Fair use is one of modern law’s most fascinating and troubling doctrines. It is amorphous and vague, as well as notoriously difficult to apply. It is, at the same time, vitally important in copyright and perhaps the most frequently raised and litigated issue in the law of intellectual property.

This Article synthesizes themes from the rich literature on fair use to fashion a novel theory of fair use that provides both a better understanding of the underlying principles and better tools for applying the doctrine.

In contrast to the dominant understanding of fair use in the literature—that fair use addresses market failures—the Article proposes that fair use is a tool that allocates a large bloc of uses directly to the public in order to limit the size of property rights that are granted to authors. The fair use doctrine, we argue, is an integral part of copyright’s sorting mechanism for, on the one hand, granting authors intellectual property rights based on their expected incentives for creation and, on the other hand, granting the public privileges based on the expected utility from direct allocation. The Article’s theory thus accords with recent Supreme Court cases by conceptualizing fair use not as an exception for costly transactions, but rather as a central feature of the copyright system that ensures productive and allocative efficiency.

This theory supports a reconceptualization of the basic structure of copyright law that both broadens fair use and makes the doctrine easier to apply. This Article favors a prima facie finding of fair use whenever the user’s category of use is one that produces widespread follow-on utility to nonusers (such as the categories of political speech or what we call “truth seeking”). This prima facie finding can be defeated only by showing that allowing such uses with respect to the particular copyrighted work would eliminate sufficient incentives for its creation.


Fair use is one of modern law’s most fascinating and troubling doctrines. It is amorphous and vague, and it is notoriously difficult to apply.1 At the same time, fair use is vitally important in copyright. The fair use defense to copyright infringement is perhaps the most frequently raised and litigated defense in the law of intellectual property.2 While fair use appears as the first in a series of defenses and mandatory licenses in the Copyright Act of 1976,3 it is seen as much more than that. Fair use is a keystone of the law of copyright; the Supreme Court has repeatedly stated that fair use is a constitutionally mandated limitation on copyright in order to avoid conflicts between the First Amendment’s free speech protections and the monopoly rights copyright owners receive over expressions.4

Yet scholars have struggled to explain why and how fair use should work. Following Professor Wendy Gordon’s immensely influential article, the most dominant theory of fair use is inextricably related to the notion of market failure.5 The market-failure approach argues that the fair use doctrine is best understood as a mechanism for enabling the use of copyrighted works without authorization when the cost of transacting to obtain the authorization is prohibitive and would block voluntary transactions.6 Fair use, in other words, is a doctrine for small uses—those on which the user and owner would agree but for the high cost of locating one another and negotiating an agreement. Gordon has since significantly developed her own positions on fair use,7 but the ingenious explanation in her original article has captured the field.8 Unfortunately, the now-standard justification suffers from several troubling features that have been only partly addressed by subsequent writings.

One problem with the market-failure explanation is descriptive. The market-failure theory characterizes fair use as an exception to—even an anomaly within—the rule that copyright owners have the right to charge for all uses (unless, of course, transaction costs prevent them from doing so). This diverges from judicial understandings of fair use that have made fair use central to the constitutional balance between copyright and free speech.9 This gap between the market-failure theory and the perceived judicial importance of fair use is likely to become more pronounced over time. As technology lowers transaction costs,10 there is less reason to recognize fair uses, dooming the market-failure explanation of fair use to eternal contraction.

A second, and more fundamental, problem with the market-failure theory is conceptual. As we show, the market-failure theory implicitly presumes that the efficient economic strategy for allocating copyright rights is to grant them in their entirety to the author. However, we show that, for intellectual property, productive efficiency may be best ensured by allocating many use privileges directly to users. Indeed, the most efficient allocation of copyright includes the grant of fair use privileges to users even when transactions for such uses would not fall prey to market failure.11

In this Article, we offer a novel theory of fair use based on this insight into the effects of allocation on productive efficiency and societal welfare. Our theory provides a basis for a robust fair use doctrine that protects the public interest in free speech and maintains the incentive structure of copyright law that is vital to the creation of expressive works. It has the added benefit of reducing, though not entirely eliminating, the ambiguity that has attended fair use cases since the doctrine’s inception.

Rejecting the conventional wisdom on fair use, we claim that copyright law should be viewed as granting not one, but two large blocs of legal protections: a grant of fair use privileges to the public and a grant of exclusive rights to authors. The grant of exclusive rights to authors (such as the rights to copy, distribute, and display) is intended to give authors the ability to profit enough from their expressions to make it worthwhile for them to continue creating.12 The grant of fair use privileges to the public, by contrast, is intended to expand use of the creations by giving the public the privilege of utilizing creative expressions for uses that create significant follow-on utility for nonusers (such as political speech and truth seeking).13 Thus, fair use is not simply an allocation of rights for minor uses. It is a grant of privileges as fundamentally important to the aims of copyright as the grant of rights to the author. In our conception, one should not look at fair use privileges as simply residual or carved out from authors’ rights. It is just as valid to view authors’ rights as residual or carved out from the fair use privilege.

