The Wrongs of Copyright’s Statutory Damages
This Article critically examines copyright’s scheme of statutory damages. It argues that the statutory framework and its implementation by the courts do not find adequate support in any of the rationales commonly invoked in the case law, including evidentiary difficulties, retribution, or deterrence. The paper then reviews an alternative rationale: the economic argument for supracompensatory damages as a mechanism for achieving optimal deterrence in the presence of substantial enforcement cost. We find that with respect to a subset of the cases where large multiplayer damages are imposed on individual infringers, the optimal deterrence rationale suffers from two grave defects, both of which are traceable to its welfarist underpinnings. One is ignoring the distributive inequity imposed on the few caught infringers who pay supracompensatory damages grossly disproportionate to their equitable share of the social burden of supporting the copyright system. The other is aggregating on a single scale qualitatively distinct human interests, of different degrees of urgency, thereby risking the swamping of fundamental human interests of a few by aggregates of modest—or even trivial—interest fragments of many. The paper concludes by briefly suggesting directions for amending these defects.
If there was ever “doctrine in search of justification,” it is copyright’s scheme of statutory damages. The Copyright Act gives a victorious plaintiff the right to elect statutory damages irrespective of establishing actual harm. It bestows on courts a broad discretion to set fixed amounts within vast monetary ranges, but provides scarce guidance on how to determine these amounts. Courts consistently stress the discretionary character of statutory damages awards, reject any rigid formulas, and rely on “weighing” a list of factors that end up providing little discipline or meaningful structure. The result is great variance in outcomes. Some awards are quite restrictive both in absolute amounts and in their proportion to estimated actual harm, while others are massively supracompensatory, sometimes involving damage multipliers in the hundreds.
One troubling aspect of this is the inconsistent and arbitrary patterns of awards. Even more troubling is the possible substantive injustice involved in a subset of the cases. The most troubling are those cases in which massive supracompensatory damages are inflicted on ordinary individuals who are unable to spread the impact of such awards through limited liability or dispersed ownership. On its face, there seems to be something fundamentally wrong with an individual of ordinary wealth having to pay over a million dollars for sharing twenty-four copyrighted sound recordings online.
The underlying source of this state of affairs is deep confusion with respect to the animating purpose and associated rationales of copyright’s statutory damages. Courts have suggested several such rationales: evidentiary difficulties, retribution, and deterrence. Upon closer examination, however, none of these rationales justifies the existing statutory damages framework. To make things worse, courts often invoke inconsistent rationales side by side without explaining the relationship between them, thereby thickening the fog in this area. Legal scholarship, for its part, has been mostly critical of copyright’s statutory damages. In addition to criticizing the arbitrary character of awards, the main critique offered in this scholarship is that significantly supracompensatory damages destabilize copyright’s balance by overdeterring access to and use of expressive materials and by chilling innovation in technologies that may be found to be facilitative of copyright infringement.
Meanwhile, judicial opinions and some of the scholarly discussion ignore what appears to be the most plausible justification for supracompensatory statutory damages: the economic argument that, in the face of substantial enforcement cost, such damages are needed for “optimal” deterrence. The argument seems compelling. Substantial enforcement cost is both a waste in itself and a cause of underdeterrence that impedes obtaining the goals of the legal regime. The proffered remedy is simple: boost up damages proportionately to underdeterrence, thereby “optimizing” the ex ante incentives of would-be violators. If infringers have only a 1% chance of getting caught, increase the sanction to one-hundred-times compensatory damages, thereby restoring deterrence to its optimal level. In light of common assertions that in the digital age—where reproduction technology is ubiquitous and infringement is decentralized and hard to detect—enforcement costs are very high, this argument seems tailor-made for justifying extreme supracompensatory damages. For the same reason, the optimal deterrence argument also casts doubt on scholarly criticisms based on overdeterrence and chilling effects. What then ensues is a speculative debate: assertions that optimal deterrence requires extreme levels of amplified sanctions are confronted by counterarguments about how strategic exploitation of threatened sanctions, combined with risk aversion on the part of users, might lead to overdeterrence.
We argue that this scholarly debate is somewhat beside the point, or at least of secondary importance, because it misses altogether the most troubling normative defects of extensive supracompensatory damages. The central wrongs of such damages are not overdeterrence, but possibly grave injustices inflicted on the individuals who suffer the amplified sanctions. When someone might lose her house or be seriously impeded in her ability to make fundamental life choices due to infringing copyright in a handful of recorded songs, it seems strange to focus the debate on whether people in general are over or underdeterred from the use of expressive works. The real wrongs presented by a subset of the cases are twofold: (1) the distributive inequity of imposing a detriment on infringers far beyond their fair share of the burdens involved in securing the benefits of the copyright system (even considering their personal responsibility for infringement); and (2) the aggregation and trading off of human interests that are qualitatively different from each other on a single scale, resulting in the danger of a large aggregate of fractions of trivial individual interests swamping the fundamental or urgent interests of the few.
The reason these concerns are currently invisible lies in the latent welfarist assumptions shared by most participants in the existing debate. The defects of copyright’s statutory damages are thus reflective of more general defects characteristic of the dominant welfarist framework of most legal–policy debates in the U.S. One such defect, which explains how the question of distributive equity vanishes from view, is the adoption of a normative criterion that focuses on the maximization of some total aggregate social number, instead of on the fairness of the effects across distinct individuals’ lives, each equally valuable for its own sake. To be clear, our present criticism of this criterion is not that fairness toward individuals should somehow “trump” social goodness. Rather, it is that there is only one unified normative criterion: fair attainment and distribution of the good, as between distinct individuals of equal moral worth. The second welfarist fallacy is that of illegitimate aggregation. It consists of the reduction of all human interests to a supposedly single quantitative metric, placing these values on a single scale and then engaging in interpersonal tradeoffs between them. The chief danger here is that of low-intensity interests of the many overwhelming high-intensity interests of the few. Ultimately this is another façade of the welfarist attempt to reduce rich and distinct human lives, with the fundamentally different spheres of interests associated with them, to one abstract social measure. The antidote is erecting barriers against such illegitimate interpersonal aggregation of fundamentally different interests, a consequence -sensitive version, if you will, of a “rights as trumps” argument.
We analyze how these two defects of welfarism play out in the context of copyright’s statutory damages. Each of the two concerns is most acute in cases where ordinary individual infringers face very large supracompensatory awards that form a significant share of their wealth. On the distributive equity side, these are cases where the appropriate distributive criterion—one that proportionally takes into account both the incremental benefit or detriment of the relevant effects and the overall holdings of the individuals on which these effects are visited—is likely to militate against amplifying sanctions well before the marginal point of optimal deterrence. On the aggregation side, the same cases are likely to raise serious concerns because the deterrence benefits are likely to be dispersed over a very large number of individuals, while concentrated monetary detriments suffered by the few are likely to trigger more urgent human interests, different in kind than those of the beneficiaries.
To be clear, we do not mount here an all-out attack on the general notion of statutory damages in copyright (or indeed elsewhere). Rather, we believe such damages, when properly designed, could play an important role in copyright. Our criticism here is of their current institutional design. Specifically, this design clings to the pure ideal of market-based compensation through property rights even when this ideal seems not to work. Insisting on implementing that ideal even under conditions of high enforcement cost and highly selective enforcement, in order to achieve “optimal” deterrence, results in grave individual inequities. These inequities are obscured by deeply problematic assumptions of the underlying welfarist framework. There are two possible ways forward. One possibility is that the system can deliver on the goals of the copyright regime without inflicting such grave inequities, in which case statutory damages should be reformed to include proper doctrinal safeguards to avoid such results. The other is that, in light of current conditions, the proper functioning of the copyright system and a robust support of expressive production through traditional property rights requires inflicting these inequities. In that case, the time has come to seriously consider alternatives to copyright’s traditional default of a “property rule” mechanism for underwriting creation, and we should look at available alternatives that do not involve such inequities.
The Article proceeds in four parts. Part I surveys the doctrinal framework of copyright’s statutory damages, both its statutory grounding and the principles and practices that have developed in the case law. Part II presents and critiques the rationales commonly offered in the case law for copyright statutory damages, including evidentiary difficulties, retribution, and simple deterrence. Part III examines the alternative rationale that is not to be found in the case law: supracompensatory sanctions as means for achieving “optimal” deterrence in the face of substantial enforcement cost. This Part first introduces the “optimal” deterrence rationale and then explains its two fundamental defects: distributive inequities and interpersonal aggregation of fundamentally distinct human interests. Part IV concludes by briefly sketching various ways for reforming copyright’s statutory damages in order to remedy their serious normative problems.
This Part provides a brief survey of the existing statutory framework of copyright’s statutory damages remedy and the principles and practices developed by courts in applying it.
The remedy of a fixed statutory sum for copyright infringement predates the American copyright regime. Formally, the only monetary remedy in the first American Copyright Act of 1790 was a per-page fixed sum to be paid by an infringer and split between the copyright owner and the government. This was a direct continuation of a tradition started by the 1710 Statute of Anne in Britain and carried over by the American state copyright statutes of the 1780s. However, a clear line divides these early origins from the modern framework. The early format, with its exclusive focus on an extra-market, fixed sanctions/fees regime, retained much of the original historical character of copyright as a public-regulation regime.
By contrast, the modern system folds in statutory damages as a supplemental layer on top of the default remedy of compensatory damages rooted in market prices. This modern framework first appeared in the 1909 Copyright Act. Its current version is the result of a substantial revision of the statutory damages provision by the 1976 Copyright Act, followed by a few later adjustments. Under this framework, statutory damages constitute an alternative remedy to actual damages or disgorgement of a defendant’s profit, one that may be elected as a matter of right by a prevailing plaintiff and done so “at any time before final judgment is rendered.”
Upon a plaintiff’s election of statutory rather than common law damages, a court has discretion to award such damages as it “considers just” within broad ranges set by the statute. Such an award is to be given “for all infringements involved in the action, with respect to any one work.” The statute provides some additional structure within which to cabin this broad discretion, in the form of differential guidelines for each of three general categories of cases. First, in the default case, the statute provides for a range of $750 to $30,000 per infringed work. Second, in cases where the copyright owner is able to show that the infringement was “committed willfully,” the discretionary ceiling is raised to $150,000. Third, in cases where the infringer is able to show that the infringement was innocent (i.e., that the infringer “was not aware and had no reason to believe that his or her acts constituted an infringement of copyright”), the discretionary floor drops to $200. Despite some disagreement and confusion in the case law about exactly what each of these standards entails, the most plausible meaning of each category’s level of fault is: willfulness requires actual knowledge or recklessness, innocence or “no reason to believe” means non-negligence, and the default category occupies the remaining middle space, namely negligent infringement.
Courts have poured further content into this broad statutory scheme by developing principles and practices for its application in litigation. On the procedural front, the decision maker with respect to the amount of statutory damages may be either the judge or, when requested by one of the parties (since 1998 as a matter of a constitutional right), a jury. In the latter case, the jury determines the statutory damages amount under instructions from the judge regarding the underlying statutory framework, including the monetary ranges. With respect to appellate review, while circuits differ on the exact standard, most are highly deferential to the lower court. With regard to the timing and circumstances of plaintiff’s decision to opt for statutory damages, courts have interpreted the plaintiff’s right of election with generosity and flexibility, allowing election (a) at any time “before final judgment is rendered”; (b) in any reasonable form; and (c) indeed, in the alternative, such that a plaintiff may request a determination of both actual and statutory damages and then only make its election after seeing both figures with courts having developed procedures to facilitate such an election process.
How, substantively, are the awards to be determined? Here there are three main issues: (a) what role, if any, should determinations of actual harm play in the setting of statutory awards?; (b) which, if any, of the three statutory categories should serve as the center of gravity?; and (c) what, if any, further considerations can and should the courts draw upon?
