Consent, user reliance, and fair use.
INTRODUCTION I. FAIR USE BASICS II. THE STANDARD VIEW: CONSENT IS IRRELEVANT A. The Seeming Logic of the Standard View B. Critiques of the Standard View III. THE CASE FOR CONSENT AS A FAIR USE FACTOR A. Efficiency and Incentives B. The Equitable Component of Fair Use C. Summary IV. A TYPOLOGY OF CONSENT AND REFUSAL A. Quantitative Dimension B. Qualitative Dimension V. CASE STUDIES A. Opt Out Systems and Information Aggregation B. Partial Consent and User Reliance Interests C. Impossibility and Orphan Works D. "Purloined" Copies and Bad Faith VI. COSTS AND CONCERNS A. Unpredictability in Application B. Judicial and Evidentiary Costs C. Inhibiting "Tolerated Uses" D. Interaction with Implied License and Other Doctrines CONCLUSION
Copyright law's fair use doctrine relies on what might be termed a binary notion of consent and refusal. If a copyright holder has given affirmative consent to a use of her work, then the use is permitted as authorized. All other cases short of affirmative consent are treated equivalently--as a refusal. As a formal matter, the fair use inquiry thus ignores the qualitative difference between, say, an express rejection of a proposed use on the one hand, and mere silence from a copyright holder on the other. This deficiency is odd given that fair use was historically described as premised on the "implied consent" of the copyright owner to reasonable uses. (1) As fair use doctrine has evolved, however, this notion has been discarded as a legal fiction. (2) Instead, judicial attention has focused on a mechanical application of the four statutory fair use factors. (3) Under the prevailing view of fair use, unless a copyright holder gives actual consent to a use, her prior conduct indicating approval or disapproval, and the reasons for her ultimate refusal, are irrelevant. (4)
Two examples may help illustrate the state of the law. Consider first the pending legal challenge to the Google Books project. (5) That case presents the question of whether Google Books--which makes the text of millions of books digitally searchable--should be permitted as a fair use, or enjoined as an infringement of authors' copyrights. (6) In constructing this massive information aggregation system, Google takes some care to accommodate differing levels of consent from copyright holders. Books in the public domain are available for download in their entirety. For in-copyright works, the default is to allow a "snippet" view: The full text is searchable, but only a line or two of text around the search term is displayed in response to a search query. (7) If the copyright holder allows it, Google will show more of the work, perhaps a "preview" of several pages from the book. (8) In contrast, if the copyright holder specifically objects, Google will not display even the snippets. (9)
The significance of this "opt out" structure presents something of a puzzle. Intuitively, if the fair use inquiry reflects an assessment of the social value of a use, the opt out functionality seems relevant. It provides a way to realize the benefits of the Google Books project while still allowing objecting rightsholders to easily choose not to participate. But if Google Books' snippets are a fair use, Google is free to override the wishes of the copyright holder. It should not matter whether a given author objects, or had an earlier opportunity to opt out of the project. on the standard view of fair use, Google's good-faith attempt to accommodate competing interests by providing an opt out is irrelevant to whether its use is fair. (10) Indeed, in reaching its conclusion that Google Books was a fair use, a recent court decision completely ignored this opt out structure. (11)
For a more analog example, consider the facts of Peter Letterese & Associates v. World Institute of Scientology Enterprises (12) In that case, the user--the Church of Scientology--had adapted aspects of a copyrighted sales manual into its training manuals. (13) This use was arguably fair--very little expression was taken--but in any event, it was made with the participation and involvement of both the original author and the copyright holder, and continued for decades without objection. (14) After a falling out between the copyright holder and user, however, an infringement action followed. The district court thought that consent was relevant to the fairness of the use. (15) The user, after all, relied on the apparent permission from the copyright holder in initially making the use, and because aspects of the original work were now incorporated into a larger whole, "undoing" the use decades later could be difficult and costly. The Eleventh Circuit, however, relied on the standard view of consent in fair use to disregard this context. (16)
This Article challenges the standard view of the role of consent in fair use cases. It argues that the form and nature of a copyright holder's refusal to consent should not always be ignored in the fair use analysis. This Article claims that consideration of consent, in certain classes of cases, is normatively desirable and encourages efficient communication between users and copyright holders. Accordingly, it urges courts to make the nature of the copyright holder's consent an explicit consideration.