This dual grant of rights to authors and privileges to users is designed to meet the dual goals of incentivizing authors to create works and encouraging efficient use of works by users. Unfortunately, the dual constitutional aims of incentivizing creation of new works and ensuring their use by the public exist in some tension. Each grant of rights to the author of a copyrighted work expands the domain of rights enjoyed by the copyright owner (increasing incentives) while simultaneously restricting the use privileges of the public in that same expression. Likewise, each grant of privileges to users increases the use of works while curbing the value of incentives granted to the authors. In deciding whether to grant any particular right to authors, lawmakers must take account of this double effect, balancing the positive effect of the increased marginal incentives for authors that are created by the grant with the negative effect of the losses of public use of the expression in accordance with the right. Conversely, each grant of a privilege to the public to use otherwise-protected expressions requires lawmakers to balance in the opposite direction: the privilege expands public use at the expense of marginal incentive effects for creation.

An important predicate of our theory is that incentivizing creativity does not necessitate granting all possible exclusive rights to authors in perpetuity. Granting this package to authors would actually result in a net social loss.14 In a world with positive transaction costs, exclusive rights to authors, by necessity, eliminate some kinds of uses by nonauthors and future authors.15 At the same time, infinite protection is not necessary to incentivize the creation of most works, and therefore it is not desirable. Indeed, the more protection granted to authors, the more likely it is that society is overpaying. Hence, the optimal incentive structure involves giving only some rights and powers to authors—just enough to motivate the creation of original works—while reserving the remaining rights and powers to the public.16 We contend that fair use is one of the key means by which this division should be effected. Fair use helps filter protections to ensure efficient allocation of uses to societally favored users while still fully maintaining the incentive effects of copyright protection for authors.

A second important predicate of our theory is the insight that, absent the need to incentivize production of expressive works, it would be best to allow the public full and unfettered use of all expressions, irrespective of their origin. This is due to the fact that expressive works are nonrivalrous in their consumption.17 That is to say, the use of an expressive work by any particular consumer does not diminish in any way the ability of another user to consume it. A million people can read William Shakespeare’s Hamlet, and the words will remain undiminished for the millionth-and-first reader.

Given this state of affairs, the optimal fair use doctrine would limit rights strictly to those necessary to incentivize creation, while leaving the public to consume copyrighted works without restriction beyond that minimum. Ordinarily, we think of this balance as being struck by giving authors time-limited exclusive rights and giving the public unlimited use of the works once those rights have expired. But, as we show, there is actually a better way to reach the balance. The law can grant rights to authors that are prima facie larger than necessary to incentivize, while at the same time cutting away from those rights by allocating extensive use privileges to the public. Allocating use privileges directly to users can produce significant societal benefits, especially when the privileges focus on uses that produce significant follow-on benefits to other members of the public. Such an allocation not only ensures the exercise of use privileges by those who find them valuable, saving the cost of transferring those rights from owner to user, but it also ensures the benefits of uses that create positive externalities for the nonconsuming public. Allocating the uses directly to users also reduces ancillary costs, such as search costs that are saved by users’ self-selection.

Accordingly, we posit that the fair use doctrine should be primarily concerned with preserving public privileges in uses that themselves produce the most widely spread benefits to the public. Specifically, the fair use doctrine should enable consumers to utilize expressive works without the consent of the author when the uses create widespread follow-on benefits throughout society.

The key to creating a viable dual-grant strategy of copyright, of course, is producing a workable dividing line between privileged uses and protected rights. In considering the allocation of authors’ rights and public privileges, it must be borne in mind that not all uses of expressive works are created equal. Some uses of expressive works bring about broad-based benefits to society. For instance, as scholarship and discussion about the First Amendment observes, the Constitution expects the state to be particularly solicitous of political speech.18 Likewise, the Copyright Clause19 reveals a societal interest in encouraging science and the pursuit of knowledge. Stated otherwise, the state seeks to promote the pursuit of truth20 as a valuable goal in and of itself.21 Other categories, such as education (“teaching,” in the terminology of the Copyright Act), are found directly in § 107 of the Act.22

Accordingly, uses of highly dispersed social value should be considered presumptively public under the auspices of the fair use doctrine, and only in extreme cases should the public be deprived of them. By contrast, other types of uses—first and foremost, standard commercial uses—should be considered to be within the scope of the rights granted exclusively to authors. In other words, in the case of standard commercial uses, the reverse legal presumption should obtain—that is, unauthorized uses are typically not fair, unless there are special circumstances that justify denying exclusivity to authors, such as market failure. Here, the value to the public of the uses is likely to be directly translated into benefits to authors, and the standard copyright protections should obtain.

