toronto), Oct. 13, 1999, at B7 (describing incorporation of Lycra and similar materials, once limited to athletic attire, in street-wear 1168 [Vol. 61:1147 Flocking also results from mutual influences and inspiration among designers. They and their assistants attend fabric and other trade shows, where they learn from suppliers what other designers have planned-sometimes with the suppliers' active encouragement. 72 Stylists, magazine editors, and buyers travel from designer to designer, cross-pollinating as they move.73 The shows are not quite simultaneous, extending across several weeks and cities, and lastminute tinkering can incorporate the influence of designers who have had earlier shows. These shared influences promote convergence around a trend, but not identical articles. For one thing, the shared influences are usually too general to produce identical articles. Moreover, each producer has substantial incentives to produce a differentiated product. A producer, faced with differentiated demand, will tend to seek out a differentiated niche to satisfy, rather than occupy the exact same space as another producer.74 Some producers are better suited for some niches than others-they may understand one segment of the market (teenagers, say, or Californians) better than another, and focus accordingly. Offering an on-trend, distinctive good may be a source of benefit to some producers, since it offers the opportunity to work with and be in communication with others on a similar problem.75 And choosing a differentiated product, rather than the exact same good offered by another producer, raises the probability that a trend supported by differentiation within flocking will get off the ground in the first place. 76 The differentiation-flocking model of production, like that of consumption, has limits. It may not apply to "it" handbags, for example. Where consumers are uninterested in differentiation-where they do not even have idiosyncratic physical needs (due to body shape or coloration), but simply want the status signaled by the item-the model may not apply. There may be apparel items collections). 72. According to one insider, "fabric salesmen have only to whisper, 'let me show you the fabrics that Saint Laurent is ordering,' and the stampede is on." Teri Agins, Copy Shops: FashionKnockoffs Hit Stores Before Originalsas Designers Seethe, WALL ST. J., Aug. 8, 1994, at Al. As the same piece explains, "[p]erhaps fake fur [an important trend one season] was merely 'in the air,' as designers like to say when such coincidences occur. But most of them can sense which way the fashion winds are blowing by attending the big textile shows held each year in Paris and Milan." Id.; see also Jonathan M. Barnett et al., The Fashion Lottery: CooperativeInnovation in Stochastic Markets 31-35 (USC Ctr. in Law, Econ. and Org., Working Paper No. C08-17, 2008), available at http://ssm.com/abstract=1241005 (emphasizing the importance of trade shows as a communication tool). 73. Christina Binkley, Runway to Rack: Finding Looks That Will Sell, WALL ST. J., Mar. 6, 2008, at Dl (noting that most sales come from pre-collections sold prior to the runway shows). 74. See, e.g., Harold Hotelling, Stabilityin Competition,39 ECON. J. 41 (1929). 75. Cf YOCHAI BENKLER, THE WEALTH OF NETWORKS: How SOCIAL PRODUCTION TRANSFORMS MARKETS AND FREEDOM 91-99 (2006) (discussing nonmonetary motivations for social production). 76. For a different model that also predicts similar but differentiated products, see Barnett et al., supranote 72, at 31 (characterizing imitation as a form of insurance). March 2009] FASHION 1169 STANFORD LAW REVIEW that are like "it" handbags. But for most apparel, there are idiosyncratic preferences, and a taste for differentiation, that make the differentiationflocking model applicable. The economic imperative to both differentiate and flock resembles the innovative production of more technologically intensive goods. Similar limited cooperation takes place in the development of a new computer operating system or DVD player, in which producers jointly struggle to get a new "standard" or "platform" off the ground. 7 There, too, it is a variety of differentiated products-"launch titles"-that contribute to the success of the shared feature by providing confidence about sufficient adoption that the platform will be a success. III. How UNREGULATED COPYING THREATENS INNOVATION Our model of trend production has two key features: a trend component that is shared by market players, and a second, differentiating component that varies for each designer. The model explains how producers collectively produce a trend: the common component fosters the sale of a diverse array of new, on-trend goods, which meet consumers' simultaneous desire for on-trend and differentiated goods. A recent, important change in industry structure-"fast-fashion" manufacturers and retailers-threatens innovation in fashion. In this Part, we explain what is new about fast fashion and why it matters. We distinguish two types of fast-fashion firms, designers and copyists, and their disparate roles. Fast-fashion designers challenge but also enhance the fashion innovation process. Fast-fashion copying, by contrast, threatens the amount of innovation and pulls the direction of innovation toward fashion's status conferral aspects and away from its expressive aspects. A. Fast-FashionCopyists Copying in fashion is not a new problem. U.S. designers in the early twentieth century-and, before that, French couturiers-were plagued by competitors who made sketches at shows or measured the seams of procured originals to discern their patterns, and then used local labor to make the copies. 78 Often, these copies could be accomplished quickly, and the copies reached the market before the original.79 77. See Timothy F. Bresnahan & Shane Greenstein, Technological Competition and the Structure ofthe Computer Industry, 47 J.INDUS. ECON. 1 (1999). 78. Sara B. Marcketti & Jean L. Parsons, Design Piracy and Self-Regulation: The Fashion Originators' Guild ofAmerica, 1932-1941, 24 CLOTHING & TEXTILES REs. J. 214, 215-17 (2006); Mary Lynn Stewart, Copying and Copyrighting Haute Couture: Democratizing Fashion, 1900-1930s, 28 FRENCH HIST. STUD. 103, 108-13 (2005). 79. Stewart, supra note 78, at 108-09. 1170 [Vol. 61:1147 March 2009] What has changed is not the fact or speed of copying, but the large scale and low cost at which rapid copies can be made. (For comparison, just think of music, where rapid copying has long been feasible, while large-scale, low-cost rapid copying is a new phenomenon.) Today, a pattern can be based upon an Internet broadcast of the runway show and transmitted electronically to a lowcost contract manufacturer overseas. 80 A gradual easing in import quotas, begun in 1995, 8 1 has increased scale and thereby lowered overseas manufacturing costs. Electronic communications and express shipping ensure that prototypes and finished articles can be brought to market quickly. As a result, thousands of inexpensive conies of a new design can be produced, from start to finish, in six weeks or less. 83 The most striking consequence of low-cost, high-scale, rapid copying is not in beating an original to market, but in the ability to wait and see which designs succeed, and copy only those. Copyists can choose a target after retailers have made their buying decisions, or even after the product reaches stores, and customers have begun to buy. 83 Such copyists can reach market well before the relevant trend has ended. 80. See, e.g., Kover, supra note 45 ("Large discounters like Target and H&M have signed major designers and can deliver fashionable clothing at cheap prices by manufacturing in countries like India and China and flying clothes to stores in the West. Computer systems can track inventories and replace sold-out items within a few days."); Fashion TV, http://www.ftv.com (last visited Jan. 31, 2009) (telecasting runway shows live); Fashion Week Daily Runway, http://www.fashionweekdaily.com/runway (last visited Jan. 31, 2009) (providing photographs of collections); New York Magazine, Fashion, http://video.nymag.com (last visited Oct. 4, 2008) (providing video ofrunway shows). 81. The 1994 Agreement on Textiles and Clothing, part of the Uruguay Round of world trade negotiations, dismantled quotas imposed by an earlier agreement, the Multifibre Arrangement of 1974. The Agreement removed some quotas immediately, and subjected the rest to a ten-year phaseout. World Trade Organization, A Summary of the Final Act of the Uruguay Round ("Agreement on Textiles and Clothing"), http://www.wto.org/english/ docs.e/legal-e/ursum e.htm#cAgreement (last visited Feb. 18, 2009). The phaseout is limited by "safeguards" that permit importers to temporarily limit the increase in quotas. Id. 82. One copyist, Forever 21, needs six weeks. Ruth La Ferla, FasterFashion,Cheaper Chic,N.Y. TIMES, May 10, 2007, at GI [hereinafter La Ferla, FasterFashion].Oscar knockoff dresses take two to four weeks to reach consumers, as of 2006, compared to twice that time just five years before. Ruth La Ferla, Night ofa ThousandKnockoffs, N.Y. TIMES, Mar. 9, 2006, at GI1 [hereinafter La Ferla, ThousandKnockoffs]. Raustiala and Sprigman argue that little has changed-that for the past twenty-five years, copying has been "easy and fast," and the increase in speed over that period "does not appear large." See Raustiala & Sprigman, supra note 11, at 1759-60. Elsewhere, however, they acknowledge a variety of factors that, in their view, increase copying speed, including "[d]igital photography, digital design platforms, the Internet, global outsourcing of manufacture, more flexible manufacturing technologies, and lower textile tariffs." Id. at 1714-15. In our view, the factors they identify are directed not only to speed-indeed, some (such as tariffs) likely have no effect on speed-but to greater scale, lower costs, and higher quality ofrapid copying. 83. Cf Agins, supra note 72 ("The brisk market in ideas has even given rise to the 'knockoff consultant.' . . . Carole Ledesma and Nathalie Jonqua . . .pose as ordinary shoppers while scouting boutiques in London, Paris and Milan. Each month, they mail 100 FASHION STANFORD LA W REVIEW Retailers and manufacturers exploit the resulting opportunity. They sell copies at a discount to the original-necessarily, given the lower quality 84 but earn a profit thanks to lower unit costs and the avoided expense of design. 85 The most notorious copyist retailer is Forever 2 1,86 though copying also extends to a wide range of department stores and specialty clothing retailers. 87 The retailers are supplied by manufacturers who, for the most part, remain anonymous. An exception is A.B.S., a prominent copyist of dresses worn to the Oscars awards ceremony and other red-carpet events. 8 8 The Appendix contains two representative examples of close copying by Forever 21. Copying is not a necessary element of the fast-fashion business model. Even retailers that sell copies do not sell only copies. And some fast-fashion firms eschew exact or close copies. For example, the two leading fast-fashion firms, Zara and H&M, avoid close copying.89Although Zara and H&M may have become conflated with Forever 21 in the public mind,90 their strategies photos of the hottest designs to 55 American clients."). 84. Fast-fashion products, it is said, are made to be worn just ten times. Pankaj Ghemawat & Jos6 Luis Nueno, ZARA: Fast Fashion 13 (HBS Case Study 9-703-497, rev. Dec. 2006) (making this point about Zara, a fast-fashion designer). 85. Cf Design Piracy Act Could Hit China, WOMEN'S WEAR DAILY, Oct. 29, 2007 (suggesting that China would be affected by the proposed Design Piracy Prohibition Act because "fast-fashion companies ... stand to lose quite a bit, as they will no longer be able to ride on the shoulders of upstream designers"). 