An economic analysis of damages rules in intellectual property law.

• Author: Blair, Roger D.

1. INTRODUCTION

   In 1987, an architect named Lebbeus Woods created a graphite pencil drawing entitled "Neomechanical Tower (Upper) Chamber," which depicts an ominous-looking chamber containing a chair mounted high upon one wall by means of a vertical rail, and a sphere suspended in front of the chair at face level.(1) Eight years later, Universal City Studios, Inc. ("Universal") released the film 12 Monkeys, a dystopian adventure involving a future world in which human beings have taken to living underground following an outbreak of a mysterious virus that kills five billion people beginning in the year 1996.(2) Near the beginning of the film, the rulers of this world bring the lead character, played by Bruce Willis,

   into a room where he is told to sit in a chair which is attached

   to a vertical rail on a wall. The chair slides up the rail

   to a horizontal ledge on the wall so that the chair is several

   yards above the ground. A sphere supported by a metal-frame

   armature descending from above is suspended directly in

   front of (Willis]. On three occasions, [Willis] returns to this

   chair.(3)

   A few weeks after the film was released, Woods filed suit in the United States District Court for the Southern District of New York, alleging that Universal and the film's director and production designer had infringed his copyright in the drawing by reproducing the work in the scene described above (albeit in another medium, film), and by distributing and exhibiting the reproduction to the public without permission.(4) Concluding that the defendants had access to Woods's work and that the scene was strikingly similar to that work, the court granted Woods's motion for a preliminary injunction.(5) Shortly thereafter, the parties agreed to a settlement pursuant to which Woods agreed to allow the continued exhibition of the film in return for an undisclosed sum of money.(6)

   Had the case proceeded all the way to trial and resulted in a judgment in Woods's favor, Woods presumably would have been entitled to a permanent injunction against the exhibition of the film,(7) and also to monetary damages for the defendants' unauthorized use of the drawing.(8) But what would those damages be? In answering this question, a policymaker with the authority to award whatever damages she believes to be appropriate would face several different, and difficult, choices; and although this example involves the law of copyright, one can easily imagine similar examples involving other types of intellectual property such as patents, trade secrets, and trademarks. One option would be to award Woods his lost profits from the sale of authorized copies of his drawing;(9) on the facts described above, however, this amount probably would be zero, because it is unlikely that the film would be a close enough substitute for, say, a book of authorized Woods reproductions so as to displace sales of the latter.(10) A second option would be to award Woods a royalty for the unauthorized use of his drawing, but this option raises the question of how to determine the amount of the royalty. Should the policymaker simply award the amount she believes to be "just"? Should she refer the decision to some expert tribunal? Or should she try to reconstruct the amount the parties themselves would have agreed upon ex ante (and if so, how)?(11) A third option would be to forgo compensation altogether and to order the defendants instead to disgorge their profit from the exhibition of the film--though surely not their entire profit, but only that portion that is attributable to the use of Woods's drawing (and how, exactly, would you go about measuring that)? Perhaps one could sympathize with a policymaker who, upon consideration of the above choices, simply throws up her hands and awards some arbitrary amount--as, arguably, the concept of presumed damages allows the trier of fact to do in some types of defamation cases.(12)

   This Article focuses on the use of economic analysis to assist our hypothetical policymaker, confronted as she is with these various and conflicting options, in crafting the optimal set of damages rules for use in intellectual property litigation. Surprisingly, this issue has remained largely unexplored in the law and economics literature to date--notwithstanding both the growing importance of intellectual property law in a technologically complex world(13) and the steady increase in the number of law and economics scholars who have begun to devote attention to this body of law.(14) To be sure, there is a substantial body of work devoted to the issue of whether intellectual property rights should be protected by what Calabresi and Melamed referred to in their famous article as "property rules" or "liability rules,"(15) with most, though not all, economic analysts of law concluding that protection under a property-rule regime is preferable to liability-rule protection.(16) To the extent that this analysis is correct, it suggests that injunctive relief should be the principal remedy available against those who infringe intellectual property rights.(17) The question nevertheless remains what sort of damages should be available to an intellectual property owner for acts of infringement occurring prior to the issuance of an appropriate injunction. Devising the correct answer to this question is of great importance if, as is likely, a significant number of infringements go undetected for more than a nominal period of time.(18)

   Our principal thesis is that the optimal set of damages rules should preserve both the incentive structure of intellectual property law and the property-like character of intellectual property rights. As we demonstrate herein, in the absence of enforcement, information, and other transaction costs, these goals require at a minimum an award that renders the infringer no better off as a result of the infringement. As a first approximation, then, the optimal rule is to award the plaintiff the royalty to which the parties would have agreed prior to the infringement, in cases in which the infringer is a more efficient user of the subject property than is the plaintiff, or the defendant's profit attributable to the infringement in cases in which he is not.(19) After eliminating the assumption of zero costs from the model, however, this rule must be modified in two crucial respects. First, to preserve the owner's incentive to create and to publish, in cases in which for whatever reason the rule fails to deter, the owner always should be able to recover her own lost profit resulting from the infringement.(20) Second, in order to avoid having courts determine the value of intellectual property and to encourage the parties to engage in voluntary bargaining ex ante, the defendant always should be required to disgorge all of his profit attributable to the infringement, unless this would result in a double recovery.(21) As a second approximation, then, the optimal rule is to award the plaintiff the greater of either her lost profits or the defendant's profit resulting from the infringement.(22) On its face, this rule appears to provide the correct incentives for optimal use, inasmuch as lost profits will exceed the defendant's profit only when the plaintiff is a more efficient user than is the infringer, and vice versa.(23)

   As we also demonstrate, however, this second approximation may be subject to further modification in light of two additional factors that inject considerable uncertainty into the analysis. The first is that an award that merely renders the infringer no better off as a result of the infringement may be an ineffective deterrent, because only a portion of all possible infringements are susceptible of detection.(24) This insight suggests that a substantial damages multiplier often may be necessary to achieve adequate deterrence. The second is that the standard of liability in intellectual property cases often is uncertain(25) and that in some instances, such as the 12 Monkeys example, the infringer will have incurred substantial sunk costs by the time his infringement is detected. These facts suggest that, on occasion, the optimal award should be lower than the initial model would advise, in order to avoid the overdeterrence of marginally lawful conduct. As a third approximation, then, the optimal rule is to award the prevailing plaintiff the greater of either a compensatory or restitutionary recovery, suitably enhanced or diminished in light of the competing interests in deterring infringements that otherwise may go undetected, and in discouraging would-be users from overcomplying with their legal obligations.(26) We recognize, of course, that this formulation does not inform the policymaker precisely how to calculate the optimal award in any given case--if anything, the analysis shows that any precise calculation of optimal damages is likely to be next to impossible in the real world(27)--and that it does not offer any easy solution, in a case such as Woods, to the problem of determining how much of a benefit the user derived from the infringement. We nevertheless conclude that this formulation provides a rational overall framework for considering damages issues, and that, notwithstanding some lack of precision, this framework greatly illuminates some vexing issues in patent, copyright, and trademark damages law.

   In Part II, we provide some background information concerning these four bodies of law, including the damages rules that govern each of them. We also briefly describe the predominant economic justifications for these bodies of law and review the literature suggesting that property-rule protection is the preferable method for safeguarding the rights of intellectual property owners. In Part III, we develop a general model of the optimal set of damages rules to apply in intellectual property litigation, and we demonstrate that the rules applicable in trade secret law track our model quite closely, whereas patent...