|
Article Excerpt
Denying patent holders extensive control over experimental use of their technologies has been widely lauded as an essential facilitative component of the innovative process, one that is to be enthusiastically embraced and expanded. If one posits the cumulative nature of innovation, the possibility of irrational holdout, and the view that patents are awarded in return for the disclosure of valuable information, widespread support for an expansive construction of the doctrine is unsurprising. The position flows naturally from a liability-based conception of intellectual property.
Yet the Federal Circuit has taken a position diametrically opposed to the majority view, limiting the experimental use doctrine to a degree that essentially renders it inoperable. This Article considers the normative foundation of the Federal Circuit's restriction and, in contrast with the prevailing academic view, concludes that it is entirely desirable. Bestowing patent holders with exclusive rights over experimental use facilitates Coasian bargaining in what is likely to be a low transaction costs environment and, more important still, is likely to enhance dynamic efficiency. Scholarship expressing a contrary view tends to place unwarranted focus on an "incentive to disclose" theory of patent rights, a theory which this Article demonstrates is neither plausible nor justified.
To the extent that considerations such as information asymmetry, transaction costs, irrationality, and bargain failure provide conceivable grounds for imposing compulsory licenses in other contexts, they have limited applicability with regard to experimental use. To import unthinkingly such considerations from other contexts is to overlook the crucial facilitative mechanisms likely to be associated with a strong property rights regime. Even horizontal competitors, who may understandably be very hesitant to license technology to each other, may nevertheless reach an agreement through reciprocal licensing deals, patent pools, grant-back provisions, or simply a sufficiently high fee. The oft-cited prospect of "irrational" holdout is overstated and, to the extent it exists, is most likely simply a reflection of a proffered licensing fee that undercompensates the patentee. Recognizing the right to exclude that lies at the heart of the patent grant fuels the sole incentive that truly matters, namely, the incentive to innovate.
The folly of the expansive view of experimental use is magnified when one considers the substance of the right it advocates. Its proponents counsel a regime of zero compensation, a price recognized by economists as necessarily inefficient. Some scholars' assertion that a zero-compensation regime will have negligible impact on patentees' "incentive to invent" is unsupportable.
INTRODUCTION I. FRAMING THE ISSUE IN CONTEXT A. A Brief Overview of the Law B. The Property Rights Movement and the Liability Faction 1. Utilitarianism as the Exclusive Foundation for Intellectual Property 2. Informing the Consequentialist Inquiry II. RESTRICTING EXPERIMENTAL USE A. Patent Policy as an "Incentive to Disclose" B. Employing the Exemption to Avoid Harm to Ex Ante Incentives C. The Critical Importance of Cumulative Innovation D. The Optimal Exemption CONCLUSION
INTRODUCTION
The degree to which information should be subject to private ownership poses one of the most hotly contested questions in contemporary academic discourse. (1) This Article covers a subset of the larger debate, considering the question of whether a patentee's right to exclude includes the authority to prevent others from experimenting on, or with, his invention. In doing so, it considers Judge Newman's argument that "[t]he right to conduct research to achieve ... knowledge need not, and should not, await expiration of the patent." (2) In contrast to prevailing academic opinion, the Article concludes, no such general right should exist.
The propertization of information flows from an economic conception of innovation. Valuable invention and expression display public good characteristics of non-excludability and non-rivalry in consumption, which cause private markets to undercompensate inventors, authors, and artists, and thus underproduce desirable goods. (3) Informed by economics, the intellectual property laws artificially bestow information goods with the defining trait of physical property--excludability-thereby creating a market. (4) But this facilitative mechanism also carries the negative effects of increasing the price of information beyond some consumers' reservation levels and, worse, possibly hindering the innovative process itself. (5) Progress in the sciences and arts depends critically on prior understanding. (6) Indeed, the overwhelming impetus for valuable discovery emanates not from individual genius or effort, but from antecedent knowledge. (7) Granting entities exclusive rights over their inventions and artistic expressions may deny others access to the information they need to engage in unquestionably valuable "cumulative innovation." (8)
Empirical investigation has yet to determine conclusively whether bestowing information with the full attributes of physical property enhances social welfare by promoting dynamic efficiency, or restricts it by creating a perverse "anticommons" in which valuable information is not shared and technological development is frustrated by holdout. (9) Ultimately, economic-minded commentators agree that the optimal level of intellectual property protection is one that maximizes the difference between the benefits created by ex ante incentives to innovate, on the one hand, and the costs of allocative inefficiency and conceivable holdup of cumulative innovation, on the other. (10) Unfortunately, imperfect information precludes one from applying such a cost-benefit analysis in a determinative fashion. Accordingly, academics have differed sharply on the level of propertization that acts as an optimal heuristic, fiercely debating the extent to which patents and copyrights can be analogized to the law governing tangible property. (11)
Those advocating the adoption of a system similar to that for tangible property rights support the near-Blackstonian position (12) that a property owner's ability to exclude others should be almost inviolate. (13) They argue that strong property rights promote the efficient allocation of valuable information through contract, further facilitate post-grant commercialization, (14) and create optimal incentives by allowing patentees and copyright holders to extract the social value of their inventions. (15) Others reject a property-based perspective, instead favoring a liability approach that would grant third parties access to intellectual-property-protected information at a fee. Such commentators typically adopt a parsimonious view. They accept that patent or copyright is necessary to overcome the nature of information as a public good, but advocate a regime that still provides the minimum level of pecuniary compensation necessary to induce innovation. (16) A third camp considers the allocative-inefficiency and blunt-incentive characteristics of intellectual property to be good reasons for replacing the patent system with a reward structure. (17) Finally, there are those who doubt that intellectual property protection is necessary to induce desirable innovation at all, (18) touting the mantra that "information just wants to be free." (19) The normative justification for this agnostic position emphasizes above all the altruistic nature of much innovation, particularly in the copyright realm. (20)
The preceding divergence in academic opinion bears witness to the myriad disagreements in the realm of intellectual property. Within this larger framework, a particularly divisive point of contention has emerged concerning a third party's right to experiment on another's patented technology. Given the Byzantine nature of some technological discoveries, it may be difficult to re-create accurately or economically many inventions without some degree of experimentation. Should an intellectual property holder be able to demand a fee for such use or even enjoin it altogether? Should a patentee have any rights over such use at all?