86. Forever 21 makes frequent appearances at fashion websites that catalog copies. One such site, Fashionista, http://www.fashionista.com (last visited Feb. 18, 2009), lists Forever 21 copies of a Marc Jacobs shirt, a Marni handbag, and dresses by Foley & Corinna, Jonathan Saunders, and Phillip Lim. The extent of copying can be gleaned from trademark and copyright suits brought against the retailer. For examples, see Liza Casabona, Retailer Forever21 Facinga Slew of Design Lawsuits, WOMEN'S WEAR DAILY, July 23, 2007, at 12, and the suits summarized infra Table 1. 87. See, e.g., Eric Wilson, Before Models Can Turn Around, Knockoffs Fly, N.Y. TIMES, Sept. 4, 2007, at Al (describing one copyist's sales to Macy's and Bloomingdales, among others); Ben Winograd & Cheryl Lu-Lien Tan, Can FashionBe Copyrighted?,WALL ST. J., Sept. 11, 2006, at BI (describing canceled wholesale orders for Ananas handbag, once copyists produced versions for "between 10% and 50% of her $285 price"). On specialty retailers, see, for example, Kover, supra note 45 (describing Abercrombie & Fitch copy of bag by designer Nicole Dreyfuss); Susan Scafidi, Karmic Relief, COUNTERFEIT CHIC, May 10, 2007, http://www.counterfeitchic.com/2007/05/karmicrelief.php (last visited Feb. 18, 2009) (describing Forth & Towne copy of Narciso Rodriguez dress); Wilson, supra (describing $130 Bebe copy of $1700 Versace dress). 88. A.B.S., "[t]he uncontested champion of red-carpet knockoffs," sells to leading department stores. La Ferla, Thousand Knockoffs, supra note 82. By 2006, the Oscarknockoff business involved more than a hundred companies, with annual sales of $300 million. Id. 89. Keith Naughton, H&M's Material Girls: The Retailer Speeds Ahead with Fast Fashions,NEWSWEEK WEB EXCLUSIVE, June 10, 2007, http://www.newsweek.com/id/33983 (quoting H&M's chief designer: "We don't copy the catwalks.... We take inspiration from what's happening in the culture, with celebrities and on the catwalks."). 90. Lau, supra note 17 ("We've all heard the fashion knockoff tales. On one hand, there's the down-market riffing on designer motifs that ranges from the H&Ms and Forever 1172 [Vol. 61:1147 are different. Like the copyists, they move product to market very quickly.9 1 But their on-trend product, reactive though it is to the latest offerings of top designers, is not a precise copy. Instead, it is an adaptation or interpretation, developed by in-house designers. The firms' difference in design practice is reflected vividly in their relative frequency of suit. We searched Westlaw and the Stanford IP Litigation Clearinghouse for copyright or trademark suits against Forever 21, H&M, and Zara between 2003 and 2008. Forever 21 was a defendant in fifty-three suits during this period, compared to two for H&M and none for Zara.92 A review of the complaints in those cases shows that most of the Forever 21 suits alleged close copying, compared to at most one close copying complaint against H&M.93 As a research tool for scholars and other interested parties, we have collected the complaints, and those brought against several other alleged copyists, and made them available online.9 A selection of infringement suits against Forever 21, limited to the years 2007 and 2008, is summarized in Table I below. 9 5 21s of the world to counterfeit duds channeled through Chinatown dens."). 91. Zara takes four to five weeks to move from conception through delivery. Ghemawat & Nueno, supra note 84, at 9. Modifications or restocking of existing designs takes just two weeks. Id. 92. See Stanford IP Litigation Clearinghouse, http://lexmachina.stanford.edu (last visited Feb. 18, 2009), and Westlaw's DOCK-FED-ALL file. The search terms included both H&M and Hennes and Mauritz, and both Zara and its corporate parent Inditex. Complaints were retrieved directly from each district court's electronic case filing system or clerk's office. 93. See Complaint at 3-4, Tokidoki, LLC v. H & M Hennes & Mauritz LP, No. 07-cv- 1565 (C.D. Cal. Mar. 9, 2007) (alleging infringement of plaintiff's heart and crossbones trademark). The second suit arose from H&M's collaboration with designer Elio Fiorucci. The "Fiorucci" trademark had been acquired by a third party, which sued H&M for allegedly using the Fiorucci name when it promoted the collaboration. See Complaint, Edwin Co. v. H&M Hennes & Mauritz LP, No. 05-cv-4435 (S.D.N.Y. May 5, 2005). 94. See Berkman Center for Internet and Society, Harvard Law School, http://hub.law.harvard.edu/fashion (last visited Feb. 18, 2009). 95. Additional evidence comes from websites such as Fashionista, http://www.fashionista.com (last visited Feb. 18, 2009), which contain frequent examples of close copying by Forever 21, but not H&M or Zara. Nor, in several dozen interviews on the subject with a wide range of industry stakeholders, did we hear any specific complaints of close copying by either firm. For an exceptional, though general, allegation of close copying by H&M, see Winograd & Tan, supranote 87, at B 1 ("Designer Catherine Malandrino ... says she has seen almost identical versions of her blouses and sweaters in such stores as H&M and Esprit."). March 2009] FASHION 1173 STANFORD LA W REVIEW Table 1.Selected U.S. Litigation Against Forever 21, 2007-200896 Plaintiff Articles at Issue Anna Sui Seventeen articles Anthropologie Ten articles Bebe Stores Twenty-eight articles Carole Hochman Nightgown with "Marilyn Monrose" fabric design Diane von Furstenberg Four wrap dresses and one blouse Harajuku Lovers Clothing with "Heart and Heart/Box design" print Harkham Industries Dress with "Shadow Fern" design Trovata Six articles Fast-fashion copyists can have a beneficial effect upon trend adoption, since they reach customers at a lower price point who would otherwise not be reached by high-end designers. 97 But this benefit can be even better supplied by fast-fashion designers, who not only offer the on-trend product at a lower price but also supply differentiating details. B. The Threatto Innovation Mass copyists undermine the market for the copied good. Copies reduce the profitability of originals, thus reducing the prospective incentive to develop new designs in the first place. The predicted result, a reduced amount of innovation, is familiar from copying in other creative industries, such as file sharing of copyrighted music and films. 96. The information in this table isdrawn from First Amended Complaint at 7,Anna Sui Corp. v. Forever 21, Inc., No. 07-cv-3235 (S.D.N.Y. June 26, 2007); Complaint at 5-6, Anthropologie, Inc. v. Forever 21, Inc., No. 07-cv-7873 (S.D.N.Y. Sept. 6, 2007); Amended Complaint at 3-13, Bebe Stores, Inc. v. Forever 21, Inc., No. 07-cv-35 (N.D. Cal. June 7, 2007); Complaint at 3-4, Carole Hochman Design Group, Inc. v. Forever 21, Inc., No. 07cv-7699 (S.D.N.Y. Aug. 20, 2007); First Amended Complaint at 5-7, Diane von Furstenberg Studio, LP v. Forever 21, Inc., No. 07-cv-2413 (S.D.N.Y. Apr. 12, 2007); Complaint at 2-3, Harajuku Lovers, LLC v. Forever 21, Inc., No. 07-cv-3881 (C.D. Cal. June 14, 2007); Complaint at 4-5, Harkham Industries, Inc. v. Forever 21, Inc., No. 08-cv-3308 (C.D. Cal. May 19, 2008); Complaint at 6-9, Trovata, Inc. v. Forever 21, Inc., No. 07-cv- 1196 (C.D. Cal. Oct. 8, 2007). 97. The designer can reach cost-conscious customers to some extent through bridge lines, see Sally Weller, Fashion's Influence on Garment Mass Production: Knowledge, Commodities and the Capture of Value 129-30 (Oct. 2003) (unpublished Ph.D. dissertation, Victoria University), available at http://wallaby.vu.edu.au/adt-VVUT/public/adtVVUT20050201.101459/index.html, albeit usually not close copies, but a fast-fashion copy is a still lower price. It is therefore no surprise that designers have issued small "capsule" collections through fast-fashion firms in many instances. See Eric Wilson, The Big Brand Theory, N.Y. TIMES, Sept. 9, 2007, § 6 (Magazine), at 74. 1174 [Vol. 61:1147 Fashion copying is different from file sharing, however, in an important respect. File sharing provides access to essentially every musical work. Fashion copyists, by contrast, are selective. They have a business to run and costs to recoup, and so only the most profitable designs are copied. Moreover, not all copies reduce producer profits. Some are relatively harmless. The selectivity of copyists, combined with the uneven effects on producer profitability, reduce the incentives of some producers-and the incentive to produce some products-more than others. Thus, mass copying can be expected to affect the directionof innovation as well, as we explain below. 1.Harmfulcopying Copyists target designs that are technically and legally easy to copy. Consider, for example, a floral-patterned dress introduced by designers Dana Foley and Anna Corinna (F&C). 9 8 As a technical matter, the dress was easy to copy. It contained no exotic fabrics, complicated tailoring, or delicate embellishments that would make accurate outsourcing difficult. 99 It lacked any exterior brand logo that would subject a copyist to trademark liability. Its shape and exterior details did not so powerfully call to mind F&C's identity that trade dress protection would be available. These facts made the dress a good target for copyists.1 00 The Appendix contains photographs of the original and a copy by Forever 21. Moreover, for a midrange designer such as F&C, the sales of the copy substitute for and hence reduce sales of the original. 10' The original dress sold for hundreds, not thousands of dollars, which is within the reach of copyists' customers. 1 02 Sometimes the substitution is made by an aggressive retailer, 98. See La Ferla, FasterFashion,supranote 82 (describing dress). 99. Difficult-to-copy details are not an absolute bar because the copyist could omit or alter them. But such changes are costly and risky, since the copyist cannot tell, without incurring substantial cost, whether the detail is essential to the design's appeal. Moreover, accuracy may be important to those consumers or retailers who know of the original and explicitly seek a close copy. 100. La Ferla, FasterFashion,supranote 82 (noting that the original and copy were "almost identical," "[fQrom their fluid cut and noodle straps to the floral panel running down their fronts"). The floral print, assuming it satisfies copyright's originality requirement, provides a possible basis for a legal claim against Forever 2 1. 101. Kover, supra note 45 (describing accessory designer's drop in monthly revenue from $50,000 to $10,000, following imitation); Eric Wilson, Simply Irresistible,N.Y. TIMES, May 21, 2008, § SPG, at I (noting return of F&C dress by customers who saw the copy); see also William Filene's Sons Co. v. Fashion Originators' Guild of Am., Inc., 90 F.2d 556, 558 (1st Cir. 1937) ("A customer who ... sees a copy ... at another store at a lower price is quite likely to think that the retailer from whom she bought the dress lacks ability to select distinctive models and that she has been overcharged. Dresses are returned and customers are lost."). 102. Even customers of modest means might "trade up." For a discussion of this phenomenon, see MICHAEL J. SILVERSTEIN ET AL., TRADING UP: THE NEW AMERICAN LuxuRY 23-25 (2003). March 2009] FASHION 1175 STANFORD LAW REVIEW rather than the final consumer. 10 3 Either way, the profits of the original designer can be much reduced. The extent of targeting, combined with the degree of substitution, explain why midrange designers account for most anecdotal complaints of design copying.1 04 They also bring most of the lawsuits that attempt to circumvent the lack of design protection by alleging copyright or trade dress violations under existing law, against Forever 21 and other fast-fashion copyists. 