To be clear, consent should not be the sole consideration in the fair use analysis, or even a factor at all in many cases. In the usual case, where there is an express, good-faith refusal from a copyright holder, the user's failure to obtain consent is not of any relevance. Fair use is designed, after all, to define the types of uses that should be permitted in spite of the copyright holder's refusal. But this does not foreclose the converse possibility--that a copyright holder's apparent consent ought to weigh in favor of fair use. As this Article argues, the nature of the copyright holder's refusal to consent has a critical role to play in some fair use cases. In particular, consent need not be ignored when it would be otherwise relevant to the traditional fair use factors. The wide-ranging, fact-sensitive fair use inquiry can readily accommodate this additional consideration.
Applying this proposal to the examples above, the fact that Google Books provides copyright holders with low-cost mechanisms to opt out--and that a copyright holder neglected to opt out despite awareness of the option--should tend to weigh in favor of fair use. Similarly, in the Letterese case, the fact that the use was created in reliance on apparent consent from the rightsholder should permit a broader scope for fair use. Ex ante, applying fair use doctrine in this matter will enable potential users to rely on external appearances of consent, even if they fall short of the exacting requirements for a binding implied license. (17) It will operate, in effect, to put a somewhat greater burden on copyright holders to affirmatively assert their rights, which are granted by default. (18)
Despite its normative appeal, the role of consent and refusal has largely gone unrecognized in the extensive commentary on the fair use doctrine. (19) Two interrelated reasons may explain why. First, as discussed above, the reasons for the user's failure to obtain consent, and the copyright holder's reasons for his refusal, are irrelevant as a matter of black letter law. (20) Second, the fair use inquiry primarily takes the perspective of the copyright user in evaluating the fairness of the use at issue: Was the alleged infringer's use "transformative"? Was the unauthorized use made for a commercial, noncommercial, or educational purpose? How much of the original work was used? (21) By contrast, the perspective taken in this Article focuses part of the inquiry on the copyright holder's actions: Why is she not consenting to the use? Did the alleged infringer ask her? Did she have an opportunity to object to the use beforehand?
The remainder of this Article will proceed as follows. Part I provides the necessary background on copyright's fair use defense, presenting the four statutory "fair use factors" and the significant judicial interpretations of the same. Part II critiques the standard view that the nature of the user's failure to obtain consent is irrelevant to the fair use inquiry. The standard view maintains that consent cannot be a fair use factor because the user's failure to obtain consent is, by definition, present in every fair use case. But this logic holds only if consent is a binary notion: either the user has affirmatively obtained consent, or he has not. As this Article argues, this binary elides a wide variety of significantly distinct conduct. Mere silence from a copyright holder, for example, is qualitatively different from an affirmative expression of approval or disapproval. A user may fail to obtain consent prior to his use because he is unable to identify the copyright owner, because he believes his use to be fair in good faith, or because he chose not to ask permission in order to avoid detection. There is no reason that the fair use inquiry should treat all these situations identically. Moreover, the standard view fails to appreciate the ways in which these intermediate varieties of consent may be relevant to the traditional fair use factors. The standard view has therefore contributed to confusion in the case law, with some courts implicitly considering consent despite its formal irrelevance.
Part Ill turns to the affirmative case that the nature of consent and refusal should be considered as a fair use factor in some cases. The argument has two key components. The first sounds in efficiency: consideration of the nature of consent encourages ex ante preference revelation by copyright holders and avoids costly ex post disputes. Current doctrine creates incentives for both users and copyright holders to act strategically in their consent-seeking interactions. For example, a copyright owner may fail to express his nonconsent early on, despite an opportunity to do so, springing up only later when the use has proven valuable and the user is locked in. Consideration of consent as a fair use factor punishes the copyright holder for failing to make her preferences known--in effect, her silence is treated as an implied assent--and thus encourages her to object when she has the opportunity. Similarly, a user who knows his use to be unfair should be discouraged from failing to request permission in order to avoid detection of that use. Considering consent can therefore operate as a preference-revealing mechanism that deters strategic conduct and limits costly ex post disputes by encouraging resolution of the status of the copyright holder's preferences prior to the use.
The second argument in favor of the proposal is based on history and the normative purposes of fair use. In short, consideration of the nature of the copyright holder's refusal, when relevant, is more faithful than the current rule to fair use's nature as an "equitable rule of reason." (22) That is, while fair use analysis is primarily concerned with factors relating to the nature of the use itself, that should not necessarily end the inquiry. Reliance, good faith, and other equitable factors relating more to the litigants than to the use can also be important in deciding what is "fair." In other words, while transaction costs, a just intellectual culture, transformative uses, and the spillover social benefits of the use are important, (23) courts should not neglect the fact that fair use is also about courts deciding cases in an equitable manner. One of fair use's virtues is its flexible, contextual, and fact-sensitive nature. (24) It is thus a feature, and not a flaw, of this Article's proposal that an identical use may be permitted in one context (e.g., when the use was made in reliance on the acquiescence of the copyright holder), but not another (when that same use was made, but no permission was sought in order to avoid detection).
Turning from theory to application, Part IV makes the proposal more concrete by developing a typology of the consent-seeking interactions between the copyright owner and potential users. Part V applies this typology to examine the role that consent plays in a variety of fair use cases.
It builds on recent work by Pamela Samuelson and Michael Madison that considers fair use cases in common patterns of "policy-relevant clusters." (25) Looking to categories of fair use cases involving user reliance and strategic behaviors, this Part critiques those courts that have ignored the nature of the copyright holder's consent. It also reveals that some courts have implicitly considered the nature of consent in their fair use analysis, despite the black letter law to the contrary. In those cases, the proposed consent framework offers a satisfying explanation of otherwise-perplexing features of fair use doctrine.
Part VI addresses potential criticisms of this Article's proposal. The Article concludes that consistency in judicial decisionmaking, as well as sound policy, would be better served by considering the nature of the copyright holder's consent or refusal in cases involving user reliance interests and strategic behavior.
I. FAIR USE BASICS
As many scholars have observed, the basic design of the Copyright Act favors protection of works: the Act pairs broad statements of the copyright owner's exclusive rights with narrow, detailed exceptions to those broad rights. (26) In this sense, the Copyright Act's default rule may be said to favor the copyright holder--once a user, for example, makes a derivative work based upon a preexisting copyrighted work, he must usually satisfy the elements of a specific statutory exception to avoid liability. (27)
The fair use doctrine, as a broad limitation on copyright's reach, is the principal exception to that general structure. Fair use was initially developed in a common law fashion by the courts as an implied exception for "reasonable" and "customary" uses; (28) it was not codified until the 1976 general Copyright Act revision. The main debate surrounding the 1976 codification of fair use centered less on the precise statutory language, and more on whether the Act should avoid congressional interference with the doctrine and decline to reference fair use at all. (29) Section 107 of the Copyright Act split the difference by codifying the fair use doctrine in general terms, but expressly stating that Congress did not intend to "freeze" the judicial development of fair use. (30) The provision reads in full:
Notwithstanding the [exclusive rights of copyright holders], the fair use of a copyrighted work, including such use by reproduction in copies or phonorecords or by any other means specified by that section, for purposes such as criticism, comment, news reporting, teaching (including multiple copies for classroom use), scholarship, or research, is not an infringement of copyright. In determining whether the use made of a work in any particular case is a fair use the factors to be considered shall include--(1) the purpose and character of the use, including whether such use is of a commercial nature or is for nonprofit educational purposes; (2) the nature of the copyrighted work; (3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and (4) the effect of the use upon the potential market for or value of the copyrighted work. The fact that a work is unpublished shall not itself bar a finding of fair use if such finding is made upon consideration of all the above factors. (31)
Despite the explicit congressional intention that this codification not disrupt the judicial development of the doctrine, the four "fair use factors" have nevertheless become the core of modern fair use doctrine, with many courts mechanically applying statutory considerations as a four factor test. (32) While a detailed discussion of each of the four factors and the judicial interpretations of them is beyond the scope of this Article and may be found elsewhere, (33) an overview of the key considerations for each factor is in order:
Factor One: The Purpose and Character of the Use. This broad and important consideration captures the intended aim of the use, with educational, newsreporting, critical, and noncommercial uses favored over commercial ones. The first factor also addresses whether a use is "transformative, " i.e., whether it "adds something new, with a further purpose or different character, altering the first with new expression, meaning, or message." (34) Courts have generally afforded "transformativeness" a broad purview beyond literal transformations such as criticisms and parodies. (35) Uses will be considered transformative if they are made for a sufficiently distinct purpose or function--such as a search engine (36) or anti-plagiarism database (37)--even though the content of the original work is not itself altered. Transformative uses, like noncommercial uses, are favored under the first fair use factor.
Factor Two: The Nature of the Copyrighted Work. The second factor relates to the type of copyrighted work at issue. It has essentially two components: whether a work is published or unpublished, and whether a work is creative, as opposed to factual or informational. Unpublished works are afforded broader copyright protection and a narrower fair use defense; (38) in fact, at common law, the use of unpublished works was never fair. (39) In the case of informational works, such as a newspaper article, the fair use defense is broader vis-a-vis creative works like a novel or a painting. (40) Indeed, factual works are generally afforded weaker copyright protection because of the fundamental copyright principle that unoriginal facts are not themselves copyrightable. (41) In short, the second fair use factor disfavors the use of unpublished works and creative works.
Factor Three: The Amount and Substantiality of the Portion Used. Factor three captures the notion that, other things being equal, a use is more likely to be fair if it copies only a small amount from the original. Copying an entire book verbatim is obviously quite different than just quoting a sentence or two. Though intuitively straightforward, two glosses on this factor complicate the analysis. First, how much was copied may be assessed on a qualitative level: its "substantiality." In other words, copying of the "heart of the work"--the main source of its economic value--may be unfair even though only a small amount of actual expression is taken. (42) For example, in Harper & Row v. Nation Enterprises, though the number of words copied from President Ford's memoir was "insubstantial, " the third factor favored the copyright holder because the small part that was taken was the primary matter that readers were interested in: why Ford had pardoned Richard Nixon. (43) The second complication is that even literal copying of an entire work may be fair if the amount taken is necessary for a legitimate purpose, such as a sufficiently transformative use. (44) The third fair use factor thus disfavors excessive copying, unless necessary for a privileged use.
Factor Four: The Effect of the Use on the Potential Market for the Copyrighted Work. The final factor, sometimes considered the most important, (45) relates to whether the use detrimentally impacts the market for the copyright owner's original work. If it has a substantial effect on the market, the use undermines the copyright owner's incentives to create the original and will tend to be found unfair. (46) This principle is subject to some important limitations. First, there is the danger of circularity in considering lost licensing revenue as a market harm: by determining whether a use is fair or unfair, the court is essentially deciding whether a licensing market should exist for this use. (47) The focus, then, should primarily be on whether the use functions as an economic substitute for the original. Second, not all economic harms are properly weighed under this factor. A dismissive book review or biting parody may undermine the market for the original, but they do not create harms that are cognizable under the Copyright Act. (48) The fourth factor only disfavors uses that diminish demand because they serve as substitutes for the original work.
Two features of these statutory factors are noteworthy for present purposes. First, the statutory factors focus primarily on the perspective of the user, not of the copyright holder. Two of the four factors--the purpose and character of the use, and the amount of the original work that was used--are wholly devoted to examining what the copyright user has done. The fourth factor--the effect of the use upon the potential market for the copyrighted work--necessarily relates both to the original and to the use, as it captures an economic interaction between the use and the original. only one factor of the four--the nature of the copyrighted work--wholly examines the issue from the perspective of the copyright holder. This factor, notably, is considered to be the least important fair use factor, and has been found to exert almost no influence on the actual outcomes of fair use cases. (49)
Second, the four statutory factors are nonexclusive by design. (50) Courts have discretion to consider a variety of other factors, and they often do so. For example, courts sometimes look to whether the user attributed the original author--as opposed to passing off copied material as his own--finding that proper attribution tends to favor fair use. (51) Similarly, deliberate distortions or errors that misrepresent the original may disfavor fair use. (52) In general, bad faith by the defendant will weigh against fair use, though this consideration is controversial. (53) The key point is that courts are entitled to extend the fair use inquiry beyond the four statutory factors. In fact, Congress has explicitly endorsed this practice. (54)
II. THE STANDARD VIEW: CONSENT IS IRRELEVANT
As the preceding section demonstrates, fair use is a flexible and wide-ranging inquiry. Despite the capacious nature of fair use, the circumstances surrounding a user's failure to obtain consent from the copyright holder are generally not considered as a matter of black letter law. This section explains and critiques the standard view that rejects the copyright holder's consent as a fair use factor.
A. The Seeming Logic of the Standard View
The logic of the standard view rests on a conflation of all conduct that falls short of actual consent. The argument runs roughly thus: The fair use defense is needed only when the use is unauthorized, for if the user has actual permission for her use, there is no infringement of the copyright and no need for the fair use defense. (55) In every fair use case, then, the user has failed to obtain consent from the copyright holder. How, then, could consent be a factor? It would appear to always cut the same way.
The Eleventh Circuit's opinion in Peter Letterese & Associates v. World Institute of Scientology provides an example of the standard view. (56) The work at issue in Letterese was a 1971 book by Les Dane called Big League Sales, which described sales closing techniques. (57) With Dane's knowledge and participation, Church of Scientology founder L. Ron Hubbard incorporated aspects of Big League Sales into Church seminars and teaching materials in the 1970s and 80s. (58) Peter Letterese, a member of the Church who had learned Dane's techniques at such seminars, acquired the copyright to Big League Sales with the intention of "making [his] own use" of the book as well as "keeping it available to the Church." (59) Mr. Letterese, however, was later excommunicated from the Church, and eventually brought a suit in 2004 against several Scientology entities for copyright infringement when they continued to use their adaptations of material from Dane's book. (60)
The district court in Letterese thought that this course of conduct was relevant to the fair use inquiry and considered "the copyright owner's actual consent" as a fair use factor. The lower court relied the evidence that "both [the author and the copyright owner] knew of, assented to, and participated in Defendants' use for decades, " along with the four traditional fair use factors, to conclude that the Church's use was fair. (61) The Eleventh Circuit rejected the consent consideration:
The district court explicitly adopted a "fifth" [fair use] factor, which it described as "the copyright owner's actual consent to the use" ... This was incorrect, both in terms of logic and precedent. As the district court itself recognized, the existence of actual consent negates the necessity of conducting a fair use analysis in the first place, as the existence of a license is an independent affirmative defense to a claim of copyright infringement. See Acuff-Rose Music, Inc. v. Jostens, Inc., 155 F.3d 140, 144 n. 2 (2d Cir.1998) ("The doctrine of 'fair use' allows the appropriation of a copyrighted work without consent under certain circumstances.") (emphasis added). Actual consent therefore is not properly a factor in fair use analysis. (62)
Courts like Letterese thus reason that because actual consent from the copyright owner negates the need for a fair use analysis, actions falling short of effective consent are irrelevant to fair use.
B. Critiques of the Standard View
1. Consent Is Not a Binary Concept
The first response to this reasoning is that the conclusion simply does not follow: The fact that actual consent negates the need for fair use does not imply that the past conduct of the copyright owner indicating consent is necessarily irrelevant. In some cases, like Letterese, there is a course of dealing that comes very close to actual consent, even though that conduct may fall short of what is necessary for an express or implied license. (63) There is no reason that this sort of "partial consent" must be treated the same as an express refusal. In fact, as this Article argues, there are reasons to consider such gradations of consent as a fair use factor in some cases. (64)
The Supreme Court has not directly addressed the issue of whether the nature of the copyright holder's refusal is properly considered as a fair use factor, but it touched on the matter in its leading fair use decision, Campbell v. Acuff-Rose Music (65) Campbell addressed whether the rap group 2 Live Crew's parody of Roy Orbison's famous song "Oh, Pretty Woman" was a fair use. (66) Before the parody was released for sale, 2 Live Crew's manager approached Acuff-Rose Music (the copyright holder) and offered to pay them a fee for the use. Acuff-Rose categorically refused to license, but 2 Live Crew released the record nonetheless. (67) While the bulk of the Campbell decision addressed other issues, (68) a footnote asked whether 2 Live Crew's request for a license evinced bad faith that should weigh against fair use:
[R]egardless of the weight one might place on the alleged infringer's state of mind ... we reject Acuff-Rose's argument that 2 Live Crew's request for permission to use the original should be weighed against a finding of fair use. Even if good faith were central to fair use, 2 Live Crew's actions do not necessarily suggest that they believed their version was not fair use; the offer may simply have been made in a good-faith effort to avoid this litigation. If the use is otherwise fair, then no permission need be sought or granted. Thus, being denied permission to use a work does not weigh against a finding of fair use. (69)
The Court in Campbell thus rejected the notion that a copyright user's request for permission to make a use should be held against him. But this does not foreclose the converse proposition: that an ex ante request for permission should weigh in favor of the user. This equivocal footnote may explain, however, the persistence of the standard view that the nature of the copyright holder's refusal is irrelevant. (70)
In sum, the logic underlying the standard view rests on a conflation of all conduct short of actual consent by the copyright holder. There is no necessary reason, however, that the fair use doctrine must conceptualize consent and refusal in this way. Instead, one may conceive of consent and refusal as a continuum, with equivocal conduct--such as silence from a copyright holder, or actions indicating but not explicitly granting consent--lying between the extremes of express approval and disapproval. Once this conceptualization is made, there is no a priori reason that such intermediate varieties of consent may not be considered in appropriate cases.
2. Consent's Relevance to the Traditional Fair Use Factors
The standard view is also flawed in that it fails to consider the possibility that intermediate varieties of consent may be relevant to the traditional statutory fair use factors. Consider the cases of "partial consent, " where the copyright holder has implicitly indicated that she approves the use, or silence from a copyright holder despite awareness of a use. In either case, the copyright holder's failure to object to a known use would tend to indicate that, at least in her view, there is little market harm resulting from the use under the fourth fair use factor, which should weigh in favor of fair use. Similarly, if the copyright holder's failure to object was motivated by strategic considerations, this would tend to indicate bad faith, which is sometimes considered under the first fair use factor.
For this reason, perhaps, some courts appear to consider aspects of consent in fair use despite its formal irrelevance. one example is Field v. Google, (71) which challenged Google's caching functionality as a massive copyright infringement. In the process of indexing the Internet's websites for its search engine, Google creates a copy of the material on each website that it temporarily archives, or "caches." (72) This cached version serves several purposes, such as improving the speed of searches and allowing users to access a website that is inaccessible due to heavy traffic or server issues. (73)
Google's caching was alleged to infringe website owners' copyrights because it creates unauthorized, wholesale reproductions of websites. (74) However, the Field court permitted it as a fair use. (75) In reaching that conclusion, the decision noted that Google provides simple mechanisms through which any website owner may easily "opt out" of caching, or of being indexed by Google entirely. (76) Field, the plaintiff, was aware of how to opt out of caching, but chose to make his works publicly available online with the specific intent of filing a copyright action against Google. (77)
on the standard understanding of fair use, as noted above, Google's opt out functionality is irrelevant. (78) Nonetheless, Field's knowing failure to opt out appears to inform the court's analysis in the Field case. This is likely because Field, the plaintiff, acted with obvious bad faith. (79) Though the court mentions opt out functionality only briefly in its discussion of the fair use factors, (80) Field's disingenuous conduct colors its analysis. Similarly, Field's failure to opt out tends to indicate that Google's caching did not cause any harm to the market for his website. (81) Field v. Google thus demonstrates that courts sometimes intuitively consider consent despite the standard view that consent is always irrelevant.
III. THE CASE FOR CONSENT AS A FAIR USE FACTOR
Having disposed of the standard view's notion that consent cannot, as a matter of logic, be considered as a fair use factor, this Part argues that the circumstances surrounding the user's failure to obtain consent from the copyright holder should be considered in fair use analysis. The first Section relies on an economic model of author-user interactions to argue that consideration of the copyright holder's consent or refusal deters costly strategic behavior, and encourages good faith dealing between authors and users. The second Section argues that considering consent accords with fair use's nature as a flexible "equitable rule of reason" that should rightly concern itself with good faith and the reliance interests of the user and copyright holder.
A. Efficiency and Incentives
1. A Model of Author-User Interaction
In abstract terms, a typical fair use case proceeds as follows: The author creates an original work. (For simplicity, the model will presume that no transfer of copyright has occurred, i.e., the original author retains the copyright. (82)) At some later time, the purported fair user wishes to make some use of the original, e.g., to quote from it or incorporate aspects into a different work.
In thinking through the model, it may be helpful to envision the situation as a close fair use case--that is, evaluation of the four traditional fair use factors (and any additional non-consent considerations) leaves the balancing test roughly in equipoise. In other words, there are reasonable arguments that use is fair under the standard factors, and roughly equally reasonable arguments that it is not.
As a concrete example, consider the facts of Harper & Row, a fair use case that divided the Supreme Court six to three against fair use. (83) Harper & Row addressed whether The Nation magazine's "scoop" of President Ford's unpublished memoir was a fair use. (84) Having received a prepublication copy from an unidentified source, the magazine quoted and paraphrased from the forthcoming memoir to produce an article detailing Ford's pardon of former President Richard Nixon. (85) Evaluation of the traditional fair use factors yields an inconclusive result in Harper & Row. on the one hand, The Nation's purpose is newsreporting (privileged under the first factor), the underlying work is factual, and the amount of actual expression taken was very small--by the dissent's count, only 300 words from a 200, 000 word manuscript. (86) on the other hand, the memoir was unpublished, the scoop directly harmed the market for the memoir, and those 300 words were arguably the "heart of the work." (87) In a close case like Harper & Row, the consent consideration may make the difference. (88)
In contrast, the model's conclusions will tend not to be critical to cases when the use is obviously fair (e.g., a brief quote for educational purposes) or obviously unfair (e.g., commercial pirating of identical copies). Because evaluation of the traditional factors is one-sided, consent is unlikely to be decisive in those contexts.