Regarding the first, courts have taken widely different approaches on the required relationship, if any, between the statutory amounts awarded and any established or estimated amount of actual harm. The main point of common assent is that actual damages, where estimatable, are indeed relevant to determining the statutory award. Further, where these are unknown, some courts will also try to approximate plaintiff’s actual harm and defendant’s profits, to serve as guideposts. At this point paths diverge. Although no court strictly places the amount of actual damages as a cap on statutory damages, some do follow the view that “assessed statutory damages should bear some relation to actual damages suffered.” Others, however, take a very different route: insisting that the plaintiff is entitled to statutory damages “whether or not there is adequate evidence of the actual damages suffered . . . or of the profits reaped by defendant,” they proceed to award substantial statutory sums even in the face of indications that actual harm was negligible. Plainly, once statutory damages are completely untethered from actual harm, the way is open for the awarding of sums that may be massive, either in terms of their absolute quantity or in proportion to actual harm, or both.
Of the three categories of willful, innocent, or the silent default of negligent infringement, which, if any, should play an anchoring role? In formal terms of course, the statutory text adopts the mid-range monetary category as the default, against which backdrop it then specifies two additional fault levels that may either raise the ceiling or lower the floor of the default range. The legislative history indicates that the heightened category of willful infringement was envisioned by Congress as applying only in “exceptional cases.” In practice, however, it appears that the application of this tripartite structure tilts toward the higher-end categories. In particular, willful infringement seems to exert a strong gravitational pull, albeit indirectly. A recent empirical study found that an overwhelming majority of copyright plaintiffs plead willful infringement, while in only a small fraction of cases litigated to judgment for the plaintiff (most settle) do courts find willfulness. This trend of overclaiming willfulness is fueled by the tendency of some courts to define the concept in a broad and lax way that potentially encompasses many infringers—including some far afield of the paradigmatic knowing distributor of verbatim copies. This strongly suggests strategic use of enhanced damages claims, either in extracting higher settlement sums from risk-averse defendants or in anchoring and ratcheting up damage awards within the normal range. Conversely, courts tend to define innocent infringement so narrowly that many nonegregious infringers find it hard to come within its harbor. Finally, the pull of the damages on the higher end is reinforced by a tendency, exhibited by at least some courts, to use, within whichever category they are operating, the maximal-damage figures or ceiling as the starting point from which to work backwards or, even more simply, as both the starting and stopping point.
Finally, are there any other considerations (besides actual harm, culpability level, and the floor–ceiling ranges) that courts draw upon, and in any case how do they purport to weigh the ones they do consider? Little guidance has been forthcoming on either front. Common starting points are a recognition that “the statute does not afford much guidance as to how courts are to fix appropriate amounts in statutory damages cases” and reference to the “wide discretion the Copyright Act affords the trial court in setting the amount of statutory damages.” Beyond that, courts rely on a flexible and somewhat shifting list of “factors,” eschewing any “mathematical formula or equation.” One commentator summarized the “most common” factors “recited” as: “the relationship between the statutory damages sought and any actual damages or profits, whether the infringement was willful or innocent, the need for deterrence, defendant’s past infringement record, defendant’s cooperation after the matter was brought to its attention, and the scope of the infringement.” Additional ones have included any deterrent effect on third parties and “the conduct and attitude of the parties.” Although courts sometimes highlight a few factors as central—typically a defendant’s culpability or conduct—there is also a general insistence that “[n]o one factor is more important than any other.” A typical analysis simply involves “consideration” of a subset of the listed factors, with no structured relationship between them and no one clear guiding aim or logic. The result: great variance of approaches and outcomes, leading one commentator to observe that in this area “standards are largely precatory.” Others observe that “courts have yet to develop a meaningful jurisprudence to calibrate how to render ‘just’ statutory damages awards” and conclude that such awards are “frequently arbitrary, inconsistent, unprincipled, and sometimes grossly excessive.” In this vast, open terrain one may find anything from minimal awards hewing close to the floors to ones tracking estimates of harm to massive awards grossly disproportionate to any harm, sometimes with little harm in sight. Among the latter kind, particularly troubling are the occasionally massive awards inflicted on private individuals, including those acting noncommercially.
The fluid state of the doctrinal standards for awarding statutory damages is a symptom of a deeper cause: a contested, somewhat obscured, and even outright confused understanding of the orienting purpose of this area of the law. In light of the dearth of statutory guidance on this issue, courts have offered several conflicting rationales for statutory damages. These are often stated in murky ways, with distinct, incompatible justifications swirled together. Upon inspection, however, none of the standard justifications offered by the courts can plausibly support anything close to the existing framework of copyright’s statutory damages.
A first possible purpose of a fixed schedule of statutory damages is to address evidentiary difficulties in establishing the actual losses suffered by the plaintiff in a particular case of infringement. In some cases, a copyright owner, due to a lack of relevant information, may be unable to prove on the required level of certainty a significant part of the losses actually caused to them by the infringing activity. When this occurs, whatever purpose is served by compensatory damages for infringement is being frustrated. Statutory damages then step in as a second best. Their role is to try to approximate, albeit imperfectly, the amount that would be recovered as actual compensation. Their awarding is a way of avoiding the least desirable result of no compensation at all, in the face of actual but hard-to-prove damages.
Evidentiary difficulties were explicitly invoked by some older authority as the rationale for various statutory damages provisions predating the modern 1976 one. There is also at least some suggestion in the legislative process leading to the 1976 provision that evidentiary difficulties were a central underlying concern. And modern courts often mention evidentiary difficulties as an important purpose of the current arrangement. At the same time, however, while some decisions centrally emphasize the importance of this purpose, more often it is listed as only one rationale among others.
Yet, the evidentiary rationale cannot justify central features of the existing statutory damages framework as applied by many courts. This rationale does not give open license for awarding large sums of damages completely detached from the actual harm suffered. Crucially, understood as a remedy for evidentiary concerns, statutory damages offer a very imperfect substitute for navigating by what is the true remedial lodestar—i.e., actual compensatory damages. If we were to take seriously the second-best status of statutory damages, two principles would follow: (1) first, such damages should not be awarded at all when actual harm is clearly established or could readily be so; and (2) second, even when actual harm remains unknown, the criteria for calculating the fixed statutory sum should be tethered as closely as possible to such harm, by way of approximation to this unknown, but perhaps estimatable, amount.
Both the statute and the case law are far removed from these principles. The statute does not limit entitlement to statutory damages to cases where establishing actual damages is impossible or difficult, and the case law is uniform in treating this entitlement as a matter of unqualified right. While some courts would keep the award close to the statutory minimum when there is no indication of actual damages, this practice is by no means uniform. Similarly with respect to tethering of the award to an approximation of actual damage, the statute makes no such requirement and simply leaves it to the courts to determine an amount within its very broad ranges. Courts, for their part, have not developed rules that strictly cap statutory damages to an approximated damage, no matter how generously defined. Even those courts that pronounce the centrality of actual damages as a “factor” refrain from using it as a cap and in practice, tend to award modest multiples of that sum. Other courts go much further, sometimes awarding dizzying multiples of estimated harm. Furthermore, there is a wide practice of allowing plaintiffs to elect statutory damages after receiving judgment with respect to actual damages, a practice that would seem senseless but for the assumption that statutory damages are permitted to be significantly higher than actual damages.
Finally, another central feature of the statutory provision—its hitching of awards to levels of culpability—has no basis at all in the evidentiary rationale. Trying to estimate levels of harm or tethering fixed amounts to such approximations simply has nothing to do with levels of fault.
In short, central aspects of both the statutory setup and the case law either have no basis in the evidentiary rationale or directly conflict with it. Absent fundamental reform, such a scheme simply cannot be said to be based on, or even track in any meaningful way, the aim of correcting for evidentiary barriers.
The primary alternative rationale offered in the case law to evidentiary difficulties is that statutory damages serve as “punishment” for the wrongful act of infringing copyright. Many courts pronounce some variation of this claim, although they often accompany it with a confusing repudiation of the “punitive” character of the remedy. In any case, merely asserting that some measure is a “punishment” neither explains its rationale nor provides meaningful guidance for its application. One has to explain what the point of punishment is, what purpose it serves. And although seldom explicitly articulated, what most likely underlies a “punishment” view of statutory damages is the notion of retribution. Retribution, at its core, is the proposition that it is simply right that a wrongdoer suffer in a manner proportionate to the severity of the wrong they committed.  That is, over and above any deterrent effect from imposing a sanction, the thought is that a wrongdoer simply “deserves” to be punished or to suffer a harm in proportion to the one they caused. As applied to copyright, the view would be that it is only right that infringers, who wrongfully harmed a copyright owner, be inflicted with some suffering or detriment proportionate to the wrong caused by their infringement.
Retribution fits, then, hand-in-glove with a key feature of the statutory scheme escaping the evidentiary justification, namely the role of fault levels in setting awards. Within the natural habitat of retribution—criminal and, to an extent, tort law—the level of the wrongdoer’s culpability is, alongside the extent of the harm caused, generally thought of as the central determinant of the magnitude of the wrong. The wrongdoer’s fault—understood as her state of mind with respect to the harmful consequence—shapes the wrongfulness of the act and that, in turn, reflects on the magnitude of the appropriate sanction. An identical act resulting in identical harm is more wrongful, and therefore requires a more severe punishment, depending on if it was committed negligently versus recklessly versus intentionally. This logic is clearly reflected in the tripartite culpability scheme of copyright’s statutory damages.
Nevertheless, there are a host of reasons why retribution is a very poor justification for copyright’s statutory damages. Most fundamentally, we simply find the core logic of lex talionis—an “eye for an eye” in the “law of revenge”—to be unpersuasive: perhaps understandable as an individual reaction but, upon reflection, indefensible and unattractive to pursue socially. No good or equitable aim is served by insisting that a wrongdoer suffer for no purpose other than itself, irrespective of any other goal such as compensation for the victim or deterrence (more on that, shortly).
In any case, even if one were to be persuaded (as we are not) by the logic of inflicting harm on a wrongdoer for its own sake in some contexts—a kind of “surplus suffering” to be imposed independent of any other consequential effects—copyright’s civil remedies appear to be a particularly ill-suited candidate for it: the typical injuries suffered by plaintiffs (i.e., to their purse rather than their person) and acts engaged in by defendants (i.e., use of nonrival goods rather than invasion of personal interests in bodies or homes) are far from those standardly featured in accounts of punishment for its sake. This is closely connected to a further point: the fact that copyright law already designates certain cases of infringement as criminal offenses punishable by criminal sanctions. It seems an implausible, indeed incoherent, legislative strategy to carefully define criteria for a subset of copyright violations to be treated as criminal offenses and mete out corresponding criminal sanctions and then, also pursue retribution—a classic justification of criminal sanctions—for conduct not rising to the level of those carefully circumscribed conditions through the civil remedy of statutory damages. At least in the absence of express congressional edict to that effect, such a convoluted structure should not be imputed to the statute.
Finally, when criminal copyright enforcement is pursued, all the safeguards of criminal procedures, such as a heightened burden of proof or various evidentiary rules, are in place. This is not the case with statutory damages. If creating an alternative route for pursuing retribution, bypassing the criminal copyright provisions, is an unattractive and implausible scheme, it is doubly so when one realizes that this civil bypass strips defendants of all the procedural and evidentiary protections of criminal procedure.
In the end, however, these additional weighty reasons are secondary. What matters most is that the logic of an eye for an eye (for no purpose other than inflicting suffering on a wrongdoer) is unattractive in general and particularly inapt as the rationale for a civil remedy for copyright infringement.
Perhaps sensing the troublesome character of the retributive aim, courts, including those who embrace the punitive view of statutory damages, seem to recoil from explicit endorsement of retribution. While judges often declare that the wrongdoing of copyright infringement should be punished, it’s hard to find an opinion that expressly adopts a crisp version of the retributive view. Instead, when a reason is at all given for the penalty, it is that of deterrence or discouraging infringement. And courts often observe that deterrence requires a supracompensatory sanction. What occurs in courts’ opinions and jury instructions then, is a conflation of rationales: retribution is imported from the criminal and tort contexts to justify sanctions gradated by fault levels, followed by a shift to deterrence as the grounds for justifying the practice. Retribution endows the structure with a veneer of logic, while its shaky normative foundations are evaded by a seamless switch to deterrence.
Deterrence suspended in a void is, however, an empty concept. One needs to know why we deter in order to know how much to deter. In particular, to justify existing statutory damages one needs to explain why deterrence justifies the nondefault categories of innocent and willful infringement and the different rates attached to them. As for innocent infringement, a moment’s thought should show that this category finds no support in deterrence. If the infringement is truly innocent, meaning non-negligent, why would we want to deter people from engaging in the relevant activity? Non-negligence means that it is unreasonable to expect infringers to avoid the infringement. In familiar cost–benefit terms, the cost of prevention—meaning the cost of discovering the infringing character of the activity—is smaller than the ex ante value of the harm. Why would we want to deter people from engaging in reasonable behavior? Moreover, since rational infringers would only invest in prevention up to the level of the ex ante expected damage they would internalize, the only way to effectively deter them in such cases is to impose a supracompensatory sanction. The result: wasteful investment in prevention outweighing the benefit of the averted harm. The alternatives are either nondeterrence or wasteful overdeterrence. In short, deterrence does not justify any damages when the infringement is truly innocent.
What of willful infringement? Perhaps surprisingly, deterrence also does not support any special treatment of this category. Once the negligence line is crossed, a “willful” state of mind simply has no additional bearing on the justification for, or the means of, deterring. Whether we regard knowing or reckless behavior as worse than negligent behavior does not affect the deterrence analysis. Both cases merit deterrence, meaning discouraging people from engaging in the relevant behavior and investing in discovering their infringing status. Under simplifying assumptions, in both cases full compensation for the damage is sufficient to deter. If, on the other hand, compensatory damages are insufficient to dissuade infringers, it is unclear why this would apply differentially to willful and negligent conducts. In both cases, once the sanction is high enough to encourage the actor to invest in effective prevention and avoid an action revealed to be infringing, any additional increment finds no justification in deterrence. Such deterrence-inert additional increments can only be justified in retributive terms, meaning the infliction of suffering on a wrongdoer for its own sake.
To support any award beyond actual damage or its approximation, one needs a deterrence theory that can justify supracompensatory damages, and this is equally true of negligence and willfulness cases. Yet, notwithstanding the frequent invocation of deterrence, no such theory is forthcoming in either the case law or the treatises. Why then might we want to achieve a level of deterrence above and beyond that already created by compensatory damages? One could try to answer this question by falling back on one of the rationales already discussed: evidentiary difficulties or retribution. Each of these answers, however, not only takes us back to the litany of problems pointed out above, but also fails to supply support or meaningful content to the idea of supracompensatory deterrence. The evidentiary rationale is about trying to approximate as a rough second-best the level of deterrence achieved by compensatory damages. Retribution, properly understood, is about inflicting a detriment on the infringer for its own sake, and thus not about deterrence at all.
What, then, is left? As Professor Depoorter notes in his important study, the one plausible, at least on first blush, theory of supracompensatory deterrence is the one that courts seem to studiously avoid in the context of copyright’s statutory damages, namely achieving optimal deterrence in the face of substantial enforcement cost. In what follows, we briefly explain this theory and then discuss its grave defects in the context of the existing statutory damages framework.
The key concept for the optimal deterrence argument is enforcement cost. The moment this cost is considered, the luster of compensatory damages as an ideal deterrence mechanism dims. The benefits of deterrence are secured with an attendant cost—that involved in the enforcement necessary for achieving it. And as importantly, once we reach the marginal point where this cost does not justify purchasing additional deterrence, the outcome is frustration of attaining full deterrence benefits. Arguably, in the field of copyright, especially in the modern digital context where potential infringement is highly decentralized, widespread, and costly to monitor, the twofold ills of enforcement cost are significant. First, any cost invested in expansive copyright enforcement is itself a social loss. And second, since the cost-effectiveness tipping point arrives much in advance of full enforcement, deterrence and its benefits are significantly degraded. With a small chance of being caught, many may infringe, thereby undermining the social benefits of the copyright system.
Can anything be done about this problem? The standard economic counsel is to increase sanctions above the compensatory level. Ramped-up sanctions increase the deterrence yield per unit of enforcement cost. Infringers happy to infringe due to a small chance of getting caught are likely to think again if, when this chance materializes, the sanction is much more painful. The result is some combination of decreased enforcement cost and increased deterrence benefits. By how much should the sanction be increased? In the setting of civil copyright, where the main factor driving underdeterrence is the low probability of detection, the answer is: until the point where the ex ante magnitude of the increased sanction discounted by the probability of detection equals that of full compensation under certainty of enforcement. A smaller sanction would underdeter and a larger one would overdeter.
Conventional economic wisdom depicts the effect of increased sanctions through the narrow lens of its boosted deterrence effect on the stylized individual figure of the potential infringer. This translates into the following broader picture of social effects. Ex ante, up to the limit of overdeterrence, boosted damages increase (otherwise deficient) deterrence, resulting in better attainment of the goals of the copyright regime. Isn’t there a countervailing ex post cost in the form of the negative effect of the sanction on caught infringers? No, comes the answer, because the (over) compensatory sanction is “just a transfer” from the defendant to the plaintiff, a self-offsetting reshuffling of identical value whose effect can therefore be ignored. This leaves as the only ground for debate the question of whether the increase in damages has crossed the overdeterrence limit, a limit which is many times the actual harm if one accepts the underlying assumptions of copyright’s radical underenforceability. The result: a powerful deterrence justification for supracompensatory damages. One that supports even very large magnitudes of increased sanctions when very high enforcement cost results in very low probability of enforcement. Under this logic, when the chance of enforcement is 1%, a hundred-fold damages award is justified. In other words, it is a justification tailored exactly for what appear to be the most troubling cases of statutory damages, in which infringers are inflicted with very large awards grossly disproportionate to any actual harm resulting from their conduct.
What is wrong with this rosy picture of boosted sanctions imposed on the few, in the name of the greater good achieved by the resultant “optimal” deterrence? Two crucial things: (1) it ignores the distributive inequity inflicted on the unlucky ones who suffer the brunt of the extreme sanction; and (2) it invites illegitimate aggregation, weighing on a single scale radically distinct kinds of interest, in a manner that threatens to sacrifice the fundamental interests of a few for the sake of modest or even trivial benefits dispersed across many. We take up each of these in turn.
The first defect of large-multiplier supracompensatory damages is their potential distributive inequity. What does distributive justice—or fairness in the overall distribution of benefits and burdens among members of society—have to do with copyright damages? The answer is that the “cost” of copyright damages—the negative effect on those who have to pay them—is a social burden, one that is undertaken in order to achieve desirable results in the form of the benefits of the copyright regime. And, like any other social burden, it should be distributed equitably.
This tends to be obscured by the fact that, in this case, the burden of supporting the social benefits of copyright takes the form of damages paid by those who infringed copyright. But this obstacle should be overcome: damages are just another form of social burden that, like any other, should be distributed equitably. To see this more clearly, assume a hypothetical prize regime in which expressive creation is supported not by a proprietary exclusionary right, but by a tax imposed on users with the proceeds distributed among creators. Further, assume that in order to achieve significant savings of administrative cost (which translate into better achievement of the benefits of the system), the tax is not imposed on all users but only on a small fraction, say 1%, of them. Each of the unlucky 1% has to pay a much greater tax in order to raise the necessary funds, but the savings are significant. Do these few have cause for complaint? They clearly do: by imposing a grossly disproportionate share of the cost of obtaining the social benefits of supporting creation on a small fraction of the beneficiaries, this regime violates any plausible principle of equitable distribution of social burdens. An equitable distribution would require, as a rough first approximation, that each pay roughly in accord with the benefit he derives from this system. While here, some pay very much more, and many others, who benefit equally, pay nothing.
Is this hypothetical fundamentally different from the imposition of the cost of the copyright regime on the few caught infringers through damages grossly disproportionate to the benefit they derive from this regime? We think not.
One obvious objection to the parallel is that it ignores personal responsibility. The hypothetical taxpayers have done no wrong and do not deserve their disproportionate burden. The infringers, by contrast, “broke the law” and bear personal responsibility for frustrating the attainment of the social benefits of the copyright regime. Didn’t they therefore forfeit their right to complain of any distributive inequity? And don’t they deserve the disproportionate burden laid on their shoulders? Not quite. One who is inclined to take the personal responsibility consideration seriously here must take as seriously the principle that the amount of responsibility be commensurate with the wrong. This principle follows from the basic logic of “just deserts” that is at work here. If an infringer does deserve to shoulder a greater burden, he only deserves to shoulder more in proportion to his fault, no less and no more. In our context, this individual fault premium may well be a fraction of the increased enforcement cost attributable to the specific actions of the infringer or those similar to it. This is likely to be insignificant relative to the magnitude of large damage multipliers imposed on a few infringers in the name of optimal deterrence. When the deterrence multiplier is large, the effect of any individual responsibility premium to be imposed on infringers will quickly wash away in the overall distributive equity calculus.
A more plausible response, but one that applies equally to the tax and copyright damages scenarios, appeals to the countervailing social benefits of the deterrence. Even if the burden imposed on the unlucky infringers caught is thought to be unfair, shouldn’t we also consider the social benefits achieved by this measure, in terms of more effectively attaining social good of the copyright regime? And isn’t it then justified to sacrifice any fairness concerns to obtain such “goodness”? This response is on target, but its framing is misguided. The real normative question here is not that of fairness versus good consequences, as in two conflicting goals that never the twain shall meet. Our aim is not to realize some notion of fairness independent of the consequences, nor to pursue good consequences irrespective of fairness, but rather to realize fairness in the pursuit of good consequences. Our enterprise, in other words, is a single, unified one, whose central question is: what is the most just alternative in terms of attaining the relevant social good and equitably distributing it among the affected individuals?
Answering this question requires a crisp understanding of several aspects of the enterprise. First, the distributive analysis here does not try to graft onto copyright some external aim, such as the correction of general social inequities writ large, in the overall distribution of income, wealth, or other goods. Rather, its aim is one squarely internal to the copyright system: to make sure that the social outcomes generated by the copyright regime are themselves equitable, an aim whose proper place, we believe, lies within the field of copyright rather than with any external tax-and-transfer scheme. Second, the question is one of substantive fairness in distributing a social good across distinct persons leading separate lives, each life being of equal worth or importance. Third, in light of this aim and that commitment, what are the relevant considerations for fairly distributing the burdens and benefits of the copyright system? A good rough starting point is that individuals should bear that share of the system’s cost that is roughly commensurate to the benefit they derive from it, on a principle of responsibility for one’s preferences. But fairness also requires taking into account a principle of priority: to the extent somebody is, through no fault of their own, worse off than others, then her claims with respect to the burdens and benefits disposed by copyright merit heightened concern. Finally, fairness also requires taking into account any countervailing benefits enjoyed by some due to measures that might clash with the first two principles.
Against this backdrop, how does the standard account of optimal deterrence fare? This standard account, based on economic efficiency, counsels that disproportionate burdens should be laid on infringers so long as these burdens are outweighed by the total social benefits of improved deterrence. The problem with this position is that it compares burdens and benefits as aggregate social sums. We are dealing, however, with a question of fairness as between separate individuals leading distinct lives. There is no entity or social “being” to which we can impute the enjoyment or suffering of aggregate costs or benefits. What then is efficiency’s criterion once it is revealed for what it is: i.e., just one contender among others for the fair way of distributing costs and benefits among distinct individuals? It is that any individual burden is justified so long as other individuals enjoy a greater marginal benefit. The fatal flaw with that should be plain: in adjudicating competing individual claims, this criterion (appropriately) takes into account countervailing benefits, but it completely ignores any greater need or priority for those individuals whose starting position is worse off. In deciding whether it is fair to impose on me a burden of ten, it is relevant if you are going to get a benefit of twenty, but it is also relevant if I have 100 fewer than you to begin with. Efficiency focuses its concern entirely on the former, ignoring completely the latter.
Is there an alternative? In other work, one of us has developed a principle of distributive equity—the principle of proportionate priority. That principle’s central guiding commitment is that persons are owed priority in the distribution of goods because, and to the extent that, they are worse off than others through no fault of their own. From that single commitment emerges a two-sided implication. First, persons enjoy distributive priority over others if and to the extent that they are worse off than these others through no fault of their own. Second, a person also enjoys distributive priority over others if and to the extent that they stand to reap a greater benefit than these others from the good being distributed. That is, in prioritizing the principle not only considers both overall levels and marginal gains, but automatically integrates the two within the ambit of a single orientation. Specifically, individuals’ priority claims are greater than others exactly to the extent that their starting point is lower than the others, with said priority weight then applied to their respective potential gains. This issues in the following principle of distributive justice: persons receive distributive priority over others so long as the relevant benefit for them, as a proportion of their existing level, is greater than the benefit for the others, as a proportion of their existing levels.
Applied to copyright statutory damages amped up by significant multipliers, the principle of proportionate priority causes distributive hazard lights to turn on. Two factors are relevant for assessing the distributive equity of such damages. The first is the substantiality of baseline compensatory damages relative to the infringer’s initial holdings. While substantial compensatory damages, as such, may be justified on personal responsibility grounds, the moment deterrence multipliers are introduced the analysis changes. To the extent that baseline damages are substantial, any added deterrence multiplier will render the infringer’s starting point significantly worse off, thereby steadily pulling the distributive calculus in her favor. The second and obvious factor is the magnitude of the damage multiplier required to achieve optimal deterrence. The larger this magnitude, the more significant the adverse effect on those who bear the burden. When the baseline damages are substantial and the multiplier is large, the marginal point where priority-adjusted burdens on the infringer outweigh deterrence benefits will be reached at a sanction level much lower than the one counseled by optimal deterrence.
The most troubling cases of excessive statutory damages fit this mold. Unlike parking tickets, the baseline of compensatory damages is very substantial. In some cases it could reach a significant share of an average individual’s wealth. Meanwhile, very large damage multipliers cause both the negative effect on infringers and the degradation of their initial starting point to swell quickly. The result, as mentioned, is that the marginal point where need-adjusted detriments to infringers outweigh social deterrence gains is reached well before the optimal deterrence level. This is especially true in the most troubling subset of cases, i.e., those that involve the imposition of very large sums on private individuals rather than business firms. When such individuals are involved, the mitigation of the concentrated effect on wealth levels through dispersal over many stakeholders and limited liability is much less likely. The possibility of a cap in the form of such individuals being judgment proof beyond a certain level of damages is no consolation. It is, rather, further indication that under the analysis of proportionate priority the results are distributively unacceptable.
Consider now a distinct possible fairness concern with “optimal” deterrence damages. The nub of the “optimal” deterrence argument is, recall, that the greater the cost of enforcement, the more we need to ratchet up supracompensatory damages against those who are actually caught: to offset the lower probability of detection requires a higher ultimate fine, to equal the same amount of expected harm. What if the enforcement cost is high enough to make the probability of detection 1-in-100? Then amp up the fines accordingly, with a multiplier of 100. What if it is 1-in-1,000? Then a multiplier of 1,000. And so on. The point is that with each turn of the ratchet, a more concentrated harm is imposed on a few, for the sake of a benefit the discrete gains from which are dispersed among a larger and larger number in ever-smaller unit amounts. At the limit, a few high-salience enforcement actions could serve to shore up the entire copyright system.
What might be wrong with such a scheme of imposing highly concentrated costs on a few for the sake of greater overall benefits spread across very many? More specifically, what is wrong with it in a way that is not already captured by a concern with distributive equity? The trouble with such a scheme is that it simplistically trades off, on a single quantitative scale, fundamental or urgent interests of a few against qualitatively distinct, modest or trivial, benefits spread over many. So long as the numbers affected on either side are disparate enough, modest or even trivial gains aggregated over a large enough number of persons will always, no matter how insignificant in quality or small in quantity the actual benefit per person, threaten to run roughshod over urgent or even fundamental interests of a few. Any qualitative difference in the competing interests will fail to be adequately captured by a metric registering only their quantitative intensity on a single scale. Such a metric simply pries apart units of “goods” or “bads” from their actual role in distinct persons’ lives, abstractly quantifying and then summing them up in the aggregate.
How are such aggregation concerns distinct from those of distribution? Distributive concerns are triggered even in “same numbers” cases—i.e., when those affected on either side of the ledger line up in number or are closely similar. Thus, even if only two persons were to be affected by a policy, one to gain and the other to lose, distributive considerations may be in play if one is significantly worse off than the other through no fault of her own. To ensure an equitable distribution, we would look not only to the parties’ respective prospects for marginal gain but also at their respective overall levels of well-being. In a purely distributive analysis, that is, our concern is simply with giving priority to those overall worse off, as a check against merely maximizing marginal gains across persons. To be sure, in the real world it is not always “same numbers” on either side: we may have a potential harm to one person that needs to be weighed against potential gains spread over 2, or 5, or 20, or 100, etc. What then? A purely distributive view would simply ignore the issue of aggregation (i.e., different numbers on either side) and continue to apply its distributive principle or function to this new population set, so that we just tally equity-adjusted harms and sum them up across the numbers affected on the one side of the ledger and compare to equity-adjusted gains summed across the numbers affected on the other side.
But at some point we may wish no longer to ignore the numbers affected on the two sides. At which point? At the point at which (a) the discrete, per-person harms versus benefits on the two sides reach a very large disparity, and (b) that disparity is masked by the countervailing disparity of numbers affected on the two sides. At this point, very significant interests concentrated among a few may be unduly sacrificed to modest—even trivial—benefits dispersed across a very large number. And it is here that a purely distributive approach is likely to run out. This is because the benefits on the many-numbers side will likely equal a high enough magnitude in the aggregate to overrun those on the other side (even when both sides are adjusted by distribution weights). With the concern being that the former are widely dispersed in very small units across a very large number, while the latter are
highly concentrated in large doses among a few—such that the aggregated sums of quantified intensity mask—indeed disfigure—the underlying qualitative significance of the affected interests.
The pursuit of distributive equity, in other words, does not by itself secure against aggregation effects: very small marginal benefits (even after these are distribution-adjusted downward) may, when accumulated over a large enough number, begin to swamp very high marginal costs (even after these are distribution-adjusted upward) that are concentrated on a very few. For instance, as a principle of purely distributive equity, the principle of proportionate priority would allow any kind of harm to one or a few for the sake of any trivial benefit to others, so long as the latter were spread across a large enough number. This is because that principle’s attention, like that of any other principle of purely distributive equity, is simply not directed to the relevant concern here: it is focused on issues of who is overall worse or better off among the similar numbers, while the issue here is about concentrated or dispersed harms and benefits across radically disparate numbers.
Does copyright’s statutory damages scheme pose such aggregation concerns, in addition to those of a purely distributive sort? On the concentrated harms side of the ledger the answer seems clear, at least in some cases: certainly the damage awards at the higher end, which can involve sums as high as $600,000 or even $1,900,000 in certain high-profile cases against private individuals acting noncommercially, are enough to raise alarm bells. Whether aggregation alarms should sound also depends on (a) how large the countervailing benefits from improved deterrence are and (b) how widely dispersed such benefits are. However, at least if the copyright system functions as hoped, there seems to be a strong prima facie case for serious aggregation concerns. The general assumption is that copyright does not simply funnel the bulk of the social value of creation into a few hands, but facilitates its dispersal over many users, consumers, and future creators. The benefits of improved deterrence, even if large, are likely to be very broadly dispersed. Meanwhile, the large award sums imposed on the few caught infringers are likely to translate into concentrated effects in realms of fundamental interests. A $1,000,000 or even $500,000 detriment inflicted upon an individual of average wealth is bound to have serious effects in areas such as housing, health, education, and more generally the ability to make and pursue basic life choices. Such effects on urgent interests of the few being swamped by a multitude of fractions of improvements for the many, in entertainment, consumption, or even education, is exactly the heart of the aggregation concern.
Where does this leave us? Let’s briefly review. First, a statutory scheme is in place where (a) damages of up to $150,000 may be awarded per infringed work; (b) damages are adjudicated in a very open-ended, largely discretionary manner, sometimes entirely delinked from any showing or estimates of the actual harm caused by the infringing activity; and (c) damage awards undergo modest, and similarly unstructured, appellate scrutiny. Second, none of the rationales most commonly adduced for this scheme—evidentiary difficulties, retribution, or standard deterrence—hold up, either as justifications for many of its key features (such as delinking from compensatory harm or heavy reliance on culpability levels) or as compelling guides for its application going forward. Third, the most plausible rationale is one that courts have shied away from advancing, that of “optimal” deterrence: imposing very high fines on those caught to save on the costs of detection and enforcement against others. There is good reason why courts are shy of embracing this rationale: its “efficient” saving of enforcement cost—so as to secure the highest overall benefits of copyright protection at lowest overall cost—is radically unfair in its incidence of these costs and benefits. And this unfairness can take either of two distinct forms: (a) first, in cases where similar numbers of people are affected on both sides, it fails to take into account basic distributive equity by adjusting or weighing the marginal costs and benefits to account for differences in how well-off overall the respective burden-bearers and beneficiaries may be; (b) second, in lopsided numbers cases, it fails to check against aggregation concerns, whereby highly concentrated costs may be imposed on a few for the sake of modest or even trivial benefits dispersed across very many.
What is to be done? We offer here a sketch of three sets of possible reforms, starting with the most modest and straightforward and ascending in scale of ambition and contention.
At the very least, our analysis in Part II suggests three reforms to existing judicial practice in this area as indispensable, based solely on considerations internal to the existing doctrinal framework and its standard rationales. First, the existing practice of awarding statutory damages with little to no attention paid to their underlying purpose is unacceptable simply on grounds of procedural fairness: the discretionary, inconsistent, and largely arbitrary awards that result are an affront to elemental rule-of-law values of transparency, horizontal equity, and prospectivity. Second, of the purposes standardly vaunted for such damages, two—retribution and standard “deterrence” claims—simply do not hold up as meaningful justifications or guides. This leaves standing only a third, evidentiary difficulties, and even this on wobbly legs given its radical ill-fit with a central feature of the statute, namely its reliance on culpability levels. Nevertheless, if we wish to hew closely to only those rationales explicitly offered by the courts, then our conclusion must be that statutory damages, moving forward, find their justification solely in evidentiary or administrative difficulties facing the determination of actual harm, and hence the awarding of such damages must remain tightly tethered to evidence or estimations of such harm. In a word, statutory damages must remain purely compensatory in character.
Courts have considerable room to follow this course even within the constraints of the statute’s ranges and the legislative history guidance that a plaintiff “is not obliged to submit proof of damages and profits” to be awarded statutory damages. Within these constraints, courts could follow the compass of compensatory damages by allowing defendants to submit evidence of actual harm or profits, engaging in estimates of plausible, maximal figures for these measures when information is absent, and emphasizing in jury instructions and judicial reasoning that the sole guiding light for exercising the discretion is the approximation of actual harm or profits and not the addition of any premium in the name of retribution or deterrence. Trial judges could use their standard tools to keep jury awards within these guidelines, and appellate courts could review lower courts’ decisions with the same aim.
What if, however, we wish to leave conventional grounds and embrace supracompensatory damages for the sake of pursuing “optimal” deterrence, despite the courts’ own reluctance to embrace that rationale? Two initial points follow immediately. First, just as in the case of the evidentiary rationale, so here judicial elaboration and pursuit of this rationale must entirely break free from the aroma of retribution that still surrounds statutory awards, enveloping the decisions and reasons in a thick fog of punitive language and sentiment that is neither apt as a matter of justification nor provides any sensible guidance. That is, principled pursuit of both evidentiary and “optimal” deterrence concerns requires studious judicial disregard of any punitive considerations. In one case, our aim is purely compensatory, and in the other supracompensatory, but in neither does it have anything to do with levels of fault or culpability.
This, however, would put the courts in somewhat of a quandary, requiring them to ignore the plain language of the statutory text, with its clear indications of the relevance of fault, language that either party may invoke in a given case. Consequently, a second indispensable implication that follows from this counsel is the need for legislative reform to the language of the statute by abolishing its use of culpability categories. Absent such reform, the fog of retribution will threaten constantly to re-envelop the dispensation of statutory damages.
Would such a statutory modification be enough? Not by our lights. This is because the underlying aim of “optimal” deterrence is itself indefensible, at least in its unqualified economic and welfarist form, owing to the criticisms raised by distributive equity and aggregation concerns. What prescriptions for reform, then, issue from these concerns? Three main options present themselves, two lying within the traditional model of copyright and a third moving beyond.
A first option is to reign in the pursuit of “optimal” deterrence by instituting judicial safeguards against the incurrence of serious inequitable harms. Recall that distributive harms are likely to occur whenever the total damages that result from applying a multiplier to a compensatory base amount to a total high enough to reduce defendants’ overall income or wealth to the point that they merit especially strong distributive concern. Aggregation concerns, meanwhile, step in when the offsetting benefits to such harms are high but accrue in very small increments dispersed across a large number of persons. What this counsels in terms of judicial safeguards is twofold: First, when the defendant is not an individual person but a firm with either limited liability or dispersed ownership, then neither distributive nor aggregation concerns are likely to prove worrying enough to require any safeguard. Second, however, when the defendant is an individual person, then an equity ceiling or cap to protect against distributive or aggregation harms is warranted, as follows: supracompensatory damages per infringing work should not pass a threshold level beyond which they would so eat into an average individual’s income or wealth that (a) the marginal benefits to others in terms of securing them the goods of copyright protection at lower enforcement cost are unlikely to outweigh the marginal harms to the defendant, when both of these are distributively adjusted to take into account their overall levels of income or wealth; or (b) even when the aggregate marginal benefits would outweigh the harms, even after both sides are distributively adjusted, nevertheless the benefits are dispersed across such a large number as to amount to trivial or modest per-person gains, while the harms burden fundamental or urgent interests of the defendant. What would that threshold level be? A precise answer would require better information than we presently have on the average income and wealth levels, and numbers of persons affected on both sides. A very rough and serviceable proxy, however, may be that when supracompensatory damages, i.e., those to which a multiplier has been applied, exceed $50,000 per infringing work, they pass that threshold.
Of course this is subject to refinement upon a closer evaluation of the relevant data, should it become available. What if that closer assessment reveals that safeguarding against distributive and aggregation harms requires such fine-grained calibration as to not be judicially administrable? At that point the answer seems clear: we should simply call off the pursuit of “optimal” deterrence in cases involving noncorporate defendants, deciding that the risk of equitable harms is not worth the chase—especially given judicial reluctance even to expressly embrace this rationale.
What if either reigning in or calling off pursuit of “optimal” deterrence threatens to undermine the very goals of copyright in a digital and online world? The contexts most in need of “optimal” deterrence are of course those most affected by the technologies enabling high-fidelity and low-cost reproduction and distribution of expressive works. In such contexts, what the distributive and aggregation concerns present is an invitation to revisit the wisdom of traditional copyright in a new world. Specifically, what they suggest is that we may wish to move away from a strong “property rule” model of copyright—whereby the enforcement of rights is secured by threat of litigation ending in ex post, case-specific remedies of injunctive relief and lost profits—and toward more of a “liability rule” model, whereby the licensing of rights is enabled in a streamlined way via ex ante, categorical royalty rates of one kind or another.
Proposals along these lines already exist in the literature of course, in varying degrees of ambition. What the present analysis adds are two not insignificant points. First, from a normative point of view, it shows how equitable concerns of distribution and aggregation further push for considering alternatives to traditional copyright, for how to secure copyright’s aim of robust cultural production while posing less of a threat to our other values in a world of radically distributed technologies of high-fidelity reproduction and easy circulation. If traditional copyright has become unsustainable without inflicting severe inequities, that is all the more reason to reconsider it. Second, from an institutional perspective, the analysis discloses the point that such alternatives may not be as far a step as often thought: to a statutory regime that already often replaces, in cases of evidentiary difficulties, compensatory damages with a statutory schedule, it proposes replacing, in cases of enforcement difficulties, ex post damages with ex ante royalties.
Copyright’s scheme of statutory damages currently suffers from two fundamental defects. The first is a failure of the courts to specify the point of such damages with any care and then follow through on that purpose with any rigor. Instead, a cacophony of confused and conflicting rationales provides cover for a doctrine in messy disarray, one neither substantively well-founded nor in accord with minimal commitments of fair process. To remedy this defect, at a minimum the courts should settle on the one rationale in their midst that has some plausibility, namely the use of statutory schedules of compensation to overcome evidentiary hurdles in establishing actual harm. Doing so properly, however, would likely require legislative reform to the statutory scheme so as to abolish the use of culpability levels, which are irrelevant from a compensatory point of view and only threaten to subvert that purpose with the fog of (unjustified) punitive rationales.
The second fundamental defect of the scheme is that the rationale often adduced by scholars, that of imposing supracompensatory damages so as to achieve “optimal” deterrence in the face of significant enforcement costs, suffers from two serious flaws traceable to cracks in its economic and welfarist foundations. Imposing multiples of compensatory damages on some infringers for the sake of deterring undetected others threatens to incur severe inequitable harms in how the burdens and benefits of the copyright system (and its enforcement) are distributed and spread. To avoid such harms we need at a minimum to delink even more firmly the pursuit of “optimal” deterrence from any aura of punitive rationales, lest these provide illicit cover for an aim that courts have been shy to explicitly embrace. Moreover, pursuit of this aim needs either to be restrained by judicial safeguards to ensure it does not incur serious equitable harms or, if such safeguards are not administrable, then its pursuit must be simply called off.
Finally, if it turns out that reigning in or calling off “optimal” deterrence means that we are no longer able, in a digital world of leaky enforcement, to secure copyright’s traditional benefits, then equity requires that we reconsider the model of traditional copyright in an online world. If the traditional property rule model—rights and remedies enforced by threat of litigation—no longer works without incurring unacceptable equitable harms, then it may be time to shift toward a liability rule system: licenses and royalties enabled by streamlined protection.
This leads to an important final realization. Although we have argued that copyright’s scheme of statutory damages is, in its current form, indefensible, there may be reasons to believe that, in revised form, a statutory scheme for copyright enforcement is indispensable. That revision would involve two changes: (1) substituting, for evidentiary purposes, statutory schedules of compensation in lieu of showing actual harm in litigation; and (2) substituting, for enforcement purposes, statutory schedules of royalties in lieu of litigating infringement in the first place.
- .Robert G. Bone, A New Look at Trade Secret Law: Doctrine in Search of Justification, 86 Calif. L. Rev. 241 (1998). Professor Bone’s searching examination of trade secret law provides a model of unflinching critical analysis that we aspire to here. See also Robert G. Bone, The (Still) Shaky Foundations of Trade Secret Law, 92 Texas L. Rev. 1803 (2014). ↑
- .17 U.S.C. § 504(c) (2018). ↑
- .Id. ↑
- .See infra text accompanying notes 56–62. ↑
- .See Capitol Records, Inc. v. Thomas-Rasset, 692 F.3d 899, 901 (8th Cir. 2012) (noting a jury award of $1.5 million for sharing twenty-four copyrighted sound recordings online). ↑
- .Pamela Samuelson et al., Statutory Damages: A Rarity in Copyright Laws Internationally, but for How Long?, 60 J. Copyright Soc’y U.S.A. 529, 530 (2013) (observing that “[v]irtually all of the law review literature in the United States has criticized the U.S. statutory damage regime”); see also Ben Depoorter, Copyright Enforcement in the Digital Age: When the Remedy Is the Wrong, 66 UCLA L. Rev. 400, 404–05 (2019) (describing how minimum statutory damages can result in large fines “due to the sheer amount of infringed works”); Peter S. Menell, This American Copyright Life: Reflections on Re-Equilibrating Copyright for the Internet Age, 61 J. Copyright Soc’y U.S.A. 235, 306–07, 315 (2014) (proposing reforms to this statutory damage regime); Pamela Samuelson & Tara Wheatland, Statutory Damages in Copyright Law: A Remedy in Need of Reform, 51 Wm. & Mary L. Rev. 439, 443 (2009) (arguing that statutory damage awards result in a potentially “significant” chilling effect); J. Cam Barker, Note, Grossly Excessive Penalties in the Battle Against Illegal File-Sharing: The Troubling Effects of Aggregating Minimum Statutory Damages for Copyright Infringement, 83 Texas L. Rev. 525, 526 (2004) (positing that minimum statutory damages, in the aggregate, impose an “unconstitutional grossly excessive penalty”). ↑
- .See Samuelson & Wheatland, supra note 6, at 443 (arguing that grossly excessive penalties could potentially chill access to expressive materials). ↑
- .Stephanie Berg, Remedying the Statutory Damages Remedy for Secondary Copyright Infringement Liability: Balancing Copyright and Innovation in the Digital Age, 56 J. Copyright Soc’y U.S.A. 265, 267 (2009). ↑
- .See infra subpart III(A). ↑
- .This is a rough simplification, abstracting from any risk preferences of would-be violators. ↑
- .See Roger D. Blair & Thomas F. Cotter, An Economic Analysis of Damages Rules in Intellectual Property Law, 39 Wm. & Mary L. Rev. 1585, 1657 (1998) (hypothesizing that the threat of statutory damages, “which may be many times greater than the actual harm or benefit” a defendant derives, may be sufficient to deter potential infringers). ↑
- .See Depoorter, supra note 6, at 415–16 (“The shadow of [potential liability] may loom large over disputes and prompt settlement concessions by risk-averse defendants, capitalized upon by copyright trolls and other new enforcement business models in recent years.”). ↑
- .To be sure, in theory welfarism is not committed to any maximizing or efficiency criterion—whether of wealth or welfare—as its ultimate aim and can factor in questions of distribution. See Louis Kaplow & Steven Shavell, Fairness Versus Welfare 16 (2002) (observing that welfarism considers distribution because it is “relevant to individuals’ well-being”). But in practice, the dominant mode of applied welfarism in U.S. law and policy—economic analysis of law—does indeed focus exclusively on efficiency in law and regulatory policy, putting aside questions of distribution as apt only for legislative tax-and-transfer. See Louis Kaplow & Steven Shavell, Why the Legal System Is Less Efficient than the Income Tax in Redistributing Income, 23 J. Legal Stud. 667, 667–68 (1994) (arguing that legal rules should not be designed to sacrifice efficiency for distributive aims because the tax system can achieve redistribution more efficiently). We believe this move is unwarranted in the present context. See infra note 109 and accompanying text. ↑
- .See Ronald Dworkin, Taking Rights Seriously xi (1978) (positing that individual rights serve as “political trumps”). ↑
- .Copyright Act of May 31, 1790, ch. 15, § 2, 1 Stat. 124, 124–25 (1790). ↑
- .Statute of Anne, 8 Ann., c. 19, § 2 (1710). ↑
- .See Oren Bracha, The Adventures of the Statute of Anne in the Land of Unlimited Possibilities: The Life of a Legal Transplant, 25 Berkeley Tech. L.J. 1427, 1448–49 (2010) (discussing the influence of the Statute of Anne on state copyright statutes). ↑
- .See Oren Bracha, Owning Ideas: The Intellectual Origins of American Intellectual Property, 1790–1909 59–60 (2016) (observing that the 1790 Act followed the Statute of Anne in providing only “forfeiture and destruction of infringing copies” and “a fine of fifty cents per each sheet” in the infringer’s possession). ↑
- .Copyright Act of 1909, Pub. L. No. 60-349, § 25(b), 35 Stat. 1075, 1081 (1909). ↑
- .Copyright Act of 1976, Pub. L. No. 94-553, 90 Stat. 2541, 2585–86 (1976) (codified as amended at 17 U.S.C. §504(c)). ↑
- .See Berne Convention Implementation Act of 1988, Pub. L. No. 100-568, § 21, 102 Stat. 2853, 2860 (1988) (enhancing statutory damages for copyright infringement); Digital Theft Deterrence and Copyright Damages Improvement Act of 1999, Pub. L. No. 106-160, § 2, 113 Stat. 1774, 1774 (1999) (codified as amended at 17 U.S.C. § 504(c) (2000)) (enhancing statutory damages for copyright infringement actions). ↑
- .17 U.S.C. § 504(c)(1) (2018). ↑
- .Id. ↑
- .Id. ↑
- .There is also a fourth, narrower category under which certain institutional infringers are exempt altogether from statutory damages (but not other remedies), provided that such an infringer “believed and had reasonable grounds for believing that his or her use of the copyrighted work was a fair use.” Id. § 504(c)(2). ↑
- .Id. § 504(c)(1). ↑
- .Id. § 504(c)(2). ↑
- .Id. ↑
- .See 4 Melville B. Nimmer & David Nimmer, Nimmer on Copyright §§ 14.04[B][a], [B][a] (2019) (describing disagreement among courts regarding the standards for innocent infringement and willful infringement, respectively). ↑
- .Erickson Prods. Inc. v. Kast, 921 F.3d 822, 833 (9th Cir. 2019) (ruling that willfulness requires “actual knowledge, willful blindness, or recklessness”). ↑
- .Childress v. Taylor, 798 F. Supp. 981, 994 (S.D.N.Y. 1992) (observing that “an infringer must meet a two-pronged burden of proof to establish innocence: (1) a subjective good faith . . . which was (2) objectively reasonable under the circumstances”). ↑
- .More precisely, within the default category, negligence is simply assumed unless the lower or higher degrees of fault are established. For the analysis of copyright infringement using tort concepts of fault, see Oren Bracha & Patrick Goold, Copyright Accidents, 96 B.U. L. Rev. 1025, 1039–40 (2016). ↑
- .Feltner v. Columbia Pictures Television, Inc., 523 U.S. 340, 355 (1998) (ruling that “the Seventh Amendment provides a right to a jury trial on all issues pertinent to an award of statutory damages under § 504(c) of the Copyright Act”). ↑
- .See 6 William F. Patry, Patry on Copyright § 22:172 (2020) (discussing election-of-damages procedures). ↑
- .See Sony BMG Music Entm’t v. Tenenbaum, 660 F.3d 487, 504 (1st Cir. 2011) (observing with approval that “[i]t is commonplace for courts to explicitly instruct juries of the maximum and minimum statutory damage awards permitted under § 504(c)”). ↑
- .See Patry, supra note 34, § 22:173 (discussing the exact standards applied by the various circuit courts to trial courts’ awards of statutory damages). ↑
- .See Harris v. Emus Records Corp., 734 F.2d 1329, 1335 (9th Cir. 1984) (“[T]he court has wide discretion in determining the amount of statutory damages to be awarded, constrained only by the specified maxima and minima.”); Nimmer & Nimmer, supra note 29, § 14.04[B][a] (observing that “as long as the district court acts within the prescribed statutory limits, its discretion will probably be upheld on appeal”). ↑
- .17 U.S.C. § 504(c)(1) (2018). ↑
- .See, e.g., Smith v. Thomas, 911 F.3d 378, 382 (6th Cir. 2018) (ruling that a plaintiff “simply has to inform the court—either orally or in writing—of his intent to seek [statutory damages]”). ↑
- .E.g., id. (“[P]laintiffs are entitled to simultaneously seek actual damages and statutory damages in the alternative.”); Curet–Velazquez v. ACEMLA de P.R., Inc., 656 F.3d 47, 57–58 (1st Cir. 2011) (citing Fed. R. Civ. P. 8(a)(3) for the proposition that plaintiff “may include relief in the alternative or different types of relief”). ↑
- .See Patry, supra note 34 (discussing procedures for requesting alternative awards and choosing in between them). ↑
- .See also H.R. Rep. No. 94-1476, at 161 (1975) (“[T]here is nothing in section 504 to prevent a court from taking account of evidence concerning actual damages and profits in making an award of statutory damages . . . .”). ↑
- .2 Paul Goldstein, Goldstein on Copyright § 126.96.36.199, at 14.45 (3d ed. 2005); Depoorter, supra note 6, at 435. ↑
- .RSO Records, Inc. v. Peri, 596 F. Supp. 849, 862 (S.D.N.Y. 1984); Patry, supra note 34, § 22:174. Some courts limit the guiding role of actual damages to cases on nonwillful infringement. See M.S.R. Imports, Inc. v. R.E. Greenspan Co., 596 F. Supp. 849, 862 (E.D. Pa. 1983) (asserting that “in the absence of a willful violation,” a statutory damages award should “attempt to approximate the normal ‘measure of damages’” approach). ↑
- .L.A. News Serv. v. Reuters Television Int’l, Ltd., 149 F.3d 987, 996 (9th Cir. 1998). ↑
- .Shapkin/Crossroads Prods., Inc. v. Legacy Home Video, Inc., Nos. 96-55650, 96-55761, 1997 WL 556312, at *2 (9th Cir. Aug. 29, 1997) (“[T]he district court has the power to discourage infringement by imposing a statutory award even where actual damages are negligible.”). ↑
- .See Samuelson & Wheatland, supra note 6, at 462–63 (providing examples of statutory damages awards disproportionate to the actual harm suffered). ↑
- .17 U.S.C. § 504(c)(1) (2018). ↑
- .Id. § 504(c)(2). ↑
- .See S. Rep. No. 94-473, at 144–45 (1975) (describing the increased maximum and decreased minimum awards in section 504(c)(2) as providing for “exceptional cases”); H.R. Rep. No. 94-1476, at 162 (1976) (same). ↑
- .Depoorter, supra note 6, at 439 (finding that “[a]lthough 80 percent of plaintiffs in all disputes claim they suffered conduct that constitutes willful infringement, courts only consider enhanced damages to be justified in just 2 percent of cases where a plaintiff wins the case”). ↑
- .Id. at 438 (finding in the case law “considerable ambiguity as it relates to potential findings of willfulness and the enhanced awards that may follow” and concluding that “[t]he flexible standard of ‘knowledge’ that courts employ . . . open[s] the door to the potential award of enhanced damages in many disputes”). ↑
- .Id. at 440. ↑
- .See, e.g., Childress v. Taylor, 798 F. Supp. 981, 994 (S.D.N.Y. 1992) (“[A]n infringer must meet a two-pronged burden of proof to establish innocence: (1) a subjective good faith belief in that state of grace which was (2) objectively reasonable under the circumstances.”); Little Mole Music v. Spike Inv., Inc., 720 F. Supp. 751, 755 (W.D. Mo. 1989) (“[M]ere non-deliberate infringement is not ‘innocent’; rather, defendants must have ‘acted in complete ignorance of the fact that [their] conduct might somehow infringe upon the rights of another party.’”); see also Samuelson & Wheatland, supra note 6, at 460 (“[T]he innocent infringer provision is essentially never used.”). ↑
- .Samuelson & Wheatland, supra note 6, at 481–83. ↑
- .E. Am. Trio Prods., Inc. v. Tang Elec. Corp., 97 F. Supp. 2d 395, 419 (S.D.N.Y. 2000). ↑
- .Fitzgerald Publ’g Co. v. Baylor Publ’g Co., 807 F.2d 1110, 1116 (2d Cir. 1986); see also Cullum v. Diamond A Hunting, Inc., 484 Fed. App’x. 1000, 1002 (5th Cir. 2012) (“Between the statutory maximum and minimum, however, the court has virtually unfettered discretion in deciding the quantum of damages to award in a copyright infringement case.”). ↑
- .UMG Recordings, Inc. v. MP3.Com, Inc., No. 00 Civ. 472(JSR), 2000 WL 1262568, at *5 (S.D.N.Y. Sept. 6, 2000). ↑
- .Patry, supra note 34, § 22:174. ↑
- .See Nat’l Football League v. PrimeTime 24 Joint Venture, 131 F. Supp. 2d 458, 473–74 (S.D.N.Y. 2001) (reviewing the factors employed by courts in determining statutory damages awards). ↑
- .See, e.g., Schiffer Publ’g, Ltd. v. Chronicle Books, LLC, No. Civ.A.03–4962, 2005 WL 67077, at *5 (E.D. Pa. Jan. 11, 2005) (“[D]efendant’s conduct is the most important factor.”); Prime Time 24 Joint Venture, 131 F. Supp. 2d at 474 (describing the defendant’s state of mind as “a key factor”). ↑
- .Patry, supra note 34, § 22:174. ↑
- .Nimmer & Nimmer, supra note 29, § 14.04[B][a]. ↑
- .Samuelson & Wheatland, supra note 6, at 459. ↑
- .Id. at 441. ↑
- .See, e.g., Sony BMG Music Entm’t v. Tenenbaum, 660 F.3d 487, 515 (1st Cir. 2011) (reversing the district court holding that a jury award of $675,000 for the infringement of copyright in thirty sound recordings violated due process); Macklin v. Mueck, No. 00-14092-CIV-MOORE, 2005 U.S. Dist. LEXIS 18026, at *3 (S.D. Fla. Jan. 28, 2005) (awarding the statutory maximum of a total of $300,000 for the infringement of copyright in two poems); L.A. Times v. Free Republic, No. 98–7840 MMM AJWX, 2000 WL 1863566, at *3 (C.D. Cal. Nov. 16, 2000) (awarding $1,000,000 for infringement by numerous postings on a website by its users of news articles accompanied by commentary). In one case, there were three jury awards for the infringement of copyright in twenty-four sound recordings: $222,000, $1,920,000, and $1,500,000. Capitol Records, Inc. v. Thomas-Rasset, 692 F.3d 899, 901–02 (8th Cir. 2012). On appeal, after the district court found the third amount in violation of the due process clause of the Fifth Amendment, plaintiffs agreed to limit their award to the original $220,000, an amount which the appellate court reinstated. Id. at 902. ↑
- .See Douglas v. Cunningham, 294 U.S. 207, 209 (1935) (observing that the purpose of statutory damages under the 1909 Act is “to give the owner of a copyright some recompense for injury done him, in a case where the rules of law render difficult or impossible proof of damages or discovery of profits”); L.A. Westermann Co. v. Dispatch Printing Co., 249 U.S. 100, 107–08 (1919) (citing Brady v. Daly, 175 U.S. 148, 154, 157 (1899), with respect to the 1909 Act); Brady, 175 U.S. at 157 (referring to “the inherent difficulty of always proving by satisfactory evidence” actual damage as the underlying reason for statutory damages for the infringement of copyright in dramatic works under the law in force at the time); Peter Pan Fabrics, Inc. v. Jobela Fabrics, Inc., 329 F.2d 194, 195–96 (2d Cir. 1964) (explaining that the statutory damages provision of the 1909 Act “allows ‘the owner of a copyright some recompense for injury done him, in a case where the rules of law render difficult or impossible proof of damages or discovery of profits’”). ↑
- .See Staff of H. Comm. on the Judiciary, 87th Cong., Report of the Register of Copyrights on the General Revision of the U.S. Copyright Law 102 (Comm. Print 1961) [hereinafter Report of the Register of Copyrights] (explaining that the main reason for copyright’s statutory damages is that “[t]he value of a copyright is, by its nature, difficult to establish, and the loss caused by an infringement is equally hard to determine”). ↑
- .See, e.g., Cable/Home Commc’n Corp. v. Network Prods., Inc., 902 F.2d 829, 850 (11th Cir. 1990) (noting that statutory damages are awarded when actual damages are difficult, if not impossible, to prove or calculate); Warner Bros. Inc. v. Dae Rim Trading, Inc., 877 F.2d 1120, 1126 (2d Cir. 1989) (same); Van Der Zee v. Greenidge, No. 03 Civ. 8659(RLE), 2006 WL 44020, at *1 (S.D.N.Y. Jan. 6, 2006) (same); Latin Am. Music, Inc. v. Spanish Broad. Sys., Inc., 866 F. Supp. 780, 782 (S.D.N.Y. 1994) (same); In re Braun, 327 B.R. 447, 450 (Bankr. N.D. Cal. 2005) (same). ↑
- .See, e.g., Report of the Register of Copyrights, supra note 68, at 103 (the purpose of statutory damages is to “assure adequate compensation to the copyright owner for his injury, and . . . to deter infringement”); Cable/Home Commc’n Corp., 902 F.2d at 851 (citing compensation and deterrence); Frank Music Corp. v. Metro-Goldwyn-Mayer, Inc., 886 F.2d 1545, 1554 (9th Cir. 1989) (same); Van Der Zee, 2006 WL 44020, at *3 (same); Granada Sales Corp. v. Aumer, No. 02 Civ. 6682(HB)(RLE), 2003 WL 21383821, at *2 (S.D.N.Y. June 2, 2003) (same); Stevens v. Aeonian Press, Inc., No. 00 Civ. 6330(JSM), 2002 WL 31387224, at *4 (S.D.N.Y. Oct. 3, 2002) (same). ↑
- .Some language in the legislative report may be read as implying that no effort to establish actual damages or the difficulty of proving them is a precondition for statutory damages eligibility. See H.R. Rep. No. 94-1476, at 161 (1976) (“[T]he plaintiff in an infringement suit is not obliged to submit proof of damages and profits and may choose to rely on the provision for minimum statutory damages.”). ↑
- .See, e.g., TD Bank N.A. v. Hill, 928 F.3d 259, 282 (3d Cir. 2019) (noting that “statutory damages may be imposed even if the accused infringer reaps nothing from infringement”); Cable/Home Commc’n Corp., 902 F.2d at 850 (observing that a plaintiff may elect statutory damages “whether or not adequate evidence exists as to the actual damages”); Harris v. Emus Records Corp., 734 F.2d 1329, 1335 (9th Cir. 1984) (“Statutory damages may be elected whether or not there is adequate evidence of the actual damages suffered by plaintiff or of the profits reaped by defendant.”). ↑
- .Goldstein, supra note 43, § 188.8.131.52, at 14.46. ↑
- .See Peer Int’l Corp. v. Pausa Records, Inc., 909 F.2d 1332, 1337 (9th Cir. 1990) (holding that the lower court did not abuse its discretion in awarding the statutory maximum). ↑
- .See Goldstein, supra note 43, § 184.108.40.206, at 14.45–14.47. ↑
- .See Samuelson & Wheatland, supra note 6, at 462–63, 487–89 (discussing cases in which courts awarded large multiples of estimated harm to injured parties). ↑
- .See supra text accompanying notes 38–41. ↑
- .There is no reason why the innocent infringer cannot inflict a much higher damage than the willful one. ↑
- .See, e.g., L.A. News Serv. v. Reuters Television Int’l, Ltd., 149 F.3d 987, 996 (9th Cir. 1998) (asserting that “statutory damages serve both compensatory and punitive purposes”); Cass Cty. Music Co. v. C.H.L.R., Inc., 88 F.3d 635, 643 (8th Cir. 1996) (referring to the “punitive sanction on infringers” role of statutory damages); Illinois Bell Tel. Co. v. Haines & Co., 905 F.2d 1081, 1089 (7th Cir. 1990) (noting that “[a] damage award greater than profits . . . [puts] potential infringers on notice” that compliance with copyright laws will cost less than violations). Some but not all decisions limit the punitive purpose to cases of willful infringement. See, e.g., On Davis v. Gap, Inc., 246 F.3d 152, 172 (2d Cir. 2001) (observing that “[t]he purpose of punitive damages—to punish and prevent malicious conduct—is generally achieved under the Copyright Act through the provisions of 17 U.S.C. § 504(c)(2), which allow increases to an award of statutory damages in cases of willful infringement”). ↑
- .See infra text accompanying notes 86–87. ↑
- .For two leading views of retributive justice, see Michael Moore, Placing Blame: A Theory of Criminal Law (1997) and Jeffrie G. Murphy, Legal Moralism and Retribution Revisited, 1 Crim. L. & Phil. 5 (2007). ↑
- .See Moore, supra note 81, at 55–58, 249 (discussing theories of “intentionality” in tort and criminal law, their associated culpability levels, and how culpability contributes to “responsibility” and fault). ↑
- .17 U.S.C. § 506 (2018). ↑
- .Note that the argument offered here is structural: its logic is that the inclusion of specific criminal penalties makes a retributive purpose for civil statutory damages implausible. This is in contrast to others who made textual arguments both for and against a punitive purpose based on the fact that the 1909 Act provided that statutory damages “shall not be regarded as a penalty,” language that was not included in the 1976 Act. See H. Tomás Gómez-Arostegui, What History Teaches Us About US Copyright Law and Statutory Damages, 5 WIPO J. 76, 79–86 (2013) (describing and criticizing such arguments). ↑
- .See Nimmer & Nimmer, supra note 29, § 15.01[A] (observing the different standards of proof in civil and criminal copyright actions). To be clear, we neither advocate retributive criminal sanctions in copyright nor argue that criminal procedural safeguards are always effective in this area. Our view is that retribution should not be pursued in copyright at all. The present argument is simply interpretive, namely: that it is highly implausible to read the statute as creating specific criminal punishments while also inflicting retributive sanctions through civil remedies where no procedural safeguards apply. ↑
- .See, e.g., Music City Music v. Alfa Foods, Ltd., 616 F. Supp. 1001, 1003 (E.D. Va. 1985) (referring to a need to “punish the wrongdoer”). ↑
- .See, e.g., Yellow Pages Photos, Inc. v. Ziplocal, LP, 795 F.3d 1255, 1271 (11th Cir. 2015) (“The Copyright Act’s statutory damages provision is designed to discourage wrongful conduct.”); Int’l Korwin Corp. v. Kowalczyk, 855 F.2d 375, 383 (7th Cir. 1988) (“When, as here, the violation is willful, deterrence of future violations is a legitimate consideration.”); Music City Music, 616 F. Supp. at 1004 (referencing “the need to deter [the] defendant from choosing to violate copyright laws in the future”). Reference to a deterrence purpose also found its way from the case law to the title and the legislative history of the 1999 amendment to the statutory damages provision. Digital Theft Deterrence and Copyright Damages Improvement Act of 1999, Pub. L. No. 106-160, 113 Stat. 1774 (codified as amended at 17 U.S.C. § 504(c) (2000)) (referencing deterrence in the Act’s title); H.R. Rep. No. 106-216, at 2 (1999) (“The purpose of H.R. 1761 is to provide more stringent deterrents to copyright infringement and stronger enforcement of the laws enacted to protect intellectual property rights.”). ↑
- .See, e.g., Unicity Music, Inc. v. Omni Commc’ns, Inc., 844 F. Supp. 504, 510 (E.D. Ark. 1994) (explaining that statutory damages are intended “both to compensate plaintiffs and deter defendants from future infringing conduct by making it clear that infringement is significantly more expensive than paying the licensing fees”); Rodgers v. Eighty Four Lumber Co., 623 F. Supp. 889, 892 (W.D. Pa. 1985) (quoting Music City Music, 616 F. Supp. at 1003) (observing that “statutory damages should exceed the unpaid license fees ‘so that defendant will be put on notice that it costs less to obey the copyright laws than to violate them’”). ↑
- .See, e.g., Comm. on Pattern Jury Instructions, Dist. Judges Ass’n, Eleventh Cir., Pattern Jury Instructions, Civil Cases § 9.32 (2018) (describing the purposes of statutory damages as follows: “compensate the copyright owner, penalize the infringer, and deter future copyright law violations”); Ninth Cir. Jury Instruction Comm., Manual of Model Civil Jury Instructions for the District Courts of the 9th Cir. § 17.35 (2017) (describing the purpose of statutory damages as “not only to compensate the plaintiff . . . but also to penalize the infringer and deter future violations of the copyright laws”). ↑
- .A perfect example of this seamless transition is Palmer v. Slaughter, No. 99-899-GMS, 2000 U.S. Dist. LEXIS 22118, at *9 (D. Del. July 13, 2000), which explains that statutory damages serve “as a deterrent by punishing the defendant for its unlawful conduct.” Id. ↑
- .Or the opportunity cost of not engaging in the activity at all. ↑
- .In other words, the familiar Learned Hand formula: B < PL. ↑
- .See Bracha & Goold, supra note 32, at 1043 (explaining the cost–benefit analysis of the negligence rule). ↑
- .This of course still leaves open the possibility of awarding damages on grounds other than deterring defendants, including simply compensating plaintiffs as merited for its own sake—but this would of course simply take us back to the evidentiary rationale. ↑
- .A. Mitchell Polinsky & Steven Shavell, Punitive Damages: An Economic Analysis, 111 Harv. L. Rev. 869, 905–06 (1998) (explaining that “reprehensibility per se” should generally not affect the imposition of increased damages, if the aim is deterrence). ↑
- .Id. at 879. ↑
- .See Depoorter, supra note 6, at 435–36 (observing that courts ignore the economic rationale of supracompensatory damages based on enforcement cost). Courts do sometimes refer to the need to deter third parties, but do not develop the full optimal deterrence argument based on enforcement cost. See, e.g., Sony BMG Music Entm’t v. Tenenbaum, 660 F.3d 487, 504 (1st Cir. 2011) (approving a jury instruction that included “the need to deter this defendant and other potential infringers”). But see Patry, supra note 34, at § 22:183 (observing that “[d]eterrence is only individual”). ↑
- .See Blair & Cotter, supra note 11, at 1693 (indicating that “due to the presence of enforcement costs, it may be optimal in some cases to enhance the plaintiff’s damages by a multiplier”). ↑
- .See, e.g., Blair & Cotter, supra note 11, at 1656 (arguing that the cost of detecting widespread private copyright infringement is “enormous”); Christopher Buccafusco & Jonathan S. Masur, Innovation and Incarceration: An Economic Analysis of Criminal Intellectual Property Law, 87 S. Cal. L. Rev. 275, 305 (2014) (discussing the challenges of detecting copyright infringement in the digital age); Jayashri Srikantiah, Note, The Response of Copyright to the Enforcement Strain of Inexpensive Copying Technology, 71 N.Y.U. L. Rev. 1634, 1647 (1996) (“Recent advances in copying technology create two main enforcement problems: increased volume and undetectability.”). ↑
- .Polinsky & Shavell, supra note 95, at 888–89. ↑
- .Id. at 874 (“[P]unitive damages ordinarily should be awarded, if, and only if, an injurer has a chance of escaping liability . . . .”). ↑
- .Id. at 887 (observing that when an injurer has a chance of escaping liability, “the proper magnitude of damages is the harm the defendant has caused, multiplied by a factor reflecting the probability of his escaping liability”). ↑
- .Id. at 889–90. ↑
- .See, e.g., Blair & Cotter, supra note 11, at 1619 (modeling optimal deterrence of copyright infringement through economic analysis). ↑
- .This is unlike criminal sanctions, where the cost of the sanction is not a mere “transfer.” See Buccafusco & Masur, supra note 99, at 312–15 (discussing various costs of copyright criminal sanctions, but failing to mention the “cost” of the suffering imposed on those who are subjected to criminal sanctions). ↑
- .See Blair & Cotter, supra note 11 (“Perhaps only the threat of a statutory damages award, which may be many times greater than the actual harm or benefit derived from the defendant’s unauthorized use, will be sufficient to prevent the value of the owner’s copyright from being destroyed by a multitude of small-scale infringing acts.”); Buccafusco & Masur, supra note 99, at 306 (arguing that in light of copyright’s enforcement difficulties, “the damages multiplier that would be necessary to raise expected penalties from infringement above expected benefits will be very large”). ↑
- .For a general discussion of distributive justice in the context of copyright, see Oren Bracha & Talha Syed, Beyond Efficiency: Consequence-Sensitive Theories of Copyright, 29 Berkeley Tech. L.J. 229, 287–313 (2014). ↑
- .And further assume that it is impossible to rotate the pool of taxpaying individuals. ↑
- .This of course faces a standard economic objection, namely that pursuit of distributive equity within copyright incurs a “double distortion” of efficiency in comparison to the single one incurred in legislative tax and transfer. See Kaplow & Shavell, supra note 13, at 669–74 (arguing that an inefficient, redistributive legal rule leaves the government with less tax revenue than an efficient rule). Whatever the merits of this objection when applied to pursuit of distributive equity writ large within fields of private law, it is to no avail against pursuit of equity in the distribution of field-specific burdens and benefits, as we propose here, for two reasons: (1) First and most fundamentally, as set out in the text, it is not the case that there is some distinct “cost” to “efficiency” that must then be weighed against the “gain” in “equity”—rather, there is simply one integrated aim, of fairly securing the net benefits of copyright across affected individuals, leading distinct lives. (2) Second, even if one were to view the “cost” to efficiency as a separate factor, it turns out that the cost within copyright is the same as that in tax-and-transfer, namely a dampening of dynamic production incentives. For further discussion of this latter point, as well as replies to other common economic objections to the pursuit of distributive aims in private law, see Bracha & Syed, supra note 107, at 289, 311–13. ↑
- .Although many may see this as straightforward enough, others—especially those deeply versed in distributive justice theory—may be surprised at this easy assimilation of the market’s distributive norm (“from each according to effective preferences, to each according to effective preferences”) under the aegis of principles of distributive equity. Two points will hopefully alleviate any alarm on this score: (1) First, most theorists of distributive justice accept the role of a responsibility norm as relevant to fair distributions and also that markets, to a first approximation, give voice to individual responsibility, in terms of persons’ choosing which preferences to cultivate and pursue in light of their social opportunity costs. Indeed, for some theorists the role of markets in this regard make them not merely of instrumental value but inherently indispensable to defining (liberal) distributive justice by serving as the vehicle through which responsibility is factored in via opportunity cost. See Ronald Dworkin, What Is Equality? Part 2: Equality of Resources, 10 Phil. & Pub. Aff. 283, 338 (1981) (suggesting that the market enforces the “fundamental requirement” that social resources “be devoted to the lives of each of its members” as measured by the opportunity cost of those resources to others). Whether or not this stronger view is merited, the weaker position is virtually universally shared by distributive justice theorists—with the singular exception of John Rawls, who agrees in principle but demurs in practice. See Talha Syed, Educational Accommodation and Distributive Equity: The Principle of Proportionate Progress, 50 Conn. L. Rev. 485, 543–44 n.212 (2018) (reviewing the literature, including Rawls’s view and criticisms of it). (2) However, and second, most theorists also recognize that responsibility is only relevant when it is reasonable to factor in, as a matter of individuals having suitably free and fair conditions for the choices to which they are held responsible, and they also recognize that markets often do not present such conditions—i.e., that markets only model reasonable responsibility to a first approximation, and that ultimately they must be qualified and corrected to account for the role of unchosen factors in their processes and outcomes. Such corrections involve, principally, giving priority to those disadvantaged through no fault of their own, either in terms of disparities in market-based earnings or the market-based costs of meeting their needs, precisely along the lines of the priority principle discussed next in the text. ↑
- .The principle in the text is formulated broadly enough so as to encompass a variety of distinct views that agree on the basic priority commitment, but otherwise disagree, both on the underlying normative basis for according such priority and on how much priority should be accorded. See, e.g., Larry S. Temkin, Inequality 247–48, 282 (1993) (advancing the “telic equality” view that equalizing is in itself valuable and serves as the basis for giving priority to the “worse” off); Harry Frankfurt, Equality as a Moral Ideal, 98 Ethics 21, 21–22 (1987) (advancing the “sufficiency” view that equalizing is not in itself valuable and so what matters is not whether someone is comparatively “worse” off than others but only whether someone is “badly” off according to an absolute standard); Derek Parfit, Equality or Priority? The Lindley Lecture at the University of Kansas 8 (Nov. 21, 1991) (advancing the noncomparative priority view as a generalization of the sufficiency view); Syed, supra note 110, at 488 (advancing the comparative priority view that what matters is neither equalizing for its own sake nor simply sufficiency or noncomparative priority, but rather fairness, which requires that persons be given priority because they are worse off than others). ↑
- .Syed, supra note 110, at 489–90. ↑
- .See id. at 533–34 (advancing and defending this principle of distributive justice based on an underlying notion of fairness whereby priority should be given to those “who are worse off through no fault of their own because and to the extent that they are worse off”). ↑
- .See supra text accompanying notes 23–28. ↑
- .To be sure, there is a limit to how effective such ratcheting up can be, with the ceiling determined by the point at which the average caught infringer will simply become judgment-proof, and hence the average potential infringer will cease to factor in additional sums into her expected harm calculus. But the concerns we raise here are likely to be triggered well before that ceiling is reached. ↑
- .Aggregation concerns have a rich pedigree in the philosophical literature, going back at least to John Taurek’s pioneering discussion of the problem in Should the Numbers Count? John Taurek, Should the Numbers Count?, 6 Phil. & Pub. Aff. 293 (1977). For additional landmarks, see Tim Scanlon, What We Owe to Each Other (1998) and Frances Kamm, Intricate Ethics (2007). The views advanced in these works, however, tend toward classically deontological ones, tracing back to Kant. Our argument here bears greater affinity with two related but distinct lines of “incommensurability” analysis, one descending from Mill, the other from Marx. The line from Mill focuses on “higher order” interests and arrives at the destination of Rawlsian “lexical priority.” See John Stuart Mill, Utilitarianism 12–16 (1863) (articulating theory of higher order pleasures and the integral role of ordering human interests when determining a utility value); John Rawls, A Theory of Justice 42–43 (1971) (describing the concept of a “lexical order”). The line from Marx focuses on “qualitatively distinct” interests and arrives at the destination of irreducibly plural capabilities per Sen and Nussbaum. See Karl Marx, Economic and Philosophic Manuscripts of 1844: Selections, in The Marx-Engels Readers 66, 101–05 (Robert C. Tucker ed., 2nd. ed. 1978) (arguing that money, the existing and active concept of value in bourgeois society, confounds and compounds all natural and human qualities that possess distinctive character); Karl Marx, On James Mill, in Karl Marx: Selected Writings 124, 125, 127–28, 130–32 (David McLellan ed., 2d ed. 2000) (noting that producers have distinct objectives and relationships with each other and with their products); Karl Marx & Friedrich Engels, The German Ideology, in Karl Marx: Selected Writings 175, 201–05 (David McLellan ed., 2d ed. 2000) (outlining the development of the theory of utility and exploitation and its emphasis on the aggregation of individual activity); Martha C. Nussbaum, Flawed Foundations: The Philosophical Critique of (a Particular Type of) Economics, 64 U. Chi. L. Rev. 1197, 1199, 1202–03 (1997) (exploring the concept of qualitatively distinct interests); Amartya K. Sen, Plural Utility, 81 Proc. Aristotelian Soc’y 193, 193–94 (1981) (arguing for a “vector view” of utility while acknowledging “the existence of many co-existing aspects of utility”). Our particular version of the argument, it bears noting, diverges from both the absolute or lexical priority accorded by Rawls and the radical incommensurability adopted by Sen and Nussbaum. Earlier deployments of this argument in distinct innovation policy contexts can be found elsewhere. See Bracha & Syed, supra note 107, at 284 n.183 (2014) (protecting from single-scale aggregation, fundamental interests in self-identification and self-expression in the context of copyright policy); William W. Fisher III & Talha Syed, A Prize System as a Partial Solution to the Health Crisis in the Developing World, in Incentives for Global Public Health: Patent Law and Access to Essential Medicines 181, 201–02 (Thomas Pogge et al. eds., 2010) (protecting from single-scale aggregation, the fundamental interests in health of those suffering from rare diseases in the context of prizes for drug research). ↑
- .See Capitol Records, Inc. v. Thomas-Rasset, 692 F.3d 899, 901 (8th Cir. 2012) (noting that a jury awarded statutory damages of $1,920,000 against a private individual); Sony BMG Music Entm’t v. Tenenbaum, 660 F.3d 487, 490 (1st Cir. 2011) (noting that a jury awarded statutory damages of $675,000 against a private individual). ↑
- .That the courts’ current manner of administering statutory damages is an affront to norms of procedural equity is of course not a point original to us. What may be somewhat more distinctive, however, is our diagnosis of the source of the problem as lying in a lack of clarity about the substantive aim(s) that should guide us in this area, rather than simply in a lack of determinate “guidelines” or “factors” untethered from any underlying purpose—and hence our insistence that the proper remedy lies with the better articulation and adherence to a defensible rationale, rather than recourse to “bright-line” rules or any other salve for easy administrability. ↑
- .H.R. Rep. No. 94-1476, at 161 (1975). ↑
- .See also Depoorter, supra note 6, at 445–46 (proposing to amend the statute to allow defendants to submit credible evidence of damages or profit that would cap statutory damages). ↑
- .This option is somewhat in line with other proposals in the literature that have as a main element the restriction of statutory damages, but it departs somewhat in its insistence on eliminating the culpability categories. Cf. Depoorter, supra note 6, at 443–44 (proposing to reduce the statutory ranges, reduce “outlier awards,” and perhaps eliminate “enhanced” statutory damages); Alan Garfield, Calibrating Copyright Statutory Damages to Promote Speech, 38 Fla. St. U. L. Rev. 1, 43 (2010) (proposing that the default rule for statutory damages should be “merely compensatory and restitutionary damages” and enhanced damages should be awarded “only in those situations in which the deterrent value of punitive damages is clearly appropriate”); Samuelson & Wheatland, supra note 6, at 509 (proposing to break down statutory damages into two provisions: one limited to estimated compensation and the other devoted to enhanced damages in cases of “egregious” infringement). ↑
- .See, e.g., Srikantiah, supra note 99, at 1647 (describing the “main enforcement problems” of the digital age as increased volume and undetectability). ↑
- .See id. (“The technology of the Internet allows users to download and distribute copies of software by pressing a few keys.”). ↑
- .See Guido Calabresi & A. Douglas Melamed, Property Rules, Liability Rules, and Inalienability: One View of the Cathedral, 85 Harv. L. Rev. 1089, 1092 (1972) (theorizing the concepts of property and liability rules). ↑
- .See, e.g., William W. Fisher III, Promises to Keep: Technology, Law, and the Future of Entertainment 202 (2004) (proposing a government-administered compensation system to replace the current copyright system); Neil Weinstock Netanel, Impose a Noncommercial Use Levy to Allow Free Peer-to-Peer File Sharing, 17 Harv. J. L. & Tech. 1, 4 (2003) (proposing a levy on noncommercial peer-to-peer media sharing). ↑