105 In addition to replacing sales, the prevalence of cheaper copies also may reduce demand for the original design. This "snob" effect10 6 may reflect a consumer's desire for distinction from lower-status consumers or from other consumers more generally. It is a negative externality of overuse with analogies in trademark and copyright. 10 7 The effect is amplified, moreover, when the same shopper visits different stores-or different floors of the same department store-selling a particular design in its original and copied forms. 108 2. Distortinginnovation The reduced profits can be expected to have a negative effect on the amount of innovation; this is a standard result of economic theory. But in addition, there is a second effect. The lack of protection against design copying, combined with the existence of trademark, trade dress, and other protections, also distorts the direction of innovation. Designers unprotected against design copying see a disproportionate effect on their profitability, and hence are discouraged from innovating-indeed, from entering in the first place. Designers who are protected by trademark and trade dress innovate in ways that play to these legal advantages. The resulting effect on the direction of innovation is to favor innovation by designers who already enjoy existing 103. See, e.g., Winograd & Tan, supra note 87 (describing cancelled wholesale orders for Ananas bag); Felix Salmon, Market Movers: Susan Scafidi on CopyrightingFashion, PORTFOLIO, Sept. 19, 2007, http://www.portfolio.com/views/blogs/market-movers/2007/09/ 19/susan-scafidi-on-copyrighting-fashion (listing examples in which initial or subsequent orders went to a copyist rather than the original designer). 104. See, e.g., Kover, supra note 45 (describing designer's experience of learning that a nearly identical version of her necklace was selling for much less at a local accessories distributor); La Ferla, FasterFashion,supra note 82 (describing F&C designer's discovery of a Forever 21 copy of her dress alongside the original on a fashion blog); Winograd &Tan, supra note 87 (describing canceled retail orders for Ananas bag after other companies provided similar, cheaper designs). 105. The pattern of suits is an imperfect proxy, because they are design piracy cases undertaken as copyright or trademark suits, the only available tools. The data do not account for instances of copying where the originator did not or could not sue. The suits tend to highlight that copying which iscostliest for originators-copying costly enough to induce a suit with uncertain prospects. 106. See Leibenstein, supranote 3,at 189. 107. See, e.g., William M. Landes & Richard A. Posner, Indefinitely Renewable Copyright, 70 U. CHI. L. REv. 471, 485-86 (2003). 108. See supra note 101. [Vol. 61:11471176 March 2009] protection by other aspects of intellectual property law, over innovation by designers-particularly small, new designers-who are not thus protected.' 09 The existence of some kinds of intellectual property protection combined with the absence of design protection also gives designers the incentive to create some kinds of products over others. Consider, for example, trade dress, which protects features of product design that serve as a source identifier, such as the distinctive hardware of a Coach handbag.1 10 In two cases, the Supreme Court considered whether trade dress protection requires a showing that consumers have come to identify the feature with its maker, so-called "secondary meaning," or instead can rely upon the inherent distinctiveness of the feature. In the first case, outside the context of fashion designs, the Court ruled that secondary meaning was not necessary, in part because it recognized that such a requirement would place "particular burdens on the startup of small companies,""' because established firms are better positioned to imbue their products with secondary meaning. However, it later ruled that secondary meaning is required for trade dress in apparel and other product designs.112 The result is to favor those incumbents with the resources to invest in the creation of secondary meaning. 113 Trademark reinforces the incumbency bias in a powerful way. Brand logos provide strong protection against copying by legitimate producers. Designers understand the value of logos as an anticopying device. 1 14 Trademark protection accompanied by a lack of design protection thereby favors those firms that have strong trademarks and disproportionately encourages production of trademark-protected goods, such as articles with logos." 5 After all, if Gucci can prohibit copies of designs that employ its trademark interlocked "G's," but not a similar work that lacks the logos, it has an incentive to employ the logo. It also encourages the production of types of items, such as handbags, for which logos (and trade dress) are highly 109. As Karl Lagerfeld put it, copying "can be very damaging for small firms, though for a house like Chanel, it means a lot less." Godfrey Deeny, Lauren Finedby ParisCourt, and So Is Berge, WOMEN'S WEAR DAILY, May 19, 1994, at 1. 110. Coach, Inc. v. We Care Trading Co., 67 F. App'x. 626, 627 (2d Cir. 2002) (per curiam). 111. Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 775 (1992); see also id. at 774 (rejecting a secondary meaning requirement out of concern for its "anticompetitive effects"). 112. Wal-Mart Stores, Inc. v. Samara Bros., Inc., 529 U.S. 205, 215-16 (2000). The SamaraBrothers Court did not address its earlier Two Pesosdicta. 113. See, e.g., Complaint, Louis Vuitton v. Limited Brands, No. 05-cv-3980 (S.D.N.Y. Apr. 13, 2005) (asserting trade dress in a new line ofbags); Scafidi, supra note 10, at 121. 114. See, e.g., RENATA MOLHO, BEING ARMANI: A BIOGRAPHY 92 (2007) (quoting Giorgio Armani, who had been skeptical about monograms as an exterior decorative element, but acceded to an eagle logo for Emporio Armani to deter copiers, "even if it did not constitute a foolproof deterrent"). 115. Scafidi, supra note 10, at 121-22; cf Raustiala & Sprigman, supra note 11, at 1723 (acknowledging that trademark may be deployed to limit design copying). FASHION 1177 STANFORD LA W REVIEW complementary. Such "logoification" affects the communicative vocabulary that fashion provides, pulling fashion toward a status-conferring function and away from the communication ofdiverse messages. 116 Incumbents that produce luxury goods have several further advantages. Many high-end articles are hard to copy with low-cost outsourcing 117 because they use expensive and distinctive materials and finishes. 118 Investments in brand image provide an additional source of protection, and hence a further source of innovation distortion. A luxury image and retail buying experience insulate some high-end products from harmful copying. Brand image cements customer loyalty, a pampering in-store experience is pleasurable, and some customers value the authenticity of purchasins an original. 119 These effects reduce substitution when copying occurs. 12 They may also discourage copying, as they leave a copyist uncertain whether an item's appeal comes from its design, or instead from the inimitable purchase experience. Large incumbents are better able to apply that investment across high volumes and a wide variety of items. 12 1 The F&Cs of the design world-less-established designers who are not large incumbents-are again at a disadvantage. 12 2 116. Copyright introduces a secondary distortion. Copyright protects distinctive fabric patterns and physically separable ornaments, thus encouraging a designer to favor patterns over solids or investments to develop a new design. 117. Bespoke copying, with high-cost manufacture and close fidelity, is still feasible. 118. See, e.g., Binkley, supranote 73 (noting designers' increasing use of embroidery and other embellishments as a way to maintain a differentiated product); Reena Jana, Put a Patent on That Pleat,Bus. WK., Mar. 31, 2008, at 65 (describing Stuart Weitzman's use of titanium-reinforced heels, which are hard to copy because the heels will snap if copied using a cheaper material); see also Anna Van Praagh et al., One of These Bags Cost £23,000, The Other's a Snip at £114, MAIL ON SUNDAY (London), Mar. 11, 2007, at 68 (describing the £23,484 Louis Vuitton Tribute Patchwork bag, made from fifteen different Louis Vuitton bags, partly to deter counterfeiters). 119. Some consumers' valuation of a bag's authenticity may not be affected negatively by the existence of copies. Even if the copies look so good as to fool even a Louis Vuitton salesperson, and many will not know whether the bag is a copy, the purchaser of the Louis Vuitton handbag may take pleasure in knowledge of its authenticity. This is similar to a preference for an authentic piece of antique furniture over an identical-looking, well-made modem reproduction-the inner valuation of authenticity. Some classics of fashion might rise to take on this elusive aura in the face of existing knockoffs, but most items of fashion do not. 120. Consider, for example, a much-admired Christian Dior dress worn by actress Charlize Theron to the Oscars ceremony a few years ago. Copyist A.B.S. made a copy that was sold in department stores to promgoers. OscarDressKnock-Offs and More,CBS NEWS, Mar. 2, 2005, http://www.cbsnews.com/stories/2005/03/02/earlyshow/living/beauty/ main677562.shtml (comparing original, "estimated to cost between $15,000 to $20,000," with A.B.S. copy selling for $200 to $300). There is no substitution here. No buyer of the copy could afford the original, and buyers of the original avoided the copy, given its lesser quality and price signal. 121. For some large incumbents, such as Christian Dior, the ready-to-wear collection is an advertisement that keeps the brand in the public eye, thereby permitting sales of profitable handbags and perfume whose sales depend upon brand image. For such firms, a decline in appropriability might push designers toward provocative but unwearable designs. 1178 [Vol. 61:1147 A common normative response against the idea of intellectual property protection for fashion design grows out of the assumption that fashion is a visible marker of status. On this theory, making it more difficult to copy fashion may seem undesirable because it would promote the ability of wealthy people to enjoy and signal their status through apparel that only they can have, and thwart those who want to purchase cheaper knockoffs of those goods. After all, if rampant copying makes available cheaper knockoffs, that may disrupt the ability of the wealthy to distinguish themselves as a group through the signal of fashion. On this view, perhaps permission to copy effectively softens the socially stratifying effects of fashion, while legal restrictions on copying would reinforce them. But there is much more to fashion than signals about status. In light of the broader and more varied communicative and expressive aspects of fashion, status is only one of a wide variety of signals that fashion makes possible. Fashion has the potential to afford a broad vocabulary for the expression of a vast range of possible messages. Conscious or not, people's fashion choices signify and communicate, with meaningful individual and collective valences. We have identified this dynamic between differentiation and flocking as the key to the experience of fashion in social life. People use fashion to signal individual differences while also partaking in common movement with the collectivity. This model has informed our analysis of the formation and function of fashion trends among producers and consumers. The current intellectual property regime, in which legal protection from design copying is lacking, tends, if anything, to push fashion consumption and production in the direction of status and luxury rather than more polyvalent innovation. In sum, we have noted two distortions. The first is toward the creation of designs that are legally more difficult to copy. Trademark and trade dress already protect the most salient status-signaling items in fashion, those adomed with logos of high-end brands. Therefore, those who want to enable effective status signal-jamming should be critical of trademark protection, and not necessarily resist copyright protection for fashion design. The second distortion is toward the creation of goods that are naturally (as opposed to legally) more difficult to copy, or goods that are more difficult for design copying to harm-for example, goods involving unusual or expensive materials or difficult workmanship. For an example, see Cathy Horyn, Offstage, ParisFussesAbout Dior,N.Y. TIMES, Jan. 23, 2000, § 9,at 1 (describing a Dior show by designer John Galliano that was "[d]rawn from a nether world of tramps and mental patients,. .. which had models draped eccentrically in newsprint-patterned silk and straitjackets"). A similar opportunity is unavailable to small, independent designers. 122. "Perhaps because Ms. Foley and Ms. Corinna have been content to remain just under the radar, companies that specialize in making cheap copies of designer fashion have been bold in appropriating their designs." Wilson, supranote 101. March 2009] FASHION 1179 STANFORD LA W REVIEW The result ofthese distortions is to push creators toward the high-end realm of status and luxury, and away from devoting creative resources to design innovation. In a regime that protected original designs from copying, we would expect to see a shift in resources from developing brand-name or luxury goods or attempting close copies of designs toward developing a richer, more polycentric language of fashion that draws on and reinvents available inspirations and influences. We would expect to see greater range and variety in fashion innovation that would enlarge the vocabulary and the set of symbols with which we may produce meaning. At bottom, though, the main reason not to accommodate the lovers of cheap fashion knockoffs is more basic. It is the same reason that we do not have a legal regime that permits people freely to make and sell photocopies of another author's book and retain the profits. It is the theory of incentives. Obviously, people always want to purchase inexpensive copies of creative works or have them for free. The reason to disallow it is not to deprive them of that benefit but rather to provide creators with an incentive to create. That is no less true in fashion. C. Is PiracyReally Beneficial? The analysis so far shows that copyists reduce the amount of innovation and distort its direction. In an influential article, Kal Raustiala and Chris Sprigman (RS) have advanced the counterintuitive argument that in the fashion industry, "piracy paradoxically benefits designers."' 123 Some observers have found their argument persuasive. 124 Here we explain why we disagree with their argument. RS start from the premise that derivation, inspiration, and borrowing are valuable and central to fashion and innovation. This general point is one that we too emphasize. But this does not make fashion relevantly different from music and film, where the same processes are important engines of innovation.125 In order to conclude, as they do, that "[w]hat works to protect the creative process in film and music will have the opposite effect on the runway,"' 126 more is needed. 123. Raustiala & Sprigman, supra note 11, at 1722 ("[P]iracy paradoxically benefits designers by inducing more rapid turnover and additional sales."); see alsoid. at 1727 ("Our core claim is that piracy is paradoxically beneficial for the fashion industry, or at least piracy is not very harmful."). 124. E.g., Orit Fischman Afori, ReconceptualizingPropertyin Designs, 25 CARDOZO ARTS & ENT. L.J. 1105 (2008); Surowiecki, supra note 12; Hal R. Varian, Why That Hoodie Your Son Wears Isn't Trademarked,N.Y. TIMES, Apr. 5, 2007, at C3; Patti Waldmeir, Why Knock-Offs Are Goodfor Fashion,FIN. TIMES, Sept. 12, 2007, at 12. 125. See, e.g., LAWRENCE LESSIG, THE FUTURE OF IDEAS: THE FATE OF THE COMMONS INA CONNECTED WORLD 8-9 (2001). 126. Raustiala & Sprigman, supra note 12. [Vol. 61:11471180 March 2009] RS argue that the proliferation of copies of a style reduces the value of the style and renders it obsolete, which, in turn, causes consumers and hence producers to move on to new designs and trends. 127 At first, producer profits are high. Then the copyists come in, and snob effects reduce the value of the good in the hands of existing users and would-be new users. New sales grind to a halt, and existing users become dissatisfied with the goods that they have. That, in turn, provides a new opportunity to sell new goods.' 28 RS call this "induced obsolescence." 129 Because the opportunity to sell new goods is profitable, and entails additional innovation, RS argue that copying benefits designers and innovation. Hence the "piracy paradox." In light of this benefit, RS conclude it is a bad idea to protect designers from piracy.' 3 ° This type of argument has long played a role in debates over design protection. 131 RS's analysis does not distinguish close copies from other relationships between fashion designs, such as interpretation, adaptation, homage, or remixing. In arguing that "growth and creativity in the fashion industry depend upon copying," 132 the "piracy" part ofthe "piracy paradox" is seemingly meant to include both close copies and the full range of remixing and trend-joining activities. 133 RS treat close copying and shared trends as indistinguishable for their purposes, referring to both phenomena as "copies., 134 We have explained 127. Raustiala & Sprigman, supra note 11, at 1722 ("[T]he absence of protection... speeds diffusion and induces more rapid obsolescence."). 128. Id. at 1721-22. 129. The practical importance of "induced obsolescence" is uncertain because obsolescence has causes other than copying, including the passage of the seasons, a change in the spirit of the times that made the item salient, desire for the new, and the innovative product of other designers. These effects may be more important sources of obsolescence of fashion designs than the proliferation ofcopies. Even with respect to the example of induced obsolescence that RS provide, see id. at 1720-21 ("widely copied" Ugg boots), the explanation that copying by others destroyed the trend does not seem more likely than alternative explanations. 130. Raustiala & Sprigman, supra note 12 ("[G]rowth and creativity in the fashion industry depend upon copying."). 131. One close observer of the fashion industry in the 1930s, rehearsing the contrary positions in the debate, summarized the argument thusly: "On the other hand, it is pointed out that imitation means the rapid obsolescence of design which stimulates invention, assures to the designer a market, and brings to the industry accelerated business all along the line." Helen Everett Meiklejohn, Dresses-The Impact of Fashionon a Business, in PRICE AND PRICE POLICIES 299, 338-39 (Walton Hamilton ed., 1938). 132. Raustiala & Sprigman, supranote 12. 133. See, e.g., Raustiala & Sprigman, supra note 11, at 1715 (concluding, after a discussion of variations on a driving shoe, that "[flrom the perspective of the music or motion picture industries, this is called 'piracy"'). 134. See, e.g., id. at 1700 (treating "slavish copies" and "loose copies" in a like manner); id. at 1724 (similar); Raustiala & Sprigman, supranote 12 ("When [designers] see something that they like, they copy it---or, in the argot of the industry, they 'reference' it."). The term "copy," "copying," or its variants, has a variety of usages in technical copyright settings. See, e.g., 17 U.S.C. § 101 (2006) (defining "copies" as "material objects, other than FASHION STANFORD LA WREVIEW above that it is important to disagregate the phenomenon of close copying from the phenomenon of trends. 13 Doing so helps make visible the effects on innovation of close copying as distinct from the effects of interpretation, inspiration, or homage. 136 As we have also explained, there are also important differences among fast-fashion firms-differences between fellow designers such as Zara and H&M and copyists such as Forever 21-and their contrasting effects upon innovation. 137 To be complete, an analysis must attend to the distinctive effect of close copyists. To consider an analogy, the argument that a broad remixing right for music benefits subsequent innovators tells us little about whether to prohibit exact copies. 138 RS's "induced obsolescence" account emphasizes the increased profitability of faster cycles of new fashion trends spurred by unchecked copying. The assumption of profitability calls to mind Dr. Seuss's famous fable of the Sneetches. 139 There, the seller offered a new fashion article-stars to adorn the chest of each Sneetch. When too many Sneetches bought the stars, phonorecords, in which a work is fixed . . . and from which the work can be perceived, reproduced, or otherwise communicated"); id. ("The term 'copies' includes the material object ... in which the work is first fixed."); Feist Publ'ns, Inc. v. Rural Tel. Serv. Co., 499 U.S. 340, 361 (1991) (using "copying" in two distinct senses, neither corresponding to the statutory definition of "copies"). As the Supreme Court explained in Feist,"Not all copying, however, is copyright infringement. To establish infringement, two elements must be proven: (1) ownership of a valid copyright, and (2) copying of constituent elements of the work that are original." Id. In the quoted passage, the first use of "copying" pertains to factual copying. One can engage in such "copying" without any actionable similarity. The second use of "copying" in Feist pertains to actionable copying. Notably, it is "constituent elements" that are subject to "copying" in this second sense, rather than the work itself. 135. See supraPart I.C. 136. RS do acknowledge that "copying may cause harm to particular originators," Raustiala & Sprigman, supranote 11, at 1727, but here, too, they mean "copying" to denote interpretation and other forms of reworking. They argue that this harm is unimportant, because a designer is "shrouded within a Rawlsian veil ofignorance," id., and does not know in advance whether she will be a net borrower or lender of new material. That uncertainty does not plausibly extend to close copies, where the designers targeted for such copying are not also engaged in copying. See also Posting of Randy Picker to The University of Chicago Law School Faculty Blog, http://uchicagolaw.typepad.com/faculty/2006/1 1/ understanding-t.html (Nov. 14, 2006, 10:56 EST) (suggesting that a firm knows whether it is mainly a target, rather than a perpetrator, of "vertical copying"). 137. Comparesupra Part III.A, with Design PiracyProhibitionAct: Hearing on H.R. 2033 Before the Subcomm. on Courts, the Internet, and Intellectual Property of the H. Comm. on the Judiciary,109th Cong. 10 (2006) (testimony of Prof. Christopher Sprigman), 2006 WL 2127110 (F.D.C.H.) ("[S]ome of the biggest copyists are European: H&M, Zara and Topshop, these retailers, and European fashion firms that copy and that reinterpret and that recontextualize and that create derivative works and do all the things that fashion firms do."), and Raustiala & Sprigman, supra note 11, at 1737 (singling out H&M and Zara as "two of the major fashion copyists"), and id. at 1759 (similar). 138. See, e.g., LESSIG, supranote 18, at 144 ("I fight for 'free culture.' My position is weakened by kids who think all culture should be free."). Lessig supports narrow copyright protection where "creativity would be hindered by the absence of this special privilege." Id. at 85. 139. DR. SEUSS, THE SNEETCHES AND OTHER STORIES (Random House 1961) (1953). 1182 [Vol. 61:1147 devaluing them, the seller provided a new service: star removal. These cycles continued, with the seller profiting from each cycle, until the Sneetches ran out of money. This account, focused upon copies as the spur to the new, neglects the ex ante effect of the fashion cycle. What made star conferral and removal so profitable was that each Sneetch failed to recognize just how short-lived his fashion success would be, and to plan accordingly. Sneetches are not lifecycle pricers, in other words. But many fashion buyers are. If copying increases, and hence the fashion lifespan of the item falls, a consumer will recognize that fact and lower her willingness to pay. In the limiting case, producers' revenue is unchanged, as consumers make unchanged periodic payments for fashion, and profits fall due to higher (because more frequently incurred) design, materials, and other costs of production. Close copies make matters worse, reducing designer profits in the meantime by reducing sales. 140 The adverse effects of copying explain why many designers oppose copying, just as they oppose counterfeiting of handbags. (RS's piracy paradox argument, if correct, ought to apply to fashion trademarks and copyrights as well.) RS pitch their paradox as an explanation for the otherwise puzzling equanimity with which designers greet copyists. 14 1 But that premise is faulty. In fact, many designers are vocal advocates against copying, 14 2 and, as Table 1 suggests, make use of the currently limited legal tools available to curb copyists.143 140. This is not to say that life-cycle pricing will always undo a determined effort to profit from a deliberately short product lifespan. There is a substantial literature on "planned obsolescence" that shows how such efforts can succeed. See, e.g., Jeremy Bulow, An Economic Theory of Planned Obsolescence, 101 Q.J. ECON. 729 (1986). Under some conditions, these models predict deliberately low durability; under others, producers choose high durability in order to discourage other firms from entering the market. The induced obsolescence account does not lay out why the conditions for optimal low durability are met here, and if they are, why producers do not take advantage of other instruments for decreasing durability, such as product design. 141. See Raustiala & Sprigman, supra note 11, at 1755-58 (contending that designers have great political power, and therefore the absence of design protection suggests that designers don't really want it); cf. Posting of Chris Sprigman to Public Knowledge, http://www.publicknowledge.org/blog/1653 (Feb. 20, 2008, 18:41 EST) (describing proposed bill as "the CFDA's little vanity project"). 142. See supra note 9 (referring to collected quotations from designers and fashion executives). To be sure, on occasion, "[d]esigners admit to a certain pride that they are being copied. But their corporate backers are not so relaxed: piracy means an inferior product that too many may mistake for the real thing." Business Sense, ECONOMIST, Mar. 6, 2004, at 6 (survey). That sense of validation-and the desire to be provocative---explain why, while Marc Jacobs the firm opposes copying, Marc Jacobs the designer (and Louis Vuitton creative director) declares not only design piracy but even counterfeiting to be "fantastic." DANA THOMAS, DELUXE: How LUXURY LOST ITS LUSTER 276 (2007). 143. See also Barnett et al., supra note 72, at 29 (compiling infringement suits reported by Women's Wear Daily). Further evidence comes from European practice, where designers use the relatively strong protection available there to curb close copies. See infra notes 173-85 and accompanying text. March 2009] FASHION 1183 STANFORD LA WREVIEW Vigorous designer opposition to copyists is not new. Designers cared so much in the 1930s that they set up an enormous, costly, and successful private system of self-help, the Fashion Originators' Guild of America, which boycotted retailers that did business with copyists, until it was enjoined as a violation of antitrust law. 144 Their decades-long lobbying effort for stronger protection has been unsuccessful, not because designers are not harmed, but because they are not sufficiently powerful. 145 The induced obsolescence account has a further evidentiary limitation. If designers did profit from "induced obsolescence" of their products, they could induce the obsolescence themselves by taking a lax approach to counterfeits, or by engineering products designed to fall apart quickly. 146 That they do not do so where it is currently feasible suggests that inducing obsolescence is not what fashion designers are engaged in. Even with protection, designers interested in an induced-obsolescence strategy could implement it by disclaiming protection against copying, or by burning out the trend more profitably on their own, without any help from copyists. 14 7 Moreover, since, as we have explained, fashion trends do not depend on copying, designers would not need to induce obsolescence through copies in order to ensure the robust trends that comprise fashion. IV. TAILORED PROTECTION FOR ORIGINAL DESIGNS The analysis up to this point explains how the increased ease of copying disrupts innovation. It reduces the amount and shifts the direction. That, in turn, undermines the formation of differentiated communicative tools. Our proposed policy response aims to preserve differentiated innovation. Our distinctive goal is to prohibit close copies while preserving flocking and differentiation in its varied forms of inspiration, homage, referencing, and quotation. The guiding principle throughout is to avoid the hypertrophy or thicket of rights that is threatened by excessive, multiple rightsholders. 148 The proposal that thus grows out of our analysis is a narrow new right that protects designers against close copies of their designs but does not protect against looser forms of similarity that may arise as designers commonly 144. See Marcketti & Parsons, supra note 78, at 226. 145. In particular, many manufacturers and retailers, including department stores, benefit from copying. 146. For a discussion of such strategies, see Barak Y. Orbach, The DurapolistPuzzle: Monopoly Power inDurable-GoodsMarkets, 21 YALE. J. ON REG. 67, 91-92 (2004). 147. RS suggest that designers' bridge lines accomplish this, see Raustiala & Sprigman, supra note 11, at 1724-25, but their example, Armani, seems inapt at least as applied to close copying, since Armani's five lines--Giorgio Armani, Armani Collezioni, Armani Jeans, Emporio Armani, and Armani Exchange--each echo the Armani style, but do not offer the same design at a lower price point. Self-protectiveness about brand image may limit the extent ofself-cannibalization. See Barnett, supranote 10, at 1406-07. 148. See MICHAEL HELLER, THE GRIDLOCK ECONOMY (2008). 1184 [Vol. 61:1147 participate in fashion trends. In recommending tailored protection for the fashion industry, we join other scholars who have urged industry-specific solutions to the regulation of innovation. 149 Part IV.A describes the scope of the proposed new right. Part IV.B considers some objections to our proposal. A. The Scope ofthe Right The proposed right has two components. First, it provides copyright protection to original works of apparel, even though these useful articles are currently not copyrightable. Second, it denies copyright protection where the later work, though arguably "substantially similar"-the usual standard for copyright liability-is also substantially different. The Copyright Act accords protection to "useful articles"-articles, such as apparel, that have "an intrinsic utilitarian function"15 0 -- only to the extent that protected features "can be identified separately from, and are capable of existing independently of," the utilitarian aspects. 151 This latter statutory requirement goes by the name of "separability." The exclusion of apparel results from a particular interpretation of separability for works that have both a functional and an expressive component, such as an item of apparel or an architectural work. Separability can take a physical or conceptual form. Physical separability is present when the article, minus the protectable element, suffers no loss of utility, and the separated element can stand alone as a work of art. 152 Physical separability suffices to protect an appliqu6 sewn onto a sweater, but not the cut, color, and appearance ofan article ofapparel. 149. See, e.g., Dan L. Burk & Mark A. Lemley, Policy Levers in Patent Law, 89 VA. L. REv. 1575 (2003) (advocating industry-specific judicial interpretation of patent doctrines); C. Scott Hemphill, Payingfor Delay: PharmaceuticalPatent Settlement as a Regulatory Design Problem, 81 N.Y.U. L. REv. 1553 (2006) (offering an industry-specific approach to antitrust law); William Fisher, The Disaggregationof Intellectual Property, HARV. L. BULL., Summer 2004 (offering a cautious endorsement of industry-specific intellectual property rules). 150. 17 U.S.C. § 101 (2006) (defining a "useful article" as "an article having an intrinsic utilitarian function that is not merely to portray the appearance of the article or to convey information"). 151. Id. For a historical account of this state of affairs, see 1 WILLIAM F. PATRY, COPYRIGHT LAW AND PRACTICE 269-70 (1994). 152. 1 PAUL GOLDSTEIN, GOLDSTEIN ON COPYRIGHT § 2.5.3 (3d ed. 2005 & 2008 Supp.). March 2009] FASHION 1185 STANFORD LAW REVIEW Most courts also recognize the possibility of conceptual separability. 153 Defining its boundaries is a notoriously difficult task, and courts and commentators have reached a wide range of views as to the proper breadth of the doctrine. 154 An expansive understanding of conceptual separability would be one way to provide protection for many designs, without the need for statutory change. That is, courts could potentially deem design aspects of a garment to be conceptually separable from a garment's usefulness, and hence protected by current copyright law. The difficulty, however, is that, as with creative works of architecture, for example, design features often are treated as inseparable from a work's function. The statutory alternative, and a more complete solution, is to take original fashion designs outside the domain of the separability regime, by adding them as a new and distinct type of copyrightable subject matter. This is a familiar part of copyright policymaking. In 1990, Congress took that step with respect to architectural works. 155 We suggest that fashion designs receive copyright 153. Compare Pivot Point Int'l, Inc. v. Charlene Prods., Inc., 372 F.3d 913, 931 (7th Cir. 2004) (en banc) (recognizing conceptual separability), and Kieselstein-Cord v. Accessories by Pearl, Inc., 632 F.2d 989, 993 (2d Cir. 1980) (same), with Esquire, Inc. v. Ringer, 591 F.2d 796, 803 (D.C. Cir. 1978) (limiting separability to physical separability). See also 1 MELVILLE B. NIMMER & DAVID NIMMER, NIMMER ON COPYRIGHT § 2.08[B][3] (2008) (concluding that conceptual separability is a valid approach because the legislative history of the 1976 Act relies approvingly upon an earlier case, Mazer v. Stein, that found conceptual separability but not physical separability). 154. The Seventh Circuit recently collected six possible tests in PivotPoint: [1] where the article's artistic features are "primary" and the utilitarian features are "subsidiary" (following Kieselstein-Cord,632 F.2d 989); [2] where the article "stimulate[s] in the mind of the beholder a concept that is separate from the concept evoked by its utilitarian function," see Carol Barnhart Inc. v. Econ. Cover Corp., 773 F.2d 411, 422 (2d Cir. 1985) (Newman, J., dissenting); [3] where the article "would still be marketable to some significant segment of the community simply because of its aesthetic qualities," see Galiano v. Harrah's Operating Co., 416 F.3d 411, 421 (5th Cir. 2005); [4] where "the artistic design was not significantly influenced by functional considerations"; [5] where "the feature[] can stand alone as a work of art traditionally conceived," and the article "in which it is embodied would be equally useful without it"; and [6] where "the artistic features are not utilitarian." Pivot Point, 372 F.3d at 923. The Seventh Circuit then devised its own test, requiring that separability exists when the article's artistic aspects can be "conceptualized as existing independently of their utilitarian function," a finding informed by "whether the design elements can be identified as reflecting the designer's artistic judgment exercised independently of functional influences." Id. at 931; see also 1 NIMMER & NIMMER, supra note 153, § 2.08[B][3] (canvassing this "fractured field"). 155. See Architectural Works Copyright Protection Act, Pub. L. No. 101-650 § 703, 104 Stat. 5089, 5133 (1990) (adding "architectural works" to subject matter of copyright); id. § 702(a), 104 Stat. at 5133 (adding "architectural work" to the definitions in 17 U.S.C. § 101); id. § 704(a), 104 Stat. at 5133 (placing limits on the copyright in an architectural work, including denial of protection for certain pictorial representations); Donald Frederick Evans & Assocs., Inc. v. Cont'l Homes, Inc., 785 F.2d 897, 901 n.7 (11 th Cir. 1986); 1 NIMMER & NIMMER, supranote 153, § 2.20 ("United States copyright law prior to [1990] did not accord protection to structures, except those few that served no utilitarian purpose."). 1186 [Vol. 61:1147 protection that runs parallel to that now granted to buildings and architectural plans. 156 What counts as infringement is a second crucial question. To begin, standard features of a design-a pinstripe, say, or an A-line silhouette-are not copyrightable features. Their appearance in a latter work would not give rise to an infringement claim. This is a familiar element of copyright law. 157 Beyond that, copyright law provides that, as to protectable elements of the work, "substantial similarity" between the two works amounts to infringement. This rule applies not only to standard copyrighted works such as books, art, film, and music, but also to newly added subject matter such as architectural works. 158 Substantial similarity varies with the circumstances. Where copyright subsists in a compilation of unprotectable parts, the copyright is sometimes said to be "thin," and protects the originator only against relatively close copies. 159 One proposed bill to protect original fashion designs applies a substantial similarity standard. 160 Our analysis of copying and trends recommends a different and narrower rule. We would prohibit only close copies, in order to support differentiation amidst flocking. If a designer copies protectable expression from an earlier work, yet also makes significant changes, the designer is no longer liable. To the extent a thin compilation copyright does not narrow substantial similarity to 156. The architectural amendment was made, in part, to comply with the Berne Convention. See 1 NIMMER & NIMMER, supra note 153, § 2.20. Arguably, the change proposed here is necessary to comply with TRIPS requirements as to industrial design. The TRIPS component of the Uruguay Round Agreement requires members to "provide for the protection of independently created industrial designs that are new or original." Agreement on Trade-Related Aspects of Intellectual Property Rights art. 25, Annex IC, Apr. 15, 1994, 33 I.L.M. 1197, 1207 (1994). United States design patents provide protection for industrial designs that are "new," but the TRIPS agreement's use of "or" suggests that designs that are original, but not new, must also receive protection. See Jerome H. Reichman, Universal Minimum Standardsof IntellectualPropertyProtectionUnder the TRIPS Component of the WTO Agreement, 29 INT'L LAW. 345, 376 (1995). Extension of copyright would afford protection to originality even without novelty. 157. 17 U.S.C. §§ 101-102 (2006). 158. Id. § 101 (including definition for "architectural works," and extending "pictorial, graphic, and sculptural works" to include architectural plans); id. § 102 (including "architectural works" in the coverage of copyright). 159. See, e.g., 4 NIMMER & NIMMER supra note 153, § 13.03 (noting that where protection is thin, the degree of required similarity required to satisfy "substantial similarity" increases); Intervest Constr., Inc. v. Canterbury Estate Homes, Inc., No. 07-12596, 2008 WL 5274274 (11 th Cir. Dec. 22, 2008) (concluding, in light of thinness of copyright in a floor plan, that differences in protectable expression were significant enough to justify conclusion that works were not substantially similar). 160. See, e.g., Design Piracy Prohibition Act, H.R. 2033, 110th Cong. §§ 2(a), (d) (2007) (adding fashion designs to types of design protected without altering "substantial similarity" infringement standard). But see Design Piracy Prohibition Act, S. 1957, 110th Cong. § 2(d) (2007) (altering applicable infringement standard to embrace only designs which are "closely and substantially similar in overall visual appearance to a protected design"). March 2009] FASHION 1187 STANFORD LAW REVIEW protection against only close copies, our proposal departs from the adage offered by Judge Learned Hand that "it is enough that substantial parts were lifted; no plagiarist can excuse the wrong by showing how much of his work he did not pirate.' 16 1 Under our proposed rule, showing a substantial difference does indeed excuse the wrong. 162 This is not a radical step, either. In 1984, an analogous right was enacted as to another copyright misfit, namely the designs of semiconductor chips, and in particular the stencil-like "mask works" used in chip production. Protection extends to reproduction, importation and distribution of the mask work in question, and to a product embodying it.163 Substantially similar products are not subject to the prohibition. There is no broad control over the path of future innovation. We propose a similar standard here. The difference has important consequences. A designer is free to join a trend once it has begun, adopting the trend feature but altering the details to satisfy particular demand for differentiation. The test we propose would ask whether an ordinary observer could discern the copy from the original. 164 This would be a test of "substantial dissimilarity." If the two works were substantially dissimilar, no infringement would be found. Like other intellectual property standards that require subjective comparison of two works, our substantial dissimilarity test can raise difficult line-drawing problems. Consider, for example, Yves Saint Laurent's famous suit against Ralph Lauren, brought under French copyright law, alleging infringement of a black tuxedo dress designed by Saint Laurent. 165 Although the two articles differed in fabric (silk rather than wool), pockets (YSL's had none), lapel width, and the substitution of black buttons for gold, the court 161. Sheldon v. Metro-Goldwyn Pictures Corp., 81 F.2d 49, 56 (2d Cir. 1936). 162. Nor do we propose any right to control the preparation of derivative works. That right "substantially overlaps the scope of the reproduction right," 2 GOLDSTEIN, supra note 152, § 7.3.1, though the degree of overlap is open to dispute. The ordinary case of a protected article that "borrows expressive elements from the original, but adds expressive elements of its own," would arguably implicate the reproduction right, rather than the derivative works right. Id. We do not mean to enter that debate. The point here is that we intend a right narrower than the usual copyright. 163. 17 U.S.C. § 905 (2006). 164. Tim Gunn, former chair of fashion design at Parsons who later gained fame on the television show ProjectRunway, says "I draw a line at something that, ifyou squint your eyes, you really can't discern it from the original." Serena French, Knock It Offi-Fashion Fights Back at Year of the Copycat; Counterfeit Counterattack,N.Y. POST, May 1, 2007, at 41. 165. Soci6t&Yves Saint Laurent Couture S.A. v. Soci6t6 Louis Dreyfus Retail Mgmt. S.A., [1994] E.C.C. 512, 514 (Trib. Comm. (Paris)). Yves Saint Laurent's version sold for $15,000, Ralph Lauren's for $1000. Yves Saint Laurent sued after seeing the Ralph Lauren dress in a French fashion magazine. The dress was shown as part of a larger editorial spread featuring women's fashion inspired by the tuxedo (in French, le smoking), see Femmes en smoking, JOURS DE FRANCE, Dec. 7, 1992, at 138-43-a nice example of differentiation amidst flocking. 1188 [Vol. 61:1147 imposed liability. 166 The Appendix contains photographs of both dresses. On our standard, the substantial differences would suffice to avoid liability, but we concede that the question is a close one. That said, these problems seem no more severe than those in ordinary copyright, trademark, or patent infringement cases. Why not go further and grant a broader right? Why not provide protection for "the cut of a dress or the sleeve of a blouse" 167 -and in essence, grant a single firm control over the exploitation of a trend? This possibility, sometimes described by intellectual property scholars as the granting of a "prospect," raises some familiar problems that are likely to be particularly acute in the fashion context. Here, as in many areas of creative endeavor, good ideas are dispersed. 168 Ideas for differentiated products that participate in the trend are scattered among many designers, and a single firm that controls the trend is less likely to get it off the ground. Identifying and negotiating with those designers who would use the feature is likely to be very costly. Moreover, many products would likely infringe multiple features, compounding the negotiation problem. At the same time, the granting of a broad right would provide no valuable incentive to upstream development. Unlike, say, a blockbuster movie or basic technology that forms the basis for downstream products, a trend feature is not the result of a single creator's deep thinking or heavy investment. Rather, trend features arise in the collective way we described in Part II. Legal control is not needed to elicit these ideas, and a legal entitlement would likely create difficult disputes over ownership, given the often simultaneous or near-simultaneous processes by which multiple designers flock to a particular idea. What should be the appropriate duration of protection? Ordinary copyright lasts for the life of the author plus seventy years. 169 Recent fashion proposals considered by Congress provide for three years of protection. 170 In our view, this is plenty of time. Most fashion articles have only a brief opportunity to recoup the cost ofdesign in any event. A short lifespan has the additional virtue of limiting the set of articles that a new design might possibly infringe. 166. Deeny, supra note 109; Michele Ingrassia, A Not-So-Little Black Dress, NEWSWEEK, June 6, 1994, at 72. The judgment was $383,000. Agins, supra note 72. The presiding judge added that the Saint Laurent dress, "I must say[,] is more beautiful-though, ofcourse, that will not influence my decision." Deeny, supranote 109, at 11. 167. Jessica Litman, The Exclusive Right to Read, 13 CARDOZO ARTS & ENT. L.J. 29, 44 (1994). 168. See, e.g., SUZANNE SCOTCHMER, INNOVATION AND INCENTIVES 38 (2004); C. Scott Hemphill, Network Neutrality and the False Promiseof Zero-PriceRegulation, 25 YALE J. ON REG. 135, 174 (2008). 169. 17 U.S.C. § 302(a) (2006). The term is ninety-five years for anonymous works, pseudonymous works, and works made for hire. Id. § 302(c). 170. Design Piracy Prohibition Act, H.R. 2033, 110th Cong. § 2(c) (2007); Design Piracy Prohibition Act, S.1957, 110th Cong. § 2(c) (2007). March 2009] FASHION 1189 STANFORD LA W REVIEW Some proposals incorporate fashion within an expansion of Chapter 13 of the Copyright Act, which was set up as a catchall for other design rights. 171 Should we take this opportunity to add other design rights such as furniture? That analysis is beyond the scope of this Article. We have not considered whether furniture or other design-intensive industries, which also lack protection, have a similar equilibrium of flocking and differentiation to preserve. Much seems different, including the role of trends, and the extent to which a trend feature coexists with differentiation. Seasonality is absent; fast fashion, too. The dynamics of furniture and other design-intensive industries await future research. B. ConsideringObjections This Subpart evaluates challenges to our argument that narrow copyright protection reduces copying, that reduced copying leads to more innovation, and that increased innovation is desirable. First, will new protection in the United States have any effect upon copying, given existing protection in Europe (among other jurisdictions'172)? The European design right protects the features and overall appearance of an article. 173 Although there is a registration system, the strong protection granted to unregistered designs makes registration unnecessary. 174 Individual states 171. Chapter 13 of the Copyright Act has the grand title "Protection of Original Designs," and protects, in seemingly general terms, "useful articles." 17 U.S.C. § 1301(a) (2006). But "useful articles" is defined therein as a "vessel hull, including a plug or mold." Id. § 1301(b)(2). The proposed Design Piracy Protection Act expands "useful articles" to include apparel, handbags, belts, and eyeglass frames. Design Piracy Prohibition Act, H.R. 2033, 110th Cong. § 2(a) (2007). 172. Although we focus upon European protection, it is notable that other jurisdictions also protect original designs. For example, Japan's industrial design right protects the "form, pattern, or color of an object or a combination of these, which appeals visually to the viewer's sense of aesthetics." Japan External Trade Organization, Investing in Japan § 5.7.1, http://www.jetro.go.jp/en/invest/setting-up/laws/section5/page7.html (last visited Feb. 18, 2009). In addition, unfair competition law applies to original designs. Id. § 5.7.2; see also Interview with Shigekazu Yamada, Nat'l Ctr. for Indus. Prop. Info. & Training, Japan Patent Office, in Tokyo, Japan (May 21, 2008) (describing seizure of counterfeit Hermes purses for violating unfair competition law). 173. Council Regulation 6/2002, art. 3, 2002 O.J. (L 3) 1, 4 (EC) (protecting "appearance of the whole or a part of a product resulting from the features of, in particular, the lines, contours, colours, shape, texture and/or materials of the product itself and/or its ornamentation"). 174. Unregistered designs are protected from copying for three years. Id., art. 19 (L 3) 7 (scope of protection); id., art. 11 (L 3) 5 (duration of protection). Registration extends the duration to twenty-five years, if renewed every five years, id., art. 12 (L 3) 5, and adds a protection against independent invention. Id., art. 19 (L 3) 7. Designers enjoy a one-year grace period after the design's public debut before registration is necessary. Id., art. 7(2) (L 3) 5; see also Hedrick, supranote 13, at 251; OFFICE FOR HARMONIZATION INTHE INTERNAL MARKET, FREQUENTLY ASKED QUESTIONS ABOUT THE COMMUNITY DESIGN: GENERAL QUESTIONS, http://oami.europa.eu/ows/rw/pages/RCD/FAQ/RCDl.ed.do (last visited Jan. 1190 [Vol. 61:1147 provide additional protection. 175 Cease-and-desist letters do much of the work of enforcement, 176 but litigation is significant too. Table 2 summarizes a few recent cases. Table 2. Selected European Litigation, 2005-2008 Case Articles at Issue Hennes & Mauritz AB v. "[A] Chinese-style dragon and flame Primark Stores pattern, a target-style design, a graffiti pattern, a . . . badge design and a floral print" 177 Monsoon v. Primark Stores Two skirts, swimwear, trousers, a scarf, and patterned socks 178 Chlo6 v. Kookai Handbag1 79 J. Choo Ltd. v. Towerstone Ltd. Handbag 180 Chanel v. Camille & Lucie Jewelry 18 1 31, 2009). 175. For example, French law includes fashion explicitly in copyrightable subject matter. CODE DE LA PROPRItTE INTELLECTUELLE art. L1i2-2 (1994), available at http://www.legifrance.gouv.fr/htmllcodes-traduits/cpialtext.htm (including, among the "works of the mind" covered by copyright law, "creations of the seasonal industries of dress and articles of fashion," that is, "industries which, by reason of the demands of fashion, frequently renew the form of their products," and naming a long list of articles, fabrics, and other products). 176. Susan Scafidi, No, No, Naf Naf, COUNTERFEIT CHIC, July 21, 2008, http://www.counterfeitchic.com/2008/07/no no-naf naf.php (last visited Feb. 18, 2009) (asserting that European companies "regularly settle" rather than litigate); see also Telephone Interview with Nathalie Moull&-Berteaux, Intellectual Prop. Dir., LVMH Fashion Group (Nov. 21, 2008) (noting firm's vigorous cease-and-desist practice against infringers); cf Video: Stop Fashion Piracy, http://www.stopfashionpiracy.com/theindustryspeaks.php (last visited Oct. 4, 2008) (quoting Robert Triefus, EVP Communications, Armani, that European protections have a substantial effect). 177. Jim Armitage, H&M Seeks Redressfrom Primarkover "Copycat" Designs Row, EVENING STANDARD (London), Mar. 8, 2005, at 35. H&M alleged damages of £100,000. Id. 178. Lucy Famdon, Monsoon Sees Red, DAILY MAIL (London), Apr. 19, 2005, at 68 (noting that Monsoon claims £200,000 in damages); Laura Peek, Copycat or Coincidence? Stores FaceCourt Clash,TIMES (London), Apr. 19, 2005, at 5. This case, like the H&M case against Primark, later settled. Lauren Veevers & Danny Fortson, Primark Chic, INDEPENDENT ON SUNDAY (London), Nov. 5, 2006, at 24. 179. Hadley Freeman, Bag Snatchers:High Street Copies Taken to Court, GUARDIAN (London), July 23, 2005, at 10. The suit proceeded under both European and UK design protection. For an earlier case under the UK design right, see Lambretta Clothing Co. v. Teddy Smith Ltd., [2004] EWCA Civ. 886 (Eng.), available at http://www.bailii.org/ew/cases/EWCA/Civ/2004/886.html (track suit with same arrangement of colors). 180. J. Choo Ltd. v. Towerstone Ltd., [2008] EWHC 346 (Eng.), available at http://oami.europa.eu/pdf/design/cdcourts/Handbags.pdf. Jimmy Choo has brought multiple suits asserting European design protection. See, e.g., New Look Withdraws 1,000 Shoes to Settle Copying Case, TIMES (London), Sept. 13, 2006, at 56 (noting that "the designer had used relatively new European legislation"). March 2009] FASHION STANFORD LAW REVIEW Case Articles at Issue Isabel Marant v. NafNaf Little black dress 182 Karen Millen Ltd. v. Dunnes Two striped shirts and a knit top 183 Stores European protection has a limited effect upon the U.S. market. Fastfashion firms based in Europe, such as Zara and H&M, are subject to design protection. We would therefore expect them to avoid close copying as to products sold in Europe. If these firms sell the same products in both Europe and the United States, then we should expect relatively few close copies in the United States as well. As discussed above, that is indeed what we observe. 184 By contrast, Forever 21 is based in the United States, and has no stores in Europe. 185 For it and other U.S.-focused copyists, European protection has no effect upon the production of close copies. Meanwhile, for U.S. designers who lack a substantial non-U.S. business, the entire market is subject to copyists. Thus, existing European protection does relatively little to help many U.S. designers. Second, will our proposed protection really reduce copying? Louis Vuitton has the resources to sue, but do smaller firms? We think the answer is yes. Under existing law, small designers already file suit. In the Forever 21 suits summarized in Table 1, many are by small designers. We see no reason to doubt they would take advantage of expanded protection. In this respect, fashion is no different from other areas of copyright, patent, and trademark, in which small plaintiffs are able to invoke their rights, 186 sometimes with the assistance of counsel retained on a contingency basis. 187 181. Katya Foreman & Emilie Marsh, Hermis, Dior Notch Counterfeit Wins, WOMEN'S WEAR DAILY, Apr. 9, 2008, at 2. Chanel was joined in this suit by Givenchy, Van Cleef & Arpels, Boucheron and Cartier; the total fine was 700,000 euros, or about $1.1 million. Id. In a separate suit filed by Christian Dior Couture, a further 150,000-euro ($230,000) fine was imposed. Id. 182. In Brief: Penaltyfor Copying, WOMEN'S WEAR DAILY, July 18, 2008, at 2; Condamnationpour copie: Naf Naf ne trouve pas cela "Marant," AGORAVOX, July 22, 2008, http://www.agoravox.fr/article.php3?id-article=42446. Naf Naf had sold a 70-euro copy of a dress that retails for 250 euros. The court imposed damages of 75,000 euros. 183. Karen Millen Ltd. v. Dunnes Stores, [2007] IEHC 449 (Ir.). 184. Raustiala and Sprigman draw the opposite conclusion from a single global product: that it shows that Zara and H&M operate with impunity in Europe. Raustiala & Sprigman, supranote 11, at 1737. 185. See Forever 21, Store Locator, http://www.forever2l.com/store/storelocator.asp (last visited Jan. 31, 2009). In addition to Forever 21, the other copyists discussed supra, such as A.B.S. and unbranded manufacturers that sell to U.S. department stores, are focused upon the U.S. market. 186. See, e.g., Rogers v. Koons, 960 F.2d 301 (2d Cir. 1992) (copyright); Big 0 Tire Dealers, Inc. v. Goodyear Tire & Rubber Co., 561 F.2d 1365 (10th Cir. 1977) (trademark); Kearns v. Ford Motor Co., 726 F. Supp. 159 (E.D. Mich. 1989) (patent). 187. This arrangement is common in patent cases. For an example in trademark and 1192 [Vol. 61:1147 March 2009] The designer will sue only if she expects a positive return on her litigation investment. Again, the existence of suits under the current regime shows that sometimes the stakes are large enough. Even where copyist manufacturers are judgment-proof, copyist retailers, generally speaking, are not. To be sure, where damages are small or difficult to calculate, deterrence is weakened, as in other areas of intellectual property. Damages here can be augmented by statutory damages 188 and awards ofattorney's fees. 189 One way to strengthen deterrence is to consider mechanisms by which designers might band together. Economies of scale in enforcement are familiar from musical collective rights organizations such as ASCAP, and from the original Fashion Originators' Guild. Like these organizations, a modem-day Guild could monitor and thereby deter unlicensed use. The new Guild, backed by law rather than the threat of boycott, would provide a credible enforcement commitment in situations where individual designers found enforcement too expensive to be worthwhile. 