i. Stage One: Prior to the Use
First, consider the situation prior to the user's decision to proceed with his contemplated use. At this point, the user may or may not know something of the author's preferences regarding uses of her work. If there is some preexisting relationship between the user and author, the author may have already communicated her preferences directly, or implicitly through her conduct. Conversely, there may be no relationship between the author and user, in which case the author can communicate her preferences only generally through a legend on the work, or--more commonly--not at all. (89)
The first decision falls to the user: Should he ask permission of the author before making his use? The benefits of asking permission are obvious. If permission is granted, the user may proceed with the use without fear of later litigation, though of course he will have to pay any agreed-upon licensing fee. The decision not to ask permission may be made for several reasons. The user may believe, based on indications from the author or otherwise, that the author approves of the use. The user may have a good faith belief that his use is fair, and that therefore no permission from the author is needed. He may not be willing to pay, or be unable to afford, the expected licensing fee. As asking permission alerts the copyright owner to the intended use, the user may decline to ask permission in order to avoid detection of his potential infringement.
Importantly, communication costs will enter into the permission-seeking calculus. In cases where the user and the author have a preexisting relationship, these costs will tend to be low--the time to make a phone call or send an email--and thus unlikely to play a significant role in the decision. (90) In a more typical case, communication costs can be significant, as it can be very difficult to ascertain who the copyright holder is and to contact that person. (91) In the extreme case, the user may be entirely unable to determine the current rightsholder, and permission seeking is a practical impossibility. (92)
When the author is aware of the intended use, she, too, has a choice to make at the pre-use stage: Should she let the user know that she approves or disapproves of the anticipated use? The decision to announce preferences or not may also be made for several reasons. The author may think that her wishes are already clear. She may not care about the anticipated use, and thus have little incentive to say anything. On the strategic side of the ledger, the author may wait to object because her bargaining position will be improved after the use is made. Communication costs are, of course, important here as well. If it is costly to contact the user, the author may not get enough benefits--either in deterring unwanted uses or in generating licensing revenue from agreed-upon uses--to make it worth her while.
If the user decides to ask permission or the author decides to announce her preferences, then some sort of negotiation will ensue. The author may demand a licensing fee for the use, which the user can agree to, refuse, or attempt to bargain down. Or the author's refusal may be categorical: a refusal to permit the use regardless of the compensation offered. If the negotiation results in an effective agreement, of course, then the use is authorized and no fair use case will ensue.
ii. Stage Two: After the Use
Next, consider the situation after the use is made, i.e., once the user has appropriated expression from the original work. If the author and user communicated beforehand, the dynamics of the situation have not changed appreciably. The author is aware of the use, and its potentially infringing nature. As the original negotiation broke down, the remaining decisions are whether either party will choose to reopen negotiations, and whether the author will bring or threaten suit.
In the situation where the author and the user have not yet communicated, the dynamics are more interesting. There are three possible scenarios. The first and simplest case is when the author never becomes aware of the use. This may arise intentionally--e.g., when the user's "avoid detection" strategy has succeeded--or unintentionally. In this situation, there will never be any author-user communication, nor any fair use suit. The second possibility is when the author was initially unaware of the use, but later gains knowledge of the use. There, the author never had an opportunity to announce her preferences earlier, and the author's decision whether to announce her preferences will take place at this stage. The author, however, will now have the additional leverage of threatening suit.
In the third case, the author has been aware of the use all along, but declined to announce preferences at stage one. It is this final category that is ripe for strategic behavior by the author. In particular, the author can "lie in wait" to take advantage of an improved post-use bargaining position. (93) The author has increased negotiating leverage after the use because the user can no longer change his course of action. Before the use is made, the user can credibly threaten to make an alternative use (e.g., relying on a different work), or to make no use at all (e.g., by excluding the portion that relied on the author's original work). After the use is made, the user's only leverage is to force the author to sue and to take his chances on a fair use defense. Simply by making the new work, the user has already infringed: contrary to popular misconception, there is no requirement that the user sell or publish the infringing work in order to be liable. (94) Indeed, the user is liable for heavy statutory damages--up to $150, 000 per work infringed--even if his use caused little or no actual economic harm to the author. (95) Unable to change course, and faced with the threat of super-compensatory damages, the user has dramatically decreased leverage in a post-use negotiation.
2. Deterring Strategic Behavior and Ex Post Disputes
Given this idealized setup, we can now ask the critical question: What sorts of consent-seeking interactions do we, as a social matter, wish to encourage, and which actions do we wish to discourage?
Two assumptions necessary for the conclusions that follow should be made be made clear at the outset. First, in order for the benefits of the consent factor to be realized, communication costs must be sufficiently low such that the costs of communication do not exceed the expected social benefits. Second, the conclusions will presume that the non-consent elements of current fair use doctrine work roughly as intended--to wit, to permit uses only when doing so is socially beneficial. (96) Although the assumption is of course artificial, it will enable us to isolate the effects of the standard view of consent and the role that consent ought to play in the fair use analysis.
Turning to the conclusions, identifying the first class of cases that our fair use law should want to discourage is relatively easy. At the least, it should seek to prevent opportunism by both the author and user. As used here, "opportunism" means strategic, bad faith behavior by an author or user. Specifically, opportunism includes intentional misrepresentations, in actions or words, regarding whether one believes a use to be fair or unfair, or whether one objects to a use. In these instances, the argument to consider the character of authorial refusal in fair use reduces to the economic justifications of equitable doctrines such as laches and unclean hands: "Equity in private law is ... aimed at preventing opportunism." (97)
An author who could easily object to a known use, but chooses not to, may be rightly penalized for a lack of diligence in asserting a claim that prejudiced the user--much like the equitable doctrine of laches. (98) After a use is made, the user has little leverage as he is "locked in" to the use, and the author can threaten to sue for statutory damages far in excess of any actual damages suffered. (99) Such a threat would likely deter users even when the social value of their planned use exceeds the true costs to the author. In other words, the author's opportunism creates a social cost. (100) An opportunistic author seeks to exploit the dynamics of the consent-seeking interaction to extract additional value from users; ex ante, this possibility may deter some socially-valuable uses.
Similarly, a user who avoids asking permission to avoid detection, knowing his use to be unfair, invokes the fair use defense with unclean hands. (101) (Of course, there may be good faith reasons to avoid detection--for example, when a user honestly believes a use to be fair but simply seeks to avoid costly litigation. This is not "opportunism" as used here, which only includes users who believe their uses to be unfair, yet act otherwise.) If we assume that current fair use doctrine works as intended to permit uses only when they are socially valuable, the possibility of escaping detection skews the user's incentives such that he may engage in socially-harmful uses, i.e., uses that cost the author more than the net social benefit. (102) One way to discourage these opportunistic behaviors is to penalize bad faith actors in the fair use analysis. (103)
What about failures in communication not attributable to bad faith? Preventing opportunism aside, is there any reason to encourage users to ask permission and authors to declare preferences? One potential reason is that early preference-revelation could promote early resolution of disputes and deter later fair use litigation. (104) Fair use litigation is uncertain and costly. (105) Moreover, there is reason to believe that pre-use negotiation is more likely to lead to a resolution--or, at least, not litigation--if the user opts to change course. Because of the author's disproportionate leverage after the use is made, pre-use negotiation will tend to be fairer and more efficient. Prior to use, the author's ability to threaten a suit for super-compensatory damages is more limited, as the user still possesses the effective counter-threat of not making the use at all. (106) In other words, the pre-use negotiation is, ceteris paribus, more likely to reflect the actual social value of the use. (107) Thus, in the absence of other externalities, (108) pre-use negotiation is more likely to lead to agreements to make uses only when they are socially beneficial.
Again, the above arguments hold only if communication costs are low. Whatever benefits there may be to early preference revelation, there are also considerable costs. For users, there are significant costs in ascertaining and contacting the rightsholder; for the author, in time lost announcing preferences, negotiating with users, or monitoring users for possible infringement. Thus, outside of the bad faith context, failure to ask permission or announce preferences should be penalized only when communication costs are low due to a preexisting relationship between the author and user, the availability of a low cost opt out system, or analogous contexts.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Abstract through III. The Case for Consent as a Fair Use Factor A. Efficiency and Incentives, p. 397-423|
|Author:||Hickey, Kevin J.|
|Publication:||Yale Journal of Law & Technology|
|Date:||Mar 22, 2014|
|Previous Article:||The new model of interest group representation in patent law.|
|Next Article:||Consent, user reliance, and fair use.|