Our reconceptualization of fair use has a significant doctrinal payoff. Most importantly, our fair use conception requires courts applying fair use to be sensitive to the degree of copyright protection offered more generally. An expansion of rights granted to authors, such as the extension of copyright protection for an additional twenty years,23 should be matched by an expansion of fair use privileges in order to once again balance the competing impulses of copyright law.

To implement our proposed conception of fair use, we propose a new doctrinal mechanism for identifying and upholding fair uses. Under our conceptualization of fair use, whenever the intent behind any use of an expressive work is for one of the aims of fair use, the use should be considered prima facie “fair” and permitted. Our definition of inherently public uses encompasses several of the presumptively fair uses in the preamble of § 107 of the Copyright Act, such as research and news reporting.24 To us, political speech, the pursuit of factual accuracy, and scientific advancement should be put on equal footing with, if not above, the illustrative uses in § 107. Within the ambit of truth seeking, we would also grant privileged status to new technological projects aimed at the enhancement and spread of knowledge, such as Google Books,25 even though they are not listed in the preamble of § 107.

At the same time, we maintain that some of the uses des­cribed as favored in fair use jurisprudence, such as parodic uses, need not be considered presumptively “fair.” More generally, we suggest as a rule of thumb that any categories of uses that create significant nonpecuniary benefits to follow-on users (that is, subsequent consumers of the expression that will be utilizing the use that is now claimed to be “fair”) should be considered fair uses.

The dual-grant theory of fair use that we advocate brings some needed coherence to the doctrine of fair use. It provides a better account of some of the key themes in recent case law, while at the same time suggesting that different results should have been reached in several of the cases. More importantly, we show how our approach can reduce some of the ambiguity surrounding fair use decisions, and we lend greater clarity to the statutory framework.

Our Article unfolds in four parts. In Part I, we explore current understandings of fair use and, in particular, the dominant market-failure theory. In Part II, we present our contribution to fair use theory by offering an alternative understanding of the fair use doctrine, focusing, in particular, on how fair use is an efficient allocation of use privileges to the public at large. In Part III, we present our contribution to fair use doctrine by proposing alterations that implement our understanding of the role of fair use. Finally, in Part IV, we explore the interface between our theory of fair use and other writings on copyright, property, and free speech.