190 A related objection is that a new right will be an effective weapon only in the hands of established designers, and will be used not against copyists, but against the very designers most in need of protection. This objection has greatest force as applied to broad design protection. It seems unlikely to pose much trouble for the narrow right against close copies that we propose here. Third, does reduced copying lead to more innovation? After all, it is sometimes argued, there is a lot of innovation already. As we have explained, that innovation is increasingly under threat, particularly innovation not already protected by trademark or investments in brand image. But there is a more basic point. The level of existing innovation, high or low, tells us little about the incremental effect of a policy change. The fact that music sales are large, despite illegal copying, hardly demonstrates that copying is good or even neutral for creators of new music. As we have explained, fashion is relevantly similar to other areas of creative production, and we expect designers to respond to economic incentives in the usual way. Strong real-world evidence that protection reduces copying, which in turn increases innovation, comes from our single national experiment with copyright, see JCWInvs., Inc. v. Novelty, Inc., 482 F.3d 910 (7th Cir. 2007). 188. See, e.g., 17 U.S.C. § 504 (2006) ($30,000 for copyright infringement, or $150,000 in the case of willful infringement). 189. See 15 U.S.C. § 1117(a) (2006) (trademark); 17 U.S.C. § 505 (2006) (copyright); 35 U.S.C. § 285 (2006) (patent). 190. In this respect, our Guild proposal resembles intellectual property enforcement insurance, which covers the insured's litigation expenses in case of a dispute. For an example, see Intellectual Property Insurance Services Corporation, IP Abatement Insurance, http://www.ipisc.com/products/insurance-policies/abatement (last visited Feb. 18, 2009). Such insurance serves to commit a rightsholder to pursue a claim. For a formal explanation, see Gerard Llobet & Javier Suarez, Patent Litigation and the Role of Enforcement Insurance (Feb. 2008) (unpublished manuscript, available at http://www.cemfi.es/-Ilobet/ PLpaper.pdf). FASHION 1193 STANFORD LA WREVIEW protection for original designs. During the heyday of the Fashion Originators' Guild, the Guild's privately enforced protection reduced copying greatly. 19 1 Moreover, contemporaneous observers understood that the prohibition of piracy caused manufacturers to shift production from copying to original design. 192 A fourth type of objection views substantial existing innovation as an argument against protection, not because protection won't increase innovation, but because it will. In particular, increased innovation might be thought undesirable if it leads to excessive product differentiation. This possibility-a kind of over-entry, in which additional entry incurs social costs but does little to better satisfy consumer wants-has long been contemplated by a large theoretical literature in economics. 193 Despite this theoretical possibility, we see no reason to conclude that it is unusually severe in fashion compared to other areas of creative production. Absent such a reason, either fashion should enjoy the higher protection of other types of creative production, or these other areas should also be denied copyright protection out of fear of excessive differentiation. 194 191. See, e.g., Guild's Work Good in Upper Brackets, N.Y. TIMES, Feb. 23, 1936, at 17 (noting general agreement among observers that the Guild's program cut piracy by 75 percent for higher-end dresses, and by 40 to 50 percent for midrange dresses). 192. See, e.g., Complete Text of Master's Report That Upholds FOGA's Style Protectionas No Monopoly, WOMEN'S WEAR DAILY, Nov. 10, 1936, at 8, 10, 39 (reprinting special master's finding, in rejecting a private antitrust challenge to the Guild, that the Guild caused some copyists to shift to origination); FashionGuild Policy Held Aid to Industry, N.Y. TIMES, June 4, 1936, at 34 (reporting testimony that the Guild had caused many former copyists to change policy without going out of business); see also Dress Trade Urged To Curb "Unethical," N.Y. TIMES, June 3, 1936, at 32 (similar); Dress War, TIME, Mar. 23, 1936, at 72 (Guild caused manufacturers of high-end dresses to begin "to do their own designing, confident that style piracy had been effectively outlawed"; moreover, as retailers returned copied dresses in a lower price range, "a number of manufacturers of these dresses, hitherto generally committed to copying higher priced dresses for a good proportion of their styles, decided that it was time to originate," and became Guild affiliates). For an earlier suggestion that the Guild offers a valuable natural experiment in evaluating design protection for fashion, see Randal C. Picker, Of Piratesand Puffy Shirts, VA. L. REV. IN BRIEF (2007), http://virginialawreview.org/inbrief.php?s=inbrief&p=2007/01/22/picker. 193. For exemplary analyses, see EDWARD H. CHAMBERLIN, THE THEORY OF MONOPOLISTIC COMPETITION (1933); Avinash K. Dixit & Joseph E. Stiglitz, Monopolistic Competition and Optimum Product Diversity, 67 AM. ECON. REV. 297 (1977); A. Michael Spence, Product Differentiationand Welfare, 66 AM. EcoN. REV. 407 (1976); see also N. Gregory Mankiw & Michael D. Whinston, Free Entry andSocial Inefficiency, 17 RAND J. ECON. 48 (1986) (making the excess entry point without relying upon product differentiation). 194. For an argument along these lines, see Michael Abramowicz, An Industrial Organization Approach to Copyright Law, 46 WM. & MARY L. REV. 33, 35-45 (2004); Michael J. Meurer, CopyrightLaw and PriceDiscrimination,23 CARDOZO L. REV. 55, 96- 97 (2001) (noting "over-harvesting" and "distraction" costs from production of close substitutes); see also Christopher S. Yoo, CopyrightandProductDifferentiation,79 N.Y.U. L. REV. 212, 260-64 (2004). 1194 [Vol. 61:1147 A final critique of fashion to revisit is that, assuming fashion is a statusseeking quest, then actions that further its spread might also raise its cost, by leaving an individual to choose between the disutility of falling behind and the social waste that accompanies catching up. 195 This concern about the status function of fashion actually supports our proposal. The primary markers of status-trademark and brand image-will exist with or without design protection. Our proposal gives protection to designs that may lack a strong status component, thereby facilitating the shift of fashion away from the status function and toward the diverse innovation we value in other creative industries. CONCLUSION The amount and kind of innovation in fashion is directly connected to its meaning-making function. We have thus directed our analysis to the role that intellectual property law can play in shaping that process through regulation of an important industry whose products are some of the most immediate means whereby people create and communicate meaning, about themselves and society. Our proposed design right would extend protection against close copies but not against looser forms of borrowing or similarity. It aims to promote innovation by allowing fashion producers and consumers to fully engage these complementary values of distinctiveness and belonging. These coexisting poles provide a key to the social dynamic of innovation. What is basic to all innovation is the constant tension and interplay between individual distinctiveness embodied in creative work and the relation of that work to others, past and present. Whether in books, music, or films, a core social dynamic of innovation is the proliferation of difference in deep interaction with the impulse to commonality. Especially visible in fashion, this dynamic pervades all areas of innovation and is instructive for intellectual property. Our analysis offashion puts into relief the contours ofan important fight in innovation policy. New copying technology alters the dynamics of innovation. In recent years, we have seen how digital file sharing of copyrighted music has changed the economics of that industry. The same is increasingly true of movies and other video content. In fashion, as in other industries, we see rapid copying becoming cheaper and more effective, and tools that enable remixing and reuse are becoming more widespread. The broad conceptual problem is that the two phenomena of copying and remixing have been conflated in the public mind, and proponents of a remix culture are reflexively associated with a permissive attitude toward copying. 196 195. For an account that emphasizes such waste, understanding fashion as a quest for the attainment of personal relative advantage, see ROBERT FRANK, LuXURY FEVER: WHY MONEY FAILS TO SATISFY INAN ERA OF EXCESS 158, 196 (2001). 196. Compare Lawrence Lessig, Essay, In Defense of Piracy, WALL ST. J., Oct. 11, March 2009] FASHION 1195 STANFORD LA WREVIEW In part this is because content owners often oppose both kinds of permissionthat is, they oppose both exact copies and subsequent interpretation, homage, and mash-up. And to be sure, some scholars and advocates favor both remix and free copying. Our analysis of fashion here highlights the need for conceptual distinction between the two phenomena in the debate about how much intellectual property protection we want to have. There is no necessary confluence or equation between a broad freedom to engage in reinterpretation and remixing, and free rein to make close copies. Here we have emphasized that such remixing is important to innovation, and that innovation is enhanced-not stymied-by protection against close copies. We believe that the line between close copying and remixing, supported by the theory of their differential effects on creators' incentives, represents an often underappreciated but most promising and urgent direction for intellectual property today. The dynamics of innovation in fashion design is a window to this important aspect of innovation generally. Our work here is intended to help ensure that free interpretation is preserved, even if free copying is not. 2008, at W3 (arguing in favor of a robust remix right for music and video), with Lessig Blog http://essig.org/blog/2008/10/newsflash-i-dontdefend-pirac.html (Oct. 13, 2008, 16:14 EST) ("News Flash: I don't 'defen[d] piracy'; "Sony to disappoint, but my new book, Remix, is not 'A Defense of Piracy,' whatever the Wall Street Journal's headline writers may think."). Lessig may have been taken for a defender of piracy not only because of his support of remixing, but also because he proposes to legalize file sharing and compensate creators by alternate means, such as a government levy on file sharing devices and services. See LESSIG, supra note 18, at 271-72. For a full analysis of one such proposal, see WILLIAM W. FISHER, PROMISES TO KEEP: TECHNOLOGY, LAW AND THE FUTURE OF ENTERTAINMENT ch. 6 (2004) (proposing compensation system whereby users buy the right to freely share files, and artists are compensated through a blanket licensing procedure). 1196 (Vol. 61:1147 March 2009] APPENDIX Figure 1. Foley & Corinna and Forever 21197 197. Adventures in Copyright Infringement, PartSix, FASHIONISTA, Apr. 12, 2007, http://fashionista.com/2007/04/ adventures-in copyrightinfrin_3.php (Foley & Corinna and Forever 21 comparison). 1197FASHION STANFORD LAW REVIEW Figure 2. Jonathan Saunders and Forever 21198 198. Dorielle Hammonds, We Love: Forever 21, LA2DAY, Aug 12, 2008, http://www.la2day.com/fashion/welove-forever_21 (Jonathan Saunders and Forever 21 comparison). 1198 [Vol. 61:1147 March 2009] Figure 3. Yves Saint Laurent and Ralph Lauren199 199. Profils d'hiver, L'OFFICIEL DE LA MODE, Sept. 1992, at 210, 211 (Yves Saint Laurent, left); and Femmes en smoking, supranote 165, at 138 (Ralph Lauren, right). FASHION 1199 STANFORD LAW REVIEW [Vol. 61:11471200