There are, initially, many appealing reasons to think that he should not have any such rights. From a deontological perspective, one might argue that an inventor who gains the luxury of a twenty-year monopoly is contractually obligated to reveal in return precisely what his invention is and how it works. (21) Indeed, some scholars have drawn this inference solely from the statutory "enablement" requirement, (22) which demands that an inventor explain the functioning of the relevant technology in a manner that would allow one skilled in the art to reproduce it "without undue experimentation." (23) Many have also spoken of an "incentive to disclose" rationale for intellectual property protection. (24) This rationale is ostensibly reflected in the Patent Act, which requires that a prospective patentee both fully disclose the "best mode" for practicing his claim (25) and satisfy the enablement requirement. Certainly, an experimental use exemption for patents could hardly be called anomalous, for one is codified in the Copyright Act. (26)
Consistent with these considerations, the overwhelming weight of academic opinion supports an expanded experimental use exemption under patent law. (27) Their view is far from radical--indeed, several foreign countries have facilitated free experimental use by statute. (28)
In light of these strongly voiced opinions, and a judicial environment increasingly hostile to the property rights movement, (29) one would expect a broad exemption for experimental use. Yet in a series of recent cases, the Federal Circuit has curtailed the doctrine to a degree that essentially renders it defunct. (30) In Madey v. Duke University, the court went so far as to refuse to exempt the infringing activity of a research university, (31) shattering the long-held presumption that academic researchers were immune from liability for patent infringement. (32) In her later dissent in Integra Lifesciences I v. Merck KGaA, Judge Newman lamented the court's decision in Madey, opining that it "essentially eliminate[d]" the well-established common law exemption and ignored patent policy's role in providing an incentive to disclose. (33)
Contrary to the weight of academic opinion, (34) this Article argues that the Federal Circuit's near elimination of the experimental use doctrine is justified. The court's holding that the exemption persists only to the extent of "philosophic inquiry" ensures that the only unauthorized use that can take place is that which has no economic consequence for the patentee in the marketplace. Providing a patentee with an exclusive right--as the Federal Circuit has done--ensures that ex ante incentives remain optimal. Allowing exclusion is appropriate in all situations except one.
When a patent is improvidently granted by the Patent and Trademark Office ("PTO')--a phenomenon that unfortunately occurs with some regularity--exclusion is decidedly not proper. (35) Given the pervasive poor quality of issued patents, there is strong normative ground for allowing rivals and others to experiment with another's patented technology to the extent necessary to assess the validity of the patent. (36) Improperly acquired patents cause serious social harm and restrict the dissemination of valuable information with no offsetting, concomitant benefit. (37) Facilitating the challenge of questionable patents therefore constitutes a strong policy mandate.
Several academics forcefully make the case for an expanded experimental use doctrine, (38) but their analysis falls short for a number of reasons. Perhaps most unsupportable is the contention that one can grant competitors freedom to experiment on an invention without harming ex ante incentives. Professor Strandburg, for instance, urges a dichotomy between experimenting "on" versus experimenting "with" patented technology and concludes that the former will not have a material effect on prospective patentees' incentive to innovate. (39) This position fails to withstand logical scrutiny. Further, proponents of an enlarged experimental use doctrine uniformly place an unwarranted focus on patent law's supposed "incentive to disclose" foundation. (40) This Article argues that the sole purpose of patent policy is to spur desirable innovation.
The strongest conceivable argument in favor of a broad experimental use exemption relates to the problem of technological holdout, identification and transaction costs, and other impediments to cumulative innovation. By ensuring that all innovators have access to the full amount of information concerning a patented invention--a level of understanding likely to be gained only through actual experimentation--society ensures that the maximum amount of valuable information is passed to the public, as early as possible. The ensuing knowledge will not necessarily directly benefit society in the short-run, for a competitor cannot normally sell a patented improvement without the original inventor's blessing, (41) but there may well be indirect or eventual benefits. A patentee's refