Considering the Role of Fairness in Copyright Fair Use

I. Introduction

The word “fair” is right there at the front of the term “fair use.” But what concepts of fairness are, and are not, inherent in copyright’s fair use doctrine? The statute begins by enumerating categories of use that are likely fair depending on the application of a multifactor test—uses such as criticism, commentary, news reporting, and education—but fair to whom?^[1] The underlying copyright holder? The user? Society at large? Certainly, those uses may be beneficial to promote discourse and knowledge, but what does that have to do with fairness?

A number of scholars have observed that copyright law has the potential to promote social justice, or conversely, to establish or reinforce stubborn social, racial, or economic hierarchies.^[2] The fair use doctrine is surely central to when and how copyright promotes or inhibits social justice. Fair use is an engine for social justice when it allows less privileged speakers to “talk back” to dominant culture, find voices and audiences, and forge paths to success. On the other hand, some social justice scholars are skeptical of fair use, which can allow more privileged speakers to engage in cultural misappropriation and exploitation with impunity. The question of when fair use promotes fairness, then, may have more to do with who is on which side of the fair use coin than with any of the doctrinally prescribed fair use factors.

Fairness, as a concept, is complicated. Common contemporary definitions of fairness tend to focus on honesty, impartiality, justness, and equality of opportunity—words whose meanings are themselves complicated and disputable.^[3] The House of Lords put it well: “Fairness is an elusive concept. It is an instinctive response to a given set of facts. Ultimately it is grounded in social and moral values. These values, or attitudes, can be stated. But they cannot be justified, or refuted, by any objective process of logical reasoning.”^[4] Of all of the many disputed and disputable notions of fairness, the present thought experiment is focused on fairness as distributive justice and the idea that systems should promote equitable distribution of resources and expressive power.

One common definition of fairness focuses on free riding: that people are owed a fair share of their costs of producing goods and services and, therefore, that it is morally unacceptable to free ride on the benefits of others’ labor without paying the costs of that labor.^[5] But that presumption is tempered by other common definitions, which make clear that, in the words of John Rawls, “[s]ocial and economic inequalities are . . . [first, to be] open to all under conditions of fair equality of opportunity; and second, they are to be to the greatest benefit of the least-advantaged members of society.”^[6] Fair equality of opportunity requires not only that social positions “be open in the formal sense, but that all should have a fair chance to attain them.”^[7] Those with the same level of talent and ability should have the same prospects of success regardless of their origins. Crucially, this requires a framework of institutions that “adjust the long-run trend of economic forces so as to prevent excessive concentrations of property and wealth, especially those likely to lead to political domination”^[8] or, additionally, domination in expressive discourse.

The argument is that this tempering—Rawls’s “difference principle”—aligns fair use with fairness.^[9] The fair use doctrine exists precisely to define circumstances where free riding is desirable.^[10] Therefore, a crucial aspect of fair use may be the notion of just enrichment. As a matter of fairness, free riding is acceptable when it yields other social or distributive benefits, such as access to knowledge, enhanced discourse, social mobility, or related drivers of “progress,” and when it does so without undue competitive harm.^[11] This notion of just enrichment as fairness may also explain why the first factor of the fair use analysis considers commerciality: as a matter of intuitive fairness, expressive copying is more easily acceptable when it does not produce a direct financial benefit for the free rider. For the same reason, commerciality is not the only factor in fair use because we are willing to accept financial benefits for the free rider if there are other social or discursive benefits to the free riding.

One appropriate benefit of free riding in the fair use context may be ameliorating distributive unfairness, including inequity in discursive opportunity and cultural market share. That premise animates the current thought experiment: would fair use be more “fair” if it expressly considered relative party status?^[12] The fair use doctrine, as it currently stands, likely helps ameliorate inequalities that copyright law otherwise promotes or perpetuates. Although empirical evidence shows that less privileged litigants do not tend to win fair use litigation, the mere existence of the doctrine permits critical, technological, and educational uses that promote equity.^[13] Fair use helps upturn artificial hierarchies between speaker and hearer that can reinforce economic, social, and historical inequalities.^[14] However, fair use may also provide avenues for the exploitation of underrepresented speakers, especially because sophisticated and repeat players can use fair use as a justification for uncompensated uses, and the expense of copyright litigation makes it difficult for underrepresented creators to seek compensation for those uses. The purpose of this Article is to explore whether and how the fair use doctrine could act more directly as a tool for promoting social justice. Part II explores the relationship between fair use and social justice under current law and theory. Part III discusses the intuitive and moral relationship between status and fairness. Part IV discusses the mechanisms, advantages, and disadvantages of more explicitly considering relative party status within the fair use analysis.

II. Fair Use and Social Justice

While copyright law may not be overtly concerned with questions of diversity, access, equity, and inclusion, the copyright regime has profound impacts on these topics.^[15] Some have argued that copyright exclusivity has significant potential as a tool for promoting dignitary and economic justice by allowing less privileged creators to professionalize, set prices for their creations, and reap income when others profit from those creations.^[16] Others have observed that copyright exclusivity has generated and perpetuated historical inequity and exploitation.^[17] Copyright ownership and enforcement do not always align with most people’s concepts of justice and fairness, because they often benefit corporate interests over individuals, and benefit the already privileged over those with fewer financial or cultural resources.^[18] Despite their seeming inconsistencies, each of these observations is correct. Copyright ownership may provide opportunities for some less privileged creators to flourish while permitting and perpetuating the exploitation of others. As with many other legal systems that govern relationships between people, social justice is neither inherent in copyright law nor inimical to it.

The relationship between social justice and copyright’s fair use doctrine is surely even more complicated. Fair use may be the central aspect of copyright social justice, and it has appropriately received significant attention in that context.^[19] As Professor Sonya Katyal explained:

[F]air use bears an intimate relationship to the way in which critical legal studies focused its gaze on the role of entitlements for minority groups. Like critical race theory, a critical approach to copyright law tends to ask the question of how entitlements are distributed and their effect on disenfranchised groups, and also to employ tools like fair use to restore some balance between property rights and social justice.^[20]

Copyright, by its nature, allows someone to own expression—it places copyright holders on a higher rung of expressive hierarchy and thus gives copyright owners the power to restrict how others express themselves. To the extent that copyright ownership favors the already powerful, that hierarchy reinforces existing power dynamics. But because fair use tempers the exclusivity of copyright ownership, it provides an avenue for counteracting copyright’s hegemonic tendencies.^[21]

Social justice may not be the express purpose of the fair use doctrine. In fact, it is hard to say with certainty what the purpose(s) of the doctrine are.^[22] Fair use—at least arguably—accomplishes many things, including remedying market failures,^[23] resolving potential conflicts between copyright and free speech,^[24] serving the public interest,^[25] promoting productive uses of copyrighted works,^[26] and enabling audiences to engage with culture as producers in addition to consumers.^[27] That said, the fact that the fair use statute specifically enumerates criticism, commentary, and education as among the activities likely to qualify as fair use surely implies that social justice fits among its broader objectives.^[28] Indeed, Justice Blackmun explicitly recognized the social justice function of fair use, describing the doctrine as “a form of subsidy . . . [that permits] limited use of [a] work for the public good.”^[29] Scholars have identified similar subsidy-like properties in fair use.^[30]

In a broader sense, where exclusive ownership creates a hierarchy between “creator” and “user,” fair use offers a framework for resisting or upturning that hierarchy.^[31] Fair uses frequently dismember works and bring them closer so that they can be examined, experimented with, critiqued, or laughed at.^[32] Thus, fair use is an inherently political act that “deactivates the apparatuses of power.”^[33] Fair use, as a concept, “invert[s] the power hierarchy inherent in the ownership aspect of copyright law by recognizing that all expression (whether the law defines it as ‘original’ or ‘derived’) depends on what came before.”^[34] Even more powerfully, fair use gives the “deriving” creator a right to utter the last word (or more accurately, the next word) over the possible objections of the “original” creator. As a practical matter, fair use can allow less privileged speakers to “talk back” to dominant culture, find their own voices and audiences by personalizing and building upon existing works, and forge paths to professional success.^[35] Fair use can also facilitate discussions of social justice by shining a light on bias in popular culture and by facilitating the creation of affinity groups where participants from different backgrounds encounter each other and share their experiences.^[36] Fair use can help less privileged educational institutions gain and share resources.^[37]

At the same time, however, some social justice scholars are suspicious of fair use,^[38] and they have cause to be. While fair use allows users to upturn copyright hierarchies, it also reinforces those same hierarchies. Notwithstanding the inherently dialogic nature of creation, copyright designates some creations as “original” to give authors rights over their entire works, and denigrates authors whose works are deemed “derivative” by making them dependent upon a legal exception for their rights.^[39] Reliance on fair use in litigation is famously uncertain and expensive,^[40] and because the fair use analysis favors noncommercial uses,^[41] it disadvantages those who seek to make a living from upturning copyright hierarchies.^[42]

Moreover, fair use may provide avenues for privileged (often large, corporate) players to engage in cultural misappropriation or exploit the creative labor of less privileged authors with relative impunity.^[43] While the fair use doctrine takes commerciality into account, it does not (and should not) exempt commercial uses from the fair use umbrella.^[44] Therefore, a privileged player’s profitable transformation of a less privileged creator’s work may appropriately be deemed fair use—thus, in a sense, depriving the less privileged creator of copyright licensing income that the privileged player could afford to pay. As Professor Lateef Mtima puts it, while “remix” may create otherwise unavailable opportunities for less privileged creators:

[I]t is likely that major copyright conglomerates will reap the ultimate commercial benefits from “relaxed” copyright protection in the digital context through mass production of remix works. In so far as African American and other marginalized authors and artists are concerned, this would merely begin the cycle of unjust exploitation anew.^[45]

Furthermore, evidence tends to show that privileged players are more likely to prevail in fair use litigation than less privileged ones. Empirical research conducted by Matthew Sag shows that courts are more likely to find fair use when “the plaintiff is a natural person and less likely [to find fair use when] the plaintiff is a corporation.”^[46] Indeed, plaintiff status may be the single best predictor of litigation outcomes. This finding is consistent with both intuitive and empirical understandings that privileged players are more likely to prevail in litigation, and, in Sag’s words, it “undermines the characterization of fair use as a redistributive tool favoring the politically and economically disadvantaged.”^[47]

One possible reason for this pattern, consistent with Sag’s findings, is that more privileged players are more likely to have sophisticated attorneys to analyze fair use in deciding whether to take or grant a license or make arguments for or against fair use when copying occurs.^[48] Furthermore, information costs associated with fair use’s uncertainty differentially benefit the well-resourced.^[49] Less measurably, but perhaps more dramatically, factors like fame, fortune, and popularity may skew fair use analyses in favor of the more privileged. Courts are more likely to recognize transformative merit in the work of participants who have acquired fame or admiration than they are in the work of upstarts.^[50] This means, in the words of Andrew Gilden, that the fair use doctrine has often “functioned in a stratified way, acting largely for the benefit of those who already enjoy a privileged cultural voice.”^[51] Furthermore, risk and uncertainty aversion may exacerbate unevenness in fair use outcomes. The somewhat-predictable-yet-uncertain nature of fair use could embolden people—but the people it emboldens are likely the people who could afford to defend themselves. Others will be more risk-averse.

III. The Role of Status in Fairness

The distributive impacts of any given fair use analysis are likely to influence our moral intuitions about whether the result is “fair.” Similar behavior will inevitably seem morally “fair” or “unfair” depending on who carries it out. For example, there is a strikingly different intuitive moral valance to a situation in which a teenager creates and sells a work of fanfiction based on a studio motion picture without paying the movie studio, than a situation in which a movie studio releases a blockbuster movie based on a teenager’s short story without paying the teenager. Perhaps our intuitions about this unfairness rely on assumptions about the magnitude of apparently unjust enrichment that the re-creator gains from free riding^[52]—the teenager probably will not make much money from selling their fanfiction, while the movie studio will probably make a lot of money from selling their blockbuster—but that is not the whole picture. It seems unfair because the typical movie studio is a privileged player with a wealth of resources and an abundance of cultural experience and influence, while the typical teenager has fewer resources and cultural influence, and is at a stage of their personal journey where writing fanfiction likely helps them develop their identity, skills, and voice.^[53] In this scenario, if the teenager were to pursue copyright remedies from the movie studio, it would be “punching up”—seeking just compensation from one with more power, wealth, and influence. If the movie studio were to pursue copyright remedies from the teenager, it would be “punching down”—engaging in a sort of abuse of power, resources, and influences that borders on bullying.

A similar set of moral intuitions govern cultural appropriation. It is problematic in a variety of ways when dominant-culture players profit from imitating the traditional knowledge and expression of historically oppressed or underrepresented peoples, regardless of whether those expressions are protectable by copyright law. On the other hand, no one complains—nor should they—when oppressed peoples profit from imitating dominant culture. It is morally troubling (at least) for a designer brand to pattern its textiles with a Native American thunderbird. It is fine for anyone to pattern their textiles with Tudor roses.^[54]

These moral intuitions have little to do with copyright incentives. In each situation, the less privileged creators did not make their work with massive profit in mind; were it not for the appropriation, they would never see riches from their work. Indeed, the more privileged appropriators are doing nothing to harm the less privileged creators. There is no unfair competition because there is no competition. There is no market usurpation because there is no meaningful market for the less privileged creators’ work without the added contributions of the privileged appropriators. Or, to put it differently, the less privileged creators would have made it anyway; they weren’t working in the hope that some designer or studio would come along and adapt their work. Any benefit they get from the designer or studio would be a windfall. And yet, there is a moral sense that by taking without paying, the privileged appropriators are being unjustly enriched. Even if it doesn’t fit well into copyright’s incentive theory, it doesn’t feel fair. But reverse the parties, and it’s fine. The enrichment becomes just.

Although these moral intuitions aren’t explicit in copyright doctrine, there is a fairness objective embedded in copyright policy that may justify incorporating them into copyright analyses. One of the primary justifications for copyright is that it advances progress by creating a mechanism for people who might not otherwise be able to afford to create works of authorship to professionalize by allowing them to set prices for works with market value.^[55] This objective implies that copyright is, at least in part, targeted at making authorship a viable profession, rather than limiting the author function to wealthy dilettantes and creators who can find wealthy patrons. Copyright is, therefore, in some senses, targeted at helping the less privileged more than the more privileged, except to the extent that the more privileged invest in the creative activities of the less privileged. Fair use not only serves as a First Amendment check on the speech-chilling potential of copyright exclusivity but also factors heavily into the constitutional goal of progress through professionalization. As Professor Jessica Silbey puts it, fair use is among the doctrines that safeguard the public domain, thereby enabling access to “essential cultural resources [for] current and future generations of artists and innovators, facilitating common welfare and community-wide justice for more people.”^[56] Less experienced creators build their skills, voices, and audiences by imitating, transforming, critiquing, and sharing their versions of copyrighted works.^[57] Consider that every year, scores of fan-artists bring their portfolios to San Diego Comic-Con, hoping their own (fair) uses of copyrighted material will garner them positions in the comics industry.^[58]

To be clear, this is not to suggest that less privileged creators are necessarily likely to make “better” or more socially beneficial uses of copyrighted works. I am merely suggesting that, from a fairness and social justice perspective, less privileged creators may need a boost to ensure that their voices are heard without undue legal challenges and that their voices are not exploited without recourse—and that such a boost aligns with the policy rationales underlying the fair use doctrine. This distinction may not go so far as to make it “fair” for the less privileged to engage in outright piracy from wealthy recording companies or movie studios, but it does make it “fair” to encourage less privileged creators to engage in expressive dialogue with those advantaged entities without having to pay or get permission from those entities. Allowing “have-nots” to talk back to “haves” is unlikely to reduce incentives for the haves to create or invest in creation, but it provides added incentives for the have-nots. Put in economic terms, there is a limit to the marginal incentive value of copyright, especially for the already prosperous.^[59] However, the marginal value of dialogic expression to the less prosperous is significant. In fairness terms, the haves do not need a leg up; the have-nots do.

Of course, not everyone may agree with the premise that providing a boost to less privileged players is “fair.” Recent debates regarding affirmative action have highlighted long-standing ideological differences between some who believe that fairness requires refraining from classification, and others who believe that engaging in classification is appropriate for purposes of seeking to overcome subordination.^[60] In a world where advantage and disadvantage work on multiple axes and individuals may be advantaged in some senses and disadvantaged in others, the very idea of using classification to achieve “group fairness” is contestable, particularly in the context of expressive speech.^[61] But in the copyright context, classification for the purpose of antisubordination seems a fitting response to the constitutional rationale that allowing expressive professionalization encourages “progress.” If the copyright system statistically benefits privileged players over less privileged ones—as it appears to do—then that is a perverse outcome. Engaging in classification to counteract that outcome would thus be fair.

The fair use doctrine has no formal mechanism for taking these considerations into account. Certain aspects of fair use may hint at considering the status of the parties, but none do it outright. For example, in the first factor, courts may (but seldom do) incorporate considerations of good faith, such as whether an alleged infringer failed to pay a customary price or otherwise violated relevant industry norms.^[62] To some extent, fair use considers fairness when it favors criticism, commentary, and education, as each of those acts has the potential to upturn discursive and social hierarchies. But fair use does not care about the larger relationship between the criticizer and the criticized; a wealthy, dominant-culture criticizer gets as much of a boost from fair use as a less privileged one does. Likewise, fair use arguably favors the less privileged by favoring noncommercial uses over commercial ones. But this means only that the movie studio and the teenager who sells their fanwork are both on the wrong side of fair use’s favoritism; in fact, a less privileged creator who wants to earn an income by engaging in fair use will run afoul of this consideration. That seems, for lack of a better term, unfair.

However, in other areas of law, doctrinal adjustments expressly account for the exploitation or social justice harm that otherwise agnostic rules could foster. For example, ambiguous contracts are interpreted against the drafter to prevent those with greater bargaining power from incorporating exploitive ambiguities into contracts they draft.^[63] Questions regarding whether a contract is unconscionable take into account the relative education, sophistication, bargaining power, and wealth of the parties.^[64] Courts take the relative wealth of the parties into account in determining whether to assess attorney’s fees^[65] and punitive damages.^[66] Is it possible to do such a thing with copyright fair use? Would it be wise to try?

IV. Considering the Role of Status in Fair Use

Would the fair use doctrine better promote progress if it expressly took into account considerations like party resources and cultural market share? What would happen if the fair use analysis expressly queried the relative power and economic positions of the participants vis a vis each other and broader society, and favored uses that favor less privileged players over uses that benefit more privileged ones—just as it favors transformative users over nontransformative ones, noncommercial users over commercial ones, and noncompeting users over those who usurp markets?

A. Defining Status as a Fair Use Consideration

Formally speaking, such an adjustment would not require any statutory change. The statute provides that the four fair use factors are nonexclusive.^[67] As a practical matter, courts have already converted the four-factor test into debatably eight factors: (1)(a) transformativeness; (1)(b) commerciality and/or educational purpose; (1)(c) good faith; (2)(a) whether the underlying work has thick or thin copyright; (2)(b) whether the underlying work is public or private; (3)(a) whether the use took a quantitatively large portion of the underlying work; (3)(b) whether the use took the “heart” of the underlying work; and (4) whether the use usurps a market for something the underlying copyright holder would likely do or authorize.^[68] Some courts have, at least implicitly, added other considerations to the list, such as whether quotations are adequately attributed as such.^[69] Adding party status to this list—not as a determining factor, but as one of many considerations—is therefore consistent with the letter of the law, and possibly its spirit.^[70]

One way of incorporating relative party status into the fair use analysis without requiring radical change is acknowledging it as a possible consideration in jury instructions. Juries surely always (whether consciously or not) incorporate their moral intuitions into their decisions, but while juries may sometimes favor the “little guy” in litigation, they may just as often be impressed by a famous name or successful corporation. The same implicit biases that make judges and juries less likely to appreciate subtle transformations or transformations common to other cultures, may make them presume that a more privileged party is righteous or that a less privileged party is shady.^[71] An explicit statement in jury instructions that jurors may incorporate status considerations into their deliberations, including whether a party could afford to pay for a license or whether a given use works to the benefit or detriment of a less privileged creator or community, might point jurors in the direction of incorporating fairness into their fair use deliberations.^[72]

This Article does not suggest that implementing this idea would be easy—but it is, nonetheless, worth considering how to make it feasible. The first challenge is determining how to identify and measure privilege and relative status. It does not go unnoticed that by aggregating different kinds of disadvantages—economic disadvantage, sociocultural discrimination, historical oppression, disability, etc.—this Article glosses over important differences in the historical and practical reasons why some copyright parties may be differently situated from each other. In life, those differences matter a great deal. However, for purposes of the current thought experiment, it’s unclear whether those differences need to be treated differently. The postulation is a “fairness” role for fair use, which could, at least theoretically, operate across multiple axes of economic, sociocultural, and historical unfairnesses. For related reasons, this approach does not require developing a hierarchy of privilege(s) and disadvantage(s). This is not to suggest that relative status should be the main fair use consideration—or that small gradations in relative status should affect the outcome. Rather, this Article suggests that comparing status, in this context, is a large-grained enterprise—we are looking for obvious power differentials, not subtle distinctions.

Thus, for example, in assessing privilege and relative status, one might compare the parties’ financial resources to determine whether there is a significant difference between them. One might look at whether one side is an individual and the other is a corporate entity on the (generally reliable) assumption that corporations tend to have more power and resources than individuals.^[73] One might look at whether any of the parties are members of traditionally underrepresented or historically disadvantaged groups, and whether they are opposed by dominant-culture or established opponents. Any and all of these seem like reasonable questions to ask. They surely will not provide definitive answers every time. What does one do, for example, with a wealthy member of a historically disadvantaged group? Or an impoverished start-up competing with a wealthy individual? In such situations, the answer may be that there simply is no obvious power differential to consider. But if the goal of the exercise is to put a thumb on the scale for less privileged players facing more privileged players in litigation, these considerations at least frame a good start.

Because disadvantages work on so many axes, it may be difficult to place limits on who should benefit from this approach. But it is not necessarily impossible. Here, too, it helps to consider the objectives of the exercise and work backward. Here, I presume, as above, that one of the primary objectives of copyright law is to advance progress by creating a mechanism for people who might not otherwise be able to afford to create works of authorship to professionalize by allowing them to set prices for works with market value.^[74] Keeping that objective in mind, this framework would examine whether a party faces barriers to success that are unrelated to their inherent ability to create works of authorship—be those barriers of finances or systemic discrimination. This approach would help people and companies who are just starting out, those who are learning their arts by making derivative works (as most creators do),^[75] and those (such as incubators, technologists, and even some investors) who make it possible for less privileged creators to grow and professionalize. It is not an easy line to draw, but it is a rational one.

The exercise would have to consider the parties’ relative status at the time of the alleged infringement, rather than the parties’ status at the time of the suit.^[76] To do otherwise would be to invite opportunistic litigation against those who build successful careers or businesses based on fair use activities. Imagine the following example: a budding African-American recording artist with few resources samples a pop music passage in a hip-hop song that sells very well. At the time of the sampling, she is a relative have-not, particularly in relation to the music publisher and record label that own copyrights to the pop song. If the fair use analysis considered her status at the time of the sampling, it would more likely lead to a finding of fair use, allowing her to build a career, fame, and fortune—thus satisfying the social justice goal. Later in her career, when she is already rich and famous, she would be safe from litigation over her earlier sampling, but if she wanted to use sampling in her new works, the same analytical approach would more likely require her to seek a license. At the same time, this approach would make it easier for new budding artists to sample her work to build their own careers. The same principle would work for tech startups that start small and get huge. Nearly every company starts small. If they engage in fair use to grow, it would be nonsensical to pull their foundation out from under them, even after they get big. At the same time, however, once they are huge, they no longer need a fair use boost, and they can start paying for new uses. It may seem odd to imagine that the same activity, if performed at different times in someone’s career, could be fair use if early but infringement if late. On the other hand, the approach would align better with distributive fairness intuitions than a system that ignored status altogether.

Also, we would often need to look past the actual parties litigating to consider the ultimate beneficiaries of a use or ultimate bearers of infringement harm. Large corporations or wealthy institutions may embark on a potentially infringing endeavor that, while it may (further) enrich the corporation or institution, ultimately serves a public interest function for less privileged users.^[77] By a similar token, a corporation or institution may sue on behalf of smaller, less privileged copyright beneficiaries.^[78] In those situations, while a privileged party’s name may be prominent in the litigation itself, the fairness interest lies at least as much with the nonparties whose interests they represent. Consider, for example, the mass-digitization efforts of HathiTrust (or Google, for that matter): while the litigants were privileged parties that benefited from their copying, their work—making information massively searchable for research, and accessible to the disabled—powerfully served the social justice interests of the less privileged.^[79] Or consider Georgia State University, making course packs for the ultimate benefit of students (a famously penurious community).^[80]

These questions are not clear-cut. If we look past the named defendants in those cases, we must also look past the Author’s Guild and Cambridge University Press (each large, wealthy institutions), where we will find a mass of individual (possibly wealthy, possibly struggling) authors.^[81] My personal intuitions—which may or may not be borne out by fact—are that in each of these situations, the bulk of any infringement damages in these cases would have benefited publishers rather than individual authors.^[82] Thus, the status inquiry may help the defendants in these particular situations more than the plaintiffs. But the question would be fact-intensive, at best.

However, sometimes even the fact-intensive inquiry of looking past a named party to find a real beneficiary may be straightforward, especially in situations where we might look past a publisher to find a single (more or less privileged) author rather than the massive collections of authors discussed above. After all, we want publishers to have the ability to invest in underrepresented authors without worrying about infringement. That said, it may more commonly be prohibitively expensive to conduct discovery into the deeper (often speculative) questions of who ultimately benefits and who ultimately suffers from any particular copying, and any answers may be indeterminate. Perhaps these situations demonstrate a significant weakness in the premise that status can reasonably be considered as part of a fair use analysis. Or perhaps these considerations just highlight the fact that relative status will be irrelevant when each side is both privileged and not privileged at once.

In fact, just as various fair use factors are more or less important in different situations, there are many cases in which relative party status would be irrelevant. For example, status would not matter when both sides are privileged (e.g., movie studios suing each other) or not privileged (a rarer situation considering the high costs of litigation). Party status will also sometimes be unimportant. For example, status would not help an individual who is engaged in a large-scale piracy operation because the other factors—nontransformativeness, commerciality, the copying of entire, thick-copyright works, and market usurpation—would easily swamp any consideration of status.

B. Positioning Status in the Fair Use Framework

Including status in the fair use analysis fits well within a number of theoretical frameworks for fair use. As a threshold matter and central to the present analysis, it aligns with a critical analysis of copyright law. Although fair use is facially neutral, it can reinforce systemic bias against nondominant speakers. Judges and juries are more likely to notice the transformed meaning or message of a use when that transformation aligns with more common, majoritarian approaches to parody, criticism, or critique, but that does not make subtler or rarer forms of transformativeness any less socially beneficial as fair uses.^[83]

For example, fans who commonly discuss the cultural harm that may result from ingrained media tropes may easily perceive the critical nature of fanvids that highlight and critique those tropes. Other audiences may not find those messages without further description and explanation. But that does not mean that subtle critique is not valuable social commentary. Indeed, the most effective way “to educate viewers about potential social harms in mainstream media . . . [may be to repurpose] examples they . . . previously viewed uncritically.”^[84] By a similar token, the centuries-old African-American art tradition of “Signifyin(g)'” on older expressions (as Henry Louis Gates, Jr. describes it) involves referring to preexisting material, imitating or repeating it with difference, recontextualizing it, and drawing attention to new connotations for it.^[85] Signifyin(g) can be particularly powerful for resisting sociocultural hierarchies because it allows creators to claim or critique the speech genres of the more powerful. Sometimes signifyin(g) is subtle to an untrained or outside observer, but no less powerful for that subtlety. For example, Dr. Dre’s “Let Me Ride” (1992) and Ice Cube’s “The Product” (1990) incorporate samples of Parliament Funkadelic and Sly and the Family Stone in ways that critique the optimistic outlook that 1960s–1970s’ funk took to race relations.^[86] In these examples, the meaning of the quoted text is entirely transformed by context, but because this form of signifyin(g) often relies on juxtaposition rather than overt commentary, it may not seem particularly transformative to outsiders.^[87]

In these scenarios, rich or famous artists—or artists who engage in the sort of critique that is obvious to dominant culture—are likely to fare better as fair use defendants simply because their methods of transformation are more broadly recognized as having artistic merit than those of budding or obscure artists or those who engage in subtler or nondominant methods of critique. At the same time, rich or famous copyright owners (record companies, publishers, movie studios) are likely to fare better as fair use plaintiffs because the market for their works is well-known, and a judge or jury might reasonably presume that they are well-positioned to exploit markets for derivative works.^[88] These trends can compound to doubly disadvantage already disadvantaged creators on both sides of the fair use coin.

Incorporating status as a thumb on the fair use scale would encourage judges and juries to look past the obvious—to find and understand more subtle (but no less important) transformations, and to observe that even two relatively-similar works may not competitively compare and may have radically different market appeals. Take, for example, two products that use images from the Harry Potter films: an “official” one made by Warner Brothers in colors that align with Hogwarts’ houses and an “unofficial” one in colors that replicate a pride flag made by a small queer-owned company. Considering relative party status would encourage a court to recognize the transformativeness in using pride colors on a product referencing the works of a notoriously transphobic author, and to recognize that—especially in light of J.K. Rowling’s vocal opposition to trans rights^[89]—the market for pride-themed goods has no meaningful overlap with the market for official Harry Potter merchandise. However, it should be made clear that this does not suggest that considering party status would make any particular use more (or less) transformative. It would merely create a context that could help illuminate transformations that might not otherwise be obvious to the untrained eye.

Another way of thinking about this approach is to say that context matters in fair use. While the fair use framework expressly considers the underlying (copyrighted) work and the resulting work and its use,^[90] it does not expressly incorporate any concept of the creative context surrounding resulting works and their uses.^[91] Rather, it incorrectly assumes that copyright is a one-size-fits-all system. Considering context—such as distributive justice and equality of resources and access—requires acknowledging the fiction of that assumption.^[92] This is consistent with a view of copyright law and fair use that centers not only on copyright owners but also on audiences of consumers, creators, and re-creators. That “consumer protection model,” as Joseph Liu describes it, mirrors other areas of law by “recogniz[ing] a distinction between sophisticated and unsophisticated parties” and mitigates the differential effects of fair use’s information costs.^[93] Relatedly, it is consistent with a growing body of literature proving that, in the aggregate, the inclusion of diverse voices is economically beneficial for innovation and growth for both creators and consumers.^[94]

This framework is also consistent with the conception of fair use as a mechanism for remedying market failure. Less resourced parties are less likely to be able to afford the rents charged by large-scale copyright holders, while better resourced parties are able to afford licenses or attorneys to analyze when and whether licenses are needed.^[95] Copyright holders may be more likely to be skeptical of license requests from members of oppressed or disadvantaged communities, perhaps because of bias or a defensive presumption that licensing uses may reveal the injustice of social hierarchies.^[96] Yet the expressions of those with fewer resources are, as a policy matter, at least as socially valuable as those with financial and cultural resources. Considering status in fair use provides a subsidy of sorts to less privileged speakers, thus remedying these market failures.

One may argue that the lack of universal affordability is not, in itself, a market failure. There are many things that the poorly resourced cannot afford, and subsidies do not step in at every turn. The law does not intervene so that everyone can afford a yacht. But the law does (at least try to) step in to ensure that everyone can afford basic needs, such as food, housing, and transportation.^[97] It steps in to (again, try to) ensure that those basic needs are not differentially expensive or available based on traits like race and familial status.^[98] It is rational that the law would step in to do something similar for the constitutionally guaranteed right to expression.

One might even consider aspects of this framework as providing a sort of reparations for African-American creators. As Professor K.J. Greene has pointed out, copyright industries have systematically exploited African-American creators and should recognize and redress this harm.^[99] While it may be impossible to compensate for those exploitations in any direct way, one indirect step would be to respect the transformatively creative expressions that African-American artists create using existing copyrighted works. Copyright history is littered with examples of disrespect.^[100] These cases have encumbered rap and hip-hop, not only influencing the way African-American artists make music but also making them more dependent on (and often thus exploited by) the licensing departments of music companies, and alienating them from copyright in their works, which become a patchwork of material they may not be able to sublicense.^[101] A system that expressly considers the history of oppression that African-American authors have faced, and affords opportunities to profit independently of corporate licensing schemes, would be a small step toward remedying that; it could even provide a form of stimulus for African-American artists.

This framework might also help address other cultural misappropriation problems. For example, scholars have illuminated the harms that cultural misappropriation can inflict on the creative, social, economic, and political interests of Native Americans and other Indigenous peoples.^[102] Taking into account the power dynamic of colonization and the way that colonization continues to harm Native American peoples would not only help balance fair use outcomes but would also allow courts to afford attention to Indigenous norms and governance within the fair use framework. It would discourage cultural misappropriation by dominant-culture players while allowing Indigenous artists to rely on fair use to reclaim misappropriated works.^[103]

More broadly, this framework would mean that less privileged creators would be able to claim copyright in their works more often. Under § 103 of the Copyright Act, creators of compilations and derivative works own copyrights only in their original contributions to those works and their rights do not extend to “any part of the work in which [preexisting] material has been used unlawfully.”^[104] The creator of an infringing compilation or derivative work may own their original contributions to the extent that they may be separated from infringing material—but that ownership may be slim to none.^[105] However, when relative status is incorporated into the fair use framework, a greater proportion of derivative works by less privileged creators become copyrightable (at least as to the aspects original to the derivative author), removing the doubt and risk of the divisibility analysis from their ownership. This provides a compounding benefit: not only are less privileged creators able to get a foot in the expressive door with less risk of liability, but once they are there, they own copyrights that may be asserted more reliably against more privileged copiers.^[106]

C. Concerns About This Approach

Despite the theoretical benefits discussed above, incorporating relative status into the fair use analysis could be both impractical and expensive, especially if the consideration becomes too granular rather than focusing on obvious status differentials. Questions of who ultimately benefits and suffers from copying are philosophically complicated to say the least. Haves and have-nots frequently work together to make creation possible, and the line between haves and have-nots is blurry and subjective. This problem is compounded by the fact that discovery into the question of relative party status is, at best, likely to be expensive and indeterminate.^[107] It risks creating a sideshow to copyright litigation that overwhelms the (rightfully) more central fair use questions. It could turn every copyright litigation into a deep dive into party finances and a “race to the bottom” of oppression and worthiness and could justify a degree of intrusion that, in other circumstances, could qualify as inappropriate discovery tactics.^[108] This expense and intrusion could do exactly the opposite of what the approach envisions—it could make litigation even more arduous and out-of-reach for less privileged litigants.

On the other hand, the “fairness” approach I advance—even when framed in the litigation-centric terms of jury instructions—is not only about litigation outcomes. It may matter even more in driving conduct outside of litigation. For example, when and whether a creator decides to publicize or sell a derivative work made without permission, when and whether a copyright holder decides to invest in threatening or initiating a lawsuit, and when and on what terms parties decide to settle disputes. Knowing ex ante that a court or jury might favor less privileged parties may significantly influence each of those decisions, which dwarf litigation in both frequency and overall importance to the expressive environment. Advance knowledge that a court may incorporate relative status in the fair use analysis may act as a disincentive (on both sides) to engage in costly litigation.

Another concern is that formally incorporating relative status into the fair use analysis risks complicating fair use precedent. While fair use analysis is already a deeply fact-intensive undertaking, precedent is nonetheless important.^[109] Would cases where more privileged parties prevailed have precedential effects for or against have-nots? Would cases where less privileged parties prevailed make otherwise bad law?^[110] Would incorporating yet another factor into the analysis make fair use outcomes even less predictable than they already are? Would precedent simply become less reliable as relative status provides one more axis on which cases could be distinguished as precedent? Or would less privileged parties be able to act with greater confidence knowing that they have a reliable consideration on their side? These things are hard to predict—and their practical effects on copying behavior may vary based on a prospective copier’s tolerance for risk. To achieve a thorough version of fairness considering relative party status as an aspect of fair use would require thinking not only about the immediate consequences of a particular litigation outcome but about the more distant precedential consequences of that litigation outcome. Put differently, it would require asking not only whether a particular disadvantaged party would be helped or harmed by a finding of fair use, but also whether a case would create a more broadly applicable rule of precedent that would help or harm future disadvantaged parties in similar situations. A perfect version of this would require an impossible level of clairvoyance, but that should not justify discarding relative party status as a relevant question.

There is also precedential value to judicial forthrightness. The Warhol opinion contains the subtle implication that the majority may have been influenced by questions of relative status; it opens with a discussion of Lynn Goldsmith’s artistic bona fides as a rock and roll photographer, perhaps to counteract assumptions that Warhol’s relative fame confers a presumption of artistic superiority.^[111] It later notes that “[i]t will not impoverish our world to require AWF to pay Goldsmith a fraction of the proceeds from its reuse of her copyrighted work.”^[112] But, if the Court’s intention was to put its thumb on the scale for the less privileged player in the saga, it did not say so. Instead, the Court created a test that imported concepts of market competition (factor 4) into the analysis of the purpose and character of the use (factor 1)^[113]—a test that is likely to hurt less privileged creators of transformative works more than it helps them. Had the Court forthrightly incorporated relative status into its analysis, it may have averted that risk.

Furthermore, one may reasonably argue that a framework like the one discussed here would create a sort of counterincentive to success. Perhaps that is “fair” in some sense because it evens the playing field, but in another sense, it defies the fairness principle that everyone should reap the full reward of their effort.^[114] A system that analyzes liability differently based on the participants’ level of success may seem like pulling the copyright rug out from under those who have relied successfully on copyright (and fair use) to garner success.

But while these concerns are real, they do not contradict the idea that relative status may provide a useful background principle in the fair use context. Incorporating ideas of relative status into the fair use analysis can refine and improve the law’s concepts of when commerciality is and is not important, and when and how transformation of copyrighted material may benefit or harm disadvantaged artists and communities. It can encourage judicial humility and discourage jurors from being impressed by fame or fortune. It may even make fair use fairer.

V. Conclusion

In the end, formulating fair use to take relative party status into account could create harm by raising the cost of copyright litigation and distracting from more central copyright issues. But despite these practical concerns, there is value in considering party status as a way of promoting social justice by aligning the fair use analysis with prevailing moral intuitions about fairness. Copyright law has been engaged in covert social engineering since its inception—creating and reinforcing expressive hierarchies in ways that tend to marginalize nondominant speakers—and there is value in considering how to counteract those effects. Some have considered other approaches to distributive unfairness in copyright law, such as fee-shifting and compulsory licensing, and each has pros and cons. But fair use is, by its nature, a doctrine about preventing (or at least alleviating) the potential harms of copyright exclusivity, so situating a remedy for copyright’s socioeconomic hierarchy in the fair use analysis is a sensible approach to consider.

And as a thought experiment, explicitly situating social justice within fair use permits us to consider how copyright could help amplify underrepresented voices and act as a tool for social equity. While practical considerations may make the present thought experiment unwise to put into operation, there are many reasons to recommend a system that expressly considers relative party status as an aspect of fair use. It would be ridiculous to argue that the less privileged participant in a copyright dispute should always prevail, but it is reasonable to put a thumb on the fair use scale for underrepresented creators. Even as it now stands, the fair use doctrine remains a tool for social justice, albeit an incomplete one. The struggle, as with so many questions in copyright, is how to accomplish more fairness, more fairly.

---

1. 17 U.S.C. § 107.

2. See infra Part II.

3. See, e.g., Fairness, Oxford English Dictionary (R.W. Burchfield ed., 1933).

4. Miller v. Miller; McFarlane v. McFarlane [2006] UKHL 24, [2006] 2 AC (HL) 618 (appeal taken from Eng.).

5. Idil Boran, Benefits, Intentions, and the Principle of Fairness, 36 Canadian J. Phil. 95, 95 (2006).

6. John Rawls, Justice as Fairness: A Restatement 42–43 (Erin Kelly ed., 2001).

7. Id. at 43.

8. See id. at 43–44.

9. Id. at 42–43.

10. See Pierre Leval, Toward a Fair Use Standard, 103 Harv. L. Rev. 1105, 1116, 1127 (1990).

11. See id. at 1114, 1116, 1126, 1136.

12. See Oren Bracha & Talha Syed, Beyond Efficiency: Consequence-Sensitive Theories of Copyright, 29 Berkeley Tech. L.J. 229, 289, 291–92, 294–95 (2014) (suggesting taking wealth into account in the fair use analysis); Molly Shaffer Van Houweling, Distributive Values in Copyright, 83 Tex. L. Rev. 1535, 1568–69 (2004) (proposing to consider defendant’s “ability to pay” in the fair use analysis for follow-on derivative uses).

13. See infra Part II.

14. See infra Part II.

15. The scholarship on this topic is wonderful, vast, and growing. For a small sampling, see generally, e.g., Keith Aoki, (Intellectual) Property and Sovereignty: Notes Toward a Cultural Geography of Authorship, 48 Stan. L. Rev. 1293 (1996); Rosemary J. Coombe, The Cultural Life of Intellectual Properties: Authorship, Appropriation, and the Law (1998); Angela R. Riley, Recovering Collectivity: Group Rights to Intellectual Property in Indigenous Communities, 18 Cardozo Arts & Ent. L.J. 175 (2000); Van Houweling, supra note 12; Olufunmilayo B. Arewa, Copyright on Catfish Row: Musical Borrowing, Porgy and Bess, and Unfair Use, 37 Rutgers L.J. 277 (2006); Olufunmilayo B. Arewa, From J.C. Bach to Hip Hop: Musical Borrowing, Copyright and Cultural Context, 84 N.C. L. Rev. 547 (2006); Margaret Chon, Intellectual Property “from Below”: Copyright and Capability for Education, 40 U.C. Davis L. Rev. 803 (2007); K.J. Greene, ''Copynorms," Black Cultural Production, and the Debate over African-American Reparations, 25 Cardozo Arts & Ent. L.J. 1179 (2008); Organization for Transformative Works, Comments of the Organization for Transformative Works to the USPTO/NTIA, Nat’l Telecomm. & Info. Admin. (Nov. 13, 2013), https://www.ntia.doc.gov/files/ntia/organization_for_transformative_works_comments.pdf [https://perma.cc/3HUK-NK75]; Bracha & Syed, supra note 12; Anjali Vats & Deidré A. Keller, Critical Race IP, 36 Cardozo Arts & Ent. L.J. 735 (2018); Madhavi Sunder, From Goods to a Good Life: Intellectual Property and Global Justice (2012); Haochen Sun, Copyright and Responsibility, 4 Harv. J. Sports & Ent. L. 263 (2013); Elizabeth L. Rosenblatt, Copyright’s One-Way Racial Appropriation Ratchet, 53 U.C. Davis L. Rev. 591 (2019); Lateef Mtima, IP Social Justice Theory: Access, Inclusion, and Empowerment, 55 Gonz. L. Rev. 401 (2019–2020); Robert Brauneis, Copyright, Music, and Race: The Case of Mirror Cover Recordings, in The Cambridge Handbook of Intellectual Property and Social Justice (Steven D. Jamar & Lateef Mtima eds., forthcoming 2023); Anjali Vats, The Color of Creatorship: Intellectual Property, Race, and the Making of Americans (2020); Jessica Silbey, Against Progress: Intellectual Property and Fundamental Values in the Internet Age (2022); Elizabeth L. Rosenblatt, User-Generated Transformation: Intellectual Property, Social Justice, and Fanworks, in The Cambridge Handbook of Intellectual Property and Social Justice, supra note 15, at 339 [hereinafter Rosenblatt, User-Generated Transformation].

16. See, e.g., Lateef Mtima, Prologue: A Social Activist’s Guide to Intellectual Property, in Intellectual Property, Entrepreneurship and Social Justice: From Swords to Ploughshares at xvii (Lateef Mtima ed., 2015) (“To gain control over your intellectual property—the products of your mind, talent, and cultural traditions—is to gain control over resources that can give you the leverage to do business in the national and global marketplace on a level playing field.”); Justin Hughes & Robert P. Merges, Copyright and Distributive Justice, 92 Notre Dame L. Rev. 513, 554–55 (2016); Tuneen E. Chisolm, Whose Song is That? Searching for Equity and Inspiration for Music Vocalists Under the Copyright Act, 19 Yale J.L. & Tech. 274, 331–33 (2017).

17. See Vats, supra note 15, at 16–18, 20–22; Greene, supra note 15, at 1217 (“If the music industry is serious in its rhetoric about ‘theft’ of IP, it should atone for the theft it itself has facilitated. The entertainment industry is arguably the prime beneficiary of special interest intellectual property legislation that seeks compensation for even trivial uses of intellectual property.”). See generally Rosenblatt, supra note 15; Elizabeth L. Rosenblatt, Social Justice and Copyright’s Excess, 6 Tex. A&M J. Prop. L. 5 (2020).

18. See, e.g., Silbey, supra note 15, at 276–77; Nicolas Suzor, Free-Riding, Cooperation, and “Peaceful Revolutions” in Copyright, 28 Harv. J.L. & Tech. 137, 140, 147–60, 192 (2014) (using a case study of pay-what-you-want systems to highlight the disconnect between copyright doctrine and consumers’ concepts of morality and fairness); Jessica Silbey, The Eureka Myth: Creators, Innovators, and Everyday Intellectual Property 27–28 (2015) (highlighting the disconnect between formal law and creator conceptions of creation and innovation); Siva Vaidhyanathan, Copyrights and Copywrongs: The Rise of Intellectual Property and How It Threatens Creativity 21–22 (2001); Glynn Lunney, Copyright’s Excess: Money and Music in the U.S. Recording Industry 195–96 (2018).

19. See, e.g., Ruth Okediji, Givers, Takers and Other Kinds of Users: A Fair Use Doctrine for Cyberspace, 53 Fla. L. Rev. 107, 153 (2001) (“To the extent that private ordering through contract or technological mechanisms disproportionately impacts those of a lower socioeconomic class or those with inferior bargaining power . . . fair use has been the indispensable and indisputable guardian of public welfare.”); Organization for Transformative Works, supra note 15, at 29–38; Betsy Rosenblatt & Rebecca Tushnet, Transformative Works: Young Women’s Voices on Fandom and Fair Use, in eGirls, eCitizens 385, 393–96 (Jane Bailey & Valerie Steeves eds., 2015); Deborah Gerhardt & Madelyn Wessel, Fair Use and Fairness on Campus, 11 N.C. J.L. & Tech. 461, 481–82 (2010) (arguing that fair use may incrementally improve resource equality among educational institutions); Sun, supra note 15, at 299–303; Steven D. Jamar, A Social Justice Perspective on the Role of Copyright in Realizing International Human Rights, 25 Pac. McGeorge Glob. Bus. & Dev. L.J. 289, 307–308 (2012); Lateef Mtima, Copyright and Social Justice in the Digital Information Society: “Three Steps” Toward Intellectual Property Social Justice, 53 Hous. L. Rev. 459, 470–71 (2015); Blake E. Reid, Copyright and Disability, 109 Calif. L. Rev. 2173, 2180–81 (2021); William W. Fisher III, Reconstructing the Fair Use Doctrine, 101 Harv. L. Rev. 1659, 1761–62 (1988).

20. Peter Goodrich et al., Panel 1: Critical Legal Studies in Intellectual Property and Information Law Scholarship, 31 Cardozo Arts & Ent. L.J. 601, 613 (2013) (citing Van Houweling, supra note 12, at 1535).

21. Some prominent cases demonstrate how fair use allows speech that resists hegemonic narratives. See, e.g., Suntrust Bank v. Houghton Mifflin Co., 268 F.3d 1257, 1270–71 (11th Cir. 2001) (retelling of Gone with the Wind from the point of view of an enslaved woman); Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 582–83 (1994) (mocking repurposing of popular crooner into hip-hop).

22. See Pamela Samuelson, Unbundling Fair Uses, 77 Fordham L. Rev. 2537, 2615–16 (2009) (discussing various justifications for fair use).

23. 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) (articulating market failure approach to fair use).

24. See Eldred v. Ashcroft, 537 U.S. 186, 194, 219–20 (2003) (reasoning that extension of copyright term did not abridge free speech because the fair use doctrine afforded latitude in the use of copyrighted works).

25. See Haochen Sun, Copyright Law as an Engine of Public Interest Protection, 16 Nw. J. Tech. & Intell. Prop. 123, 136–39 (2019).

26. See, e.g., Sony Corp. v. Universal City Studios, Inc., 464 U.S. 417, 449–51 (1984); Authors Guild, Inc. v. HathiTrust, 755 F.3d 87, 95 (2d Cir. 2014); Google LLC v. Oracle Am., Inc., 141 S. Ct. 1183, 1198 (2021).

27. See Joseph P. Liu, Panel III: Fair Use, Notice Failure, and the Limits of Copyright as Property, 96 B.U. L. Rev. 833, 835–36 (2016); Jessica Silbey, Panel III: Fairer Uses, 96 B.U. L. Rev. 857, 864 (2016) (discussing Liu’s approach).

28. 17 U.S.C. § 107; see also Silbey, supra note 15, at 227.

29. Sony Corp., 464 U.S. at 478 (Blackmun, J., dissenting).

30. Jane C. Ginsburg, Authors and Users in Copyright, 45 J. Copyright Soc’y U.S.A. 1, 15 (1997) (emphasizing the “redistribution” of value from copyright owners to preferred classes of users); Jane C. Ginsburg, Copyright, Common Law, and Sui Generis Protection of Databases in the United States and Abroad, 66 U. Cin. L. Rev. 151, 169 (1997) (viewing fair use as a subsidy from copyright owner in favor of uses with public benefits); 2 Paul Goldstein, Goldstein on Copyright § 12.2.2.1(a), at 12:28 (3d ed. Supp. 2013) (observing that the “commercial-noncommercial distinction” in fair use can be characterized as “a covert subsidy to worthy nonprofit enterprises such as schools and universities”); Robert P. Merges, The End of Friction? Property Rights and Contract in the “Newtonian” World of On-Line Commerce, 12 Berkeley Tech. L.J. 115, 135 (1997) (characterizing fair use as a tax on copyright owners and a subsidy in favor of certain classes of users).

31. See Elizabeth L. Rosenblatt, Fair Use as Resistance, 9 U.C. Irvine L. Rev. 377, 388–92 (2019).

32. See M. M. Bakhtin, Epic and Novel, in The Dialogic Imagination: Four Essays 3, 23 (Michael Holquist ed., Caryl Emerson & Michael Holquist, trans., 1981).

33. Cf. Giorgio Agamben, Profanations 76–77 (Jeff Fort, trans., Zone Books 2007) (2005) (“Once profaned, that which was unavailable and separate loses its aura and is returned to use.”).

34. See Rosenblatt, supra note 31.

35. See Organization for Transformative Works, supra note 15, at 29–38; see generally Rosenblatt & Tushnet, supra note 19; Anupam Chander & Madhavi Sunder, Everyone’s a Superhero: A Cultural Theory of “Mary Sue” Fan Fiction as Fair Use, 95 Calif. L. Rev. 597, 603 (2007); Rosenblatt, User-Generated Transformation, supra note 15; Evans C. Anyanwu, Note, Let’s Keep It on the Download: Why the Educational Use Factor of the Fair Use Exception Should Shield Rap Music from Infringement Claims, 30 Rutgers Comput. & Tech. L.J. 179, 196–98 (2004) (arguing that rap music sampling serves an educational function in the dissemination of African-American culture and thus should be allowed under the fair use doctrine as an educational use).

36. See, e.g., Rebecca Tushnet, Scary Monsters: Hybrids, Mashups, and Other Illegitimate Children, 86 Notre Dame L. Rev. 2133, 2138–43 (2011) (examining transformative vids); Brief for Electronic Frontier Foundation & Organization for Transformative Works as Amici Curiae Supporting Petitioner at 15–17, Andy Warhol Found. for the Visual Arts v. Goldsmith, 143 S. Ct. 1258 (2023) (No. 21-869) (discussing critical fanvids); @fernacular, Welcome to: If Male Superhero Costumes were Designed Like Female Superhero Costumes!, Tumblr (Feb. 18, 2012), https://fernacular.tumblr.com/post/17814450235/welcome-to-if-male-superhero-costumes-were [https://perma.cc/9R9A-LLMH]; Rosenblatt, User-Generated Transformation, supra note 15.

37. See Gerhardt & Wessel, supra note 19, at 529.

38. Lateef Mtima, Copyright Social Utility and Social Justice Interdependence: A Paradigm for Intellectual Property Empowerment and Digital Entrepreneurship, 112 W. Va. L. Rev. 97, 145–50 (2009) (“There are cogent arguments to the [e]ffect that in light of past injustices, it is only fair that members of marginalized groups be allowed wide latitude in connection with the unauthorized use of copyrighted works. There is much merit to this position. As a pervasive intellectual property empowerment strategy, however, the efficacy of such an approach is limited.” Professor Mtima favors a compulsory licensing scheme over fair use, noting that “[a]s many an African American rap artist has begun to appreciate, she who samples today shall herself be sampled tomorrow.” (citations omitted)).

39. See 17 U.S.C. §§ 102–03.

40. Some scholars have identified predictable patterns in fair use litigation, but predictability and certainty are very different, and the prospect of litigation defense is expensive, which may discourage many potential fair users. See generally Barton Beebe, An Empirical Study of U.S. Copyright Fair Use Opinions, 1978–2005, 156 U. Pa. L. Rev. 549 (2008) (identifying predictable patterns in fair use jurisprudence); Matthew Sag, Predicting Fair Use, 73 Ohio St. L.J. 47 (2012) (assessing the predictability of fair use outcomes in litigation); Samuelson, supra note 22 (identifying policy-based clusters in fair use jurisprudence).

41. Goldsmith, 143 S. Ct. at1279–80.

42. See Rosenblatt, supra note 31, at 392; cf. Rebecca Tushnet, My Fair Ladies: Sex, Gender, and Fair Use in Copyright, 15 Am. U. J. Gender Soc. Pol’y & L. 273, 300–04 (2007) (arguing that the analysis of commercial use under factor one tends to disfavor a predominantly female class of authors and consumers).

43. See, e.g., Trevor G. Reed, Fair Use as Cultural Appropriation, 109 Calif. L. Rev. 1373, 1394–1400 (2021) (discussing and providing examples of conduct perpetuating cultural misappropriation). A great deal of cultural appropriation and exploitation is unrelated to fair use and rather depends on the fact that many cultural products—such as collectively created, oral-tradition, or unfixed works—are unprotected by copyright and thus free for the taking, so to speak. The contours of the fair use doctrine have no effect on these exploitations, except to the extent that fair use permits those whose cultural production has been incorporated into copyrighted works to criticize those copyrighted works without facing paradoxical copyright liability. See id. at 1399–1401.

44. 17 U.S.C. § 107; see also Authors Guild v. Google, Inc., 804 F.3d 202, 219 (2d Cir. 2015) (“Many of the most universally accepted forms of fair use, such as news reporting and commentary, quotation in historical or analytic books, reviews of books, and performances, as well as parody, are all normally done commercially for profit.”).

45. Mtima, supra note 38, at 139–40 (citation omitted).

46. Sag, supra note 40, at 78.

47. Id.; see also Tushnet, supra note 42, at 300 (arguing that the analysis of market harm under factor four and commercial use under factor one tend to disfavor a predominantly female class of authors and consumers); Robert Brauneis & Dotan Oliar, An Empirical Study of the Race, Ethnicity, Gender, and Age of Copyright Registrants, 86 Geo. Wash. L. Rev. 46, 94 (2018) (arguing that “[j]udges exercising their discretion in fair use cases (as well as other contexts) would be right to be self-conscious about the gendered disparate impact of their decisions”).

48. See Liu, supra note 27, at 842.

49. Id.

50. Andrew Gilden, Raw Materials and the Creative Process, 104 Geo. L.J. 355, 375–88 (2016) (discussing the phenomenon in detail).

51. Id. at 378.

52. See Silbey, supra note 15, at 241–43 (providing examples in which creators’ views on copying vary based on a user’s profitability and ability to pay).

53. See id. at 252–55.

54. Neither of these situations is likely to implicate copyright. As noted above, much cultural misappropriation falls outside of copyright’s ambit as it concerns appropriation of oral-tradition, collectively-created, very old, or unfixed knowledge. See Rosenblatt, supra note 15, at 617–24. Issues surrounding copyright protectability and social justice fall outside the scope of this Article.

55. See Glynn S. Lunney, Jr., Reexamining Copyright’s Incentives-Access Paradigm, 49 Vand. L. Re v. 483, 581 (1996).

56. Silbey, supra note 15, at 222.

57. See Organization for Transformative Works, supra note 15, at 29–38.

58. See Beth Accomando, Comic-Con Is Not Over for Artists Who Did Portfolio Reviews, KPBS (July 26, 2022, 6:50 PM), https://www.kpbs.org/news/arts-culture/2022/07/26/comic-con-is-not-over-for-artists-who-did-portfolio-reviews [https://perma.cc/LM86-5PJ3].

59. See Silbey, supra note 15, at 275 (“[R]epeat players succeed not because they have learned to play the game but because their relative wealth and influence from past successes accumulate to assure their dominance and future successes . . . signs of institutional failure according to norms of equal justice in a democratic society.”).

60. See generally Students for Fair Admissions v. President and Fellows of Harvard Coll., 143 S. Ct. 2141 (2023) (applying the anticlassification approach in rejecting constitutionality of the antisubordination approach); Reva B. Siegel, Equality Talk: Antisubordination and Anticlassification Values in Constitutional Struggles over Brown, 117 Harv. L. Rev. 1470, 1474–75 (2004) (discussing the historical rise of the distinction between anticlassification and antisubordination values).

61. See Citizens United v. FED, 558 U.S. 310, 340, 372 (2010) (holding that the First Amendment stands against restrictions distinguishing among different speakers).

62. See, e.g., Harper & Row Publishers, Inc. v. Nation Enters., 471 U.S. 539, 562–63 (1985); Google LLC v. Oracle Am., Inc., 141 S. Ct. 1183, 1204 (2021) (considering common custom and practice in relevant industry); Wall Data, Inc. v. L.A. Sheriff’s Dept., 447 F.3d 769, 778 (9th Cir. 2006). It is not meant to imply that considering good faith would necessarily advance social justice. On the contrary, the Campbell case (in which 2 Live Crew sought and was denied a license but could assert the fair use defense despite its arguable bad faith in continuing nonetheless) demonstrates how notions of “good faith” could be used against speakers who wish to engage in rebellious expression. See Campbell v. Acuff-Rose Music, Inc., 510 U.S. 569, 572–74, 594 (1994).

63. See Restatement (Second) of Contracts § 206 & cmt. a (Am. L. Inst. 1981).

64. See, e.g., City of Raton v. Ark. River Power Auth., 760 F. Supp. 2d 1132, 1154 (D.N.M. 2009).

65. See, e.g., Faraci v. Hickey-Freeman, Co., 607 F.2d 1025, 1028 (2d Cir. 1979) (“[F]ee awards are at bottom an equitable matter, [and] courts should not hesitate to take the relative wealth of the parties into account.” (citation omitted)).

66. See Restatement (Second) of Torts § 908(2) (Am. L. Inst. 1979). Relative party status is not typically a consideration in assessing tort liability more generally. See 30 Am. Jur. Trials 711 § 1 (Aug. 2023). However, I would argue that the fair use analysis is distinguishable from more general questions of tort liability. General tort law does not depend on the constitutional notion of “progress” for its justification, or on the foundational premise of promoting expression the way that copyright and fair use do. Moreover, while most tortfeasors cannot reasonably be expected to know—much less take—a victim’s relative status into account ex ante, a putative fair user is more likely to know whose work they are using and where the respective parties stand. Nevertheless, while it is outside the scope of this Article, the arguments herein likely apply to questions regarding copyright damages and fees as well—taking relative status into account in determining the amount and appropriateness of damages is at least as sensible as taking relative status into account in assessing fair use. Cf. Ga.–Pacific Corp. v. U.S. Plywood Corp., 318 F. Supp. 1116, 1117, 1120 (S.D.N.Y. 1970) (permitting consideration of relative size and wealth of parties in patent damages analysis); Adams Lab’ys, Inc. v. Jacobs Eng’g Co., 761 F.2d 1218, 1226 (7th Cir. 1985) (holding that in the liability phase of patent litigation, allowing plaintiff to use its disadvantaged status as a small company to appeal “to the sympathy of jurors through references to the relative wealth of the defendants in contrast to the relative poverty of the plaintiff[] . . . may be cause for reversal”); Innovention Toys, LLC v. MGA Ent., Inc., No. 07–6510, 2012 WL 5384880, at *4 (E.D. La. Nov. 1, 2012) (distinguishing between evidence of relative wealth for purposes of appealing to juror sympathy and evidence of relative wealth for purposes of assessing patent damages).

67. 17 U.S.C. § 107.

68. See Sag, supra note 40, at 54–55 (breaking down fair use factors into subconsiderations); see also supra note 56 and accompanying text (good faith falls under “the purpose and character of the use” umbrella).

69. See, e.g., Warner Bros. Ent. Inc. v. RDR Books, 575 F. Supp. 2d 513, 544 (S.D.N.Y. 2008) (considering Harry Potter lexicon’s extensive quotation without the use of quotation marks).

70. This is a different question from whether a particular instance of copying serves the public interest, although it is related. Public interests, such as access to knowledge, also help disadvantaged populations. However, here, I am considering an intervention that expressly favors disadvantaged litigants—even individual artists whose work may have no “public” impact. For discussion of the public interest as a fair use consideration, see, e.g., Sun, supra note 25, at 145–46; Rebecca Tushnet, Copy This Essay: How Fair Use Doctrine Harms Free Speech and How Copying Serves It, 114 Yale L.J. 535, 562 (2004) (arguing that nontransformative uses may provide social benefit through access to knowledge).

71. See infra text accompanying note 83.

72. See Zahr K. Said, Jury-Related Errors in Copyright, 98 Ind. L.J. 749, 807–10 (2023) (calling for an approach that provides juries with a clear understanding of what they can do and consider).

73. See Sag, supra note 40, at 66–67, 78.

74. See supra Part III.

75. See Organization for Transformative Works, supra note 15, at 50–52; Christine Nishiyama, Inspiration vs. Imitation: How to Copy as an Artist, Skill Share Blog (Dec. 8, 2021), https://www.skillshare.com/en/blog/inspiration-vs-imitation-how-to-copy-as-an-artist/ [https://perma.cc/N6AT-WHKK] (describing copying as learning and discussing differences between imitation, derivation, and plagiarism). George Bernard Shaw is often credited with the sentiment: “Imitation is not just the sincerest form of flattery—it’s the sincerest form of learning.” Id.

76. This is consistent with the Supreme Court’s analysis in Warhol v. Goldsmith, which assesses the commerciality of discrete uses. Thus, the mere creation of a work may be fair use, and display of the work in a museum may be fair use, but later licensing of the work may not be fair use. See Andy Warhol Found. for the Visual Arts, Inc. v. Goldsmith, 143 S. Ct. 1258, 1291 (2023) (Gorsuch, J., concurring) (further explaining the majority’s decision). Likewise, a relative status analysis would best take a snapshot at the time of the discrete use in question rather than asking about status in the abstract.

77. See, e.g., Authors Guild, Inc. v. HathiTrust, 755 F.3d 87, 91 (2d Cir. 2014); Authors Guild v. Google, Inc., 804 F.3d 202, 228–29 (2d Cir. 2015); Cambridge Univ. Press v. Patton, 769 F.3d 1232, 1263–64 (11th Cir. 2014).

78. See, e.g., United States v. Am. Soc’y of Composers, Authors & Publishers, 599 F. Supp. 2d 415, 419 (S.D.N.Y. 2009) (detailing ASCAP’s argument that “ringback tones” do not constitute fair use).

79. See Authors Guild, Inc., 755 F.3d at 91; Authors Guild, 804 F.3d at 228–29. More research would be needed to determine whether this framework would create a principled distinction between the mass digitization conducted for HathiTrust and Google Book search and the mass digitization conducted by artificial intelligence (AI) training companies and services, such as Common Crawl, Stable Diffusion, and ChatGPT. At first glance, it appears that digitization for research and accessibility more directly works to the benefit of the less privileged, whereas digitization for AI training is either neutral or potentially harmful (in the job-replacement sense) for these groups. On the other hand, training AIs on the most complete data sets possible not only increases the value of AIs to all users (regardless of their degree of privilege) but may also have particular benefits for underrepresented communities. Commentators have observed that limiting AI training to public-domain data sets can lead to serious algorithmic bias. See, e.g., Amanda Levendowski, How Copyright Law Can Fix Artificial Intelligence’s Implicit Bias Problem, 93 Wash. L. Rev. 579, 610–19 (2018). There is, of course, no guarantee that just because a work is copyrighted it is any less biased than an uncopyrighted work; there is no shortage of bias in the universe of copyrighted works. Nevertheless, including the works of underrepresented creators in AI training data is, on the whole, likely to reduce the bias of the AI’s output.

80. Cf. Cambridge Univ. Press, 769 F.3d at 1238–41, 1260–84 (reversing, vacating, and remanding for a more comprehensive analysis as to whether coursepack uses constituted fair use) (The court noted that “because of the circumstances of this case, some of the factors weigh more heavily on the fair use determination than others.”).

81. See Silbey, supra note 15, at 216–20, 256–57.

82. See id. at 276–77 (providing some evidence of this intuition).

83. See supra notes 48–51 and accompanying text.

84. See Brief for Electronic Frontier Foundation and Organization for Transformative Works, supra note 36, at 15–16.

85. See Henry Louis Gates, Jr., The Signifying Monkey: A Theory of African-American Literary Criticism 51–52, 88 (1988).

86. See Joanna Demers, Steal This Music 82–85 (2006) (discussing “Let Me Ride” and “The Product”).

87. See id. at 85.

88. See supra notes 48–51 and accompanying text.

89. See Aja Romano, Is J.K. Rowling Transphobic? Let’s Let Her Speak for Herself, Vox, https://www.vox.com/culture/23622610/jk-rowling-transphobic-statements-timeline-history-controversy [https://perma.cc/6HWG-8ECP] (last updated Mar. 16, 2023, 10:17 AM).

90. See Andy Warhol Found. for the Visual Arts, Inc. v. Goldsmith, 143 S. Ct. 1258, 1277 (2023) (“The same copying may be fair when used for one purpose but not another.”).

91. 17 U.S.C. § 107.

92. See Silbey, supra note 27, at 864–66.

93. See Liu, supra note 27, at 842–49, 851–54 (describing the impact of resource differential on fair use information costs and discussing consumer protection model for fair use).

94. See Joel Waldfogel, The Welfare Effect of Gender-Inclusive Intellectual Property Creation: Evidence from Books (Nat’l Bureau of Econ. Rsch., Working Paper No. 30987, 2023), http://www.nber.org/papers/w30987 [https://perma.cc/6XEV-J9DX]; Chang-Tai Hsieh et al., The Allocation of Talent and U.S. Economic Growth, 87 Econometrica 1439, 1472–73 (2019) (finding that “falling barriers [to the occupational choice facing women and Black men in highly-skilled occupations] explain[s] roughly 40% of aggregate growth in market GDP per person”).

95. See Liu, supra note 27, at 842–44.

96. Cf. Rosenblatt, supra note 17, at 17 (noting that some copyright holders can refuse to license their works when it would be risky).

97. See generally 42 U.S.C. (The Public Health and Welfare).

98. E.g., 42 U.S.C. §§ 2000a, 3604 (Public Accommodations).

99. Greene, supra note 15, at 1189–1200, 1216–19.

100. See Rosenblatt, supra note 15, at 608–10.

101. See Rosenblatt, supra note 17, at 16–21.

102. See, e.g., Riley, supra note 15, at 197–201; Reed, supra note 43, at 1385–90.

103. See Reed, supra note 43, at 1385–99.

104. 17 U.S.C. § 103.

105. Id.; see Eden Toys, Inc. v. Florelee Undergarment Co., 697 F.2d 27, 34–35 (2d Cir. 1982) (holding that maker of derivative work has copyright in noninfringing increments of expression); Pickett v. Prince, 207 F.3d 402, 405–07 (7th Cir. 2000) (holding that the maker of a derivative work had no copyright when infringement “pervade[d]” derivative work).

106. Cf. Rosenblatt, supra note 17, at 8–9.

107. Cf. Brian T. Fitzpatrick, Twombly and Iqbal Reconsidered, 87 Notre Dame L. Rev. 1621, 1638–43 (2012) (discussing rising discovery costs); Penelope Eileen Bryan, Women’s Freedom to Contract at Divorce: A Mask for Contextual Coercion, 47 Buff. L. Rev. 1153, 1175–80 (1999) (discussing prohibitive costs and indeterminacy of discovery into personal finances in matrimonial context).

108. Fed. R. Civ. P. 26(b)(2)(C) (requiring proportionality of discovery in certain circumstances); see also, e.g., Hottenstein v. City of Sea Isle City, No. 11–740, 2014 WL 12616926, at *1 (D.N.J. June 19, 2014) (limiting scope of discovery into financial matters even when plaintiff was entitled to “a fair assessment of defendants’ financial picture”).

109. See, e.g., LaChapelle v. Fenty, 812 F. Supp. 2d 434, 448 (S.D.N.Y. 2011) (noting both the fact-intensive nature of fair use and the importance of precedent).

110. For example, one consideration that could lurk beneath the surface of the 2023 Warhol v. Goldsmith case is the fact that the Warhol Foundation is a famous and wealthy organization, while Lynn Goldsmith is a more obscure and less wealthy artist. While considerations of relative status might justify putting a “thumb on the scale” for Goldsmith in that case, the ultimate holding—which acknowledges but dismisses the transformative meaning and message of Warhol’s work and rules that it does not constitute fair use because Warhol’s (transformative) work was commercialized for arguably the same purpose as Goldsmith’s underlying work—may, in precedential effect, undermine social justice by making it much riskier for less-privileged artists to commercialize their transformative works. See generally Warhol Found. for the Visual Arts, Inc. v. Goldsmith, 143 S. Ct. 1258 (2023).

111. Id. at 1256.

112. Id. at 1286.

113. Id. at 1276 (“A use that shares the purpose of a copyrighted work, by contrast, is more likely to provide ‘the public with a substantial substitute for matter protected by the [copyright owner’s] interests in the original wor[k] or derivatives of [it],’ which undermines the goal of copyright.” (citing Authors Guild v. Google, Inc., 804 F.3d 202, 207 (2d Cir. 2015))); id. at 1290 (Gorsuch, J., concurring) (“[U]nder the first fair-use factor the salient point is that the purpose and character of the Foundation’s use involved competition with Ms. Goldsmith’s image. To know that much is to know the first fair-use factor favors Ms. Goldsmith.”).

114. See John C. Winfrey, Charity Versus Justice in Locke’s Theory of Property, 42 J. Hist. Ideas 423, 435–38 (1981) (discussing tension between moral concepts of charity and fairness in John Locke’s labor-desert theory); Mark Bartholomew, Trademark Morality, 55 Wm. & Mary L. Rev. 85, 110–14 (2013) (discussing labor-desert fairness in trademark context).