  • 1. See Michael W. Carroll, Fixing Fair Use, 85 NC L Rev 1087, 1093–95 (2007); Barton Beebe, An Empirical Study of U.S. Copyright Fair Use Opinions, 1978–2005, 156 U Pa L Rev 549, 551–53 (2008).
  • 2. Paul Siegel, Communication Law in America 237 (Rowman & Littlefield 2014). See also David Nimmer, “Fairest of Them All” and Other Fairy Tales of Fair Use, 66 L & Contemp Probs 263, 266–67 (Winter/Spring 2003).
  • 3. Pub L No 94-553, 90 Stat 2541, codified as amended at 17 USC § 101 et seq. For the statutory section on fair use, see 17 USC § 107.
  • 4. See, for example, Eldred v Ashcroft, 537 US 186, 219–20 (2003) (describing fair use as a “built-in First Amendment accommodation[ ]”); Harper & Row, Publishers, Inc v Nation Enterprises, 471 US 539, 555–60 (1985) (describing the “First Amendment protections . . . embodied in . . . the latitude for scholarship and comment traditionally afforded by fair use”).
  • 5. See generally Wendy J. Gordon, Fair Use as Market Failure: A Structural and Economic Analysis of the Betamax Case and Its Predecessors, 82 Colum L Rev 1600 (1982).
  • 6. See generally id.
  • 7. See generally, for example, Wendy J. Gordon, Excuse and Justification in the Law of Fair Use: Transaction Costs Have Always Been Part of the Story, 50 J Copyright Society USA 149 (2003) (describing the specific instances when market failure should justify a finding of fair use); Wendy J. Gordon, Market Failure and Intellectual Property: A Res­ponse to Professor Lunney, 82 BU L Rev 1031 (2002) (describing how the original market-failure test for fair use has been misinterpreted as a mechanism to limit fair use).
  • 8. For some illustrative examples of the application of this framework, see generally Brad A. Greenberg, Copyright Trolls and Presumptively Fair Uses, 85 U Colo L Rev 53 (2014); Christina Chung, Note, Holy Fandom, Batman! Commercial Fan Works, Fair Use, and the Economics of Complements and Market Failure, 19 BU J Sci & Tech L 367 (2013); Jennifer M. Urban, How Fair Use Can Help Solve the Orphan Works Problem, 27 Berkeley Tech L J 1379 (2012); Ryan Kairalla, Work as Weapon, Author as Target: Why Parodies That Target Authors (Not Just Their Works) Should Be Fair Uses, 2 NYU J Intel Prop & Enter L 227 (2012); Matthew Sag, Predicting Fair Use, 73 Ohio St L J 47 (2012); Lydia Pallas Loren, Redefining the Market Failure Approach to Fair Use in an Era of Copyright Permission Systems, 5 J Intel Prop L 1 (1997).
  • 9. See Sarl Louis Feraud International v Viewfinder, Inc, 489 F3d 474, 482 (2d Cir 2007) (discussing how fair use “balances the competing interests of the copyright laws and the First Amendment”); Eldred, 537 US at 220 (noting that fair use doctrine is one of copyright’s “traditional First Amendment safeguards”); A&M Records, Inc v Napster, Inc, 239 F3d 1004, 1028 (9th Cir 2001) (“We note that First Amendment concerns in copyright are allayed by the presence of the fair use doctrine.”); Nihon Keizai Shimbun, Inc v Comline Business Data, Inc, 166 F3d 65, 74–75 (2d Cir 1999) (“First Amendment concerns [relating to copyright injunctions] are protected by and coextensive with fair use doctrine.”); Maxtone-Graham v Burtchaell, 631 F Supp 1432, 1435 (SDNY 1986) (“To satisfy [ ] First Amendment requirements, [c]ourts have mitigated the chilling effects of copyright law upon free expression . . . by developing . . . Fair Use.”).
  • 10. Claudio U. Ciborra, Teams, Markets and Systems: Business Innovation and Information Technology 74–76 (Cambridge 1993). See also generally Antonio Cordella, Transaction Costs and Information Systems: Does IT Add Up?, 21 J Info Tech 195 (2006).
  • 11. See generally Glynn S. Lunney Jr, Fair Use and Market Failure: Sony Revisited, 82 BU L Rev 975 (2002).
  • 12. Justin Hughes, Fair Use across Time, 50 UCLA L Rev 775, 797 (2003); Joseph P. Liu, Copyright and Time: A Proposal, 101 Mich L Rev 409, 428 (2002); Stewart E. Sterk, Rhetoric and Reality in Copyright Law, 94 Mich L Rev 1197, 1198–1204 (1996).
  • 13. The terminology of rights and privileges is taken from the classic typology of Professor Wesley Hohfeld. See generally Wesley Newcomb Hohfeld, Some Fundamental Legal Conceptions as Applied in Judicial Reasoning, 23 Yale L J 16 (1913).
  • 14. Oren Bracha and Talha Syed, Beyond the Incentive–Access Paradigm? Product Differentiation & Copyright Revisited, 92 Tex L Rev 1841, 1884–85 (2014); Shyamkrishna Balganesh, Foreseeability and Copyright Incentives, 122 Harv L Rev 1569, 1589–91 (2009).
  • 15. See Gordon, 82 Colum L Rev at 1613 (cited in note 5).
  • 16. William M. Landes and Richard A. Posner, An Economic Analysis of Copyright Law, 18 J Legal Stud 325, 341–44 (1989).
  • 17. Lunney, 82 BU L Rev at 994–95 (cited in note 11).
  • 18. See Lillian R. BeVier, The First Amendment and Political Speech: An Inquiry into the Substance and Limits of Principle, 30 Stan L Rev 299, 304–22 (1978) (analyzing a variety of sources to conclude that the First Amendment, at a minimum, “protects the process of forming and expressing the will of the majority according to which our representatives must govern”).
  • 19. US Const Art I, § 8, cl 8. We refer to this clause as the “Copyright Clause,” as we are interested in only its reference to copyright law. The Clause refers to patent law as well, and it might more accurately be labeled the “Intellectual Property Clause.”
  • 20. See Hustler Magazine, Inc v Falwell, 485 US 46, 52 (1988) (stressing the significance of the “truth-seeking function” of the First Amendment); Patrick Garry, The First Amendment and Freedom of the Press: A Revised Approach to the Marketplace of Ideas Concept, 72 Marq L Rev 187, 197 (1989); Stanley Ingber, The Marketplace of Ideas: A Legitimizing Myth, 1984 Duke L J 1, 6–8.
  • 21. In this Article, we take these constitutional values as givens.
  • 22. 17 USC § 107 (“[T]he fair use of a copyrighted work, including . . . for purposes such as criticism, comment, news reporting, teaching[,] . . . scholarship, or research, is not an infringement of copyright.”).
  • 23. Such an extension was implemented by the Sonny Bono Copyright Term Extension Act, Pub L No 105-298, 112 Stat 2827 (1998), codified as amended at 17 USC §§ 108, 203, 301–04.
  • 24. 17 USC § 107.
  • 25. See Part III.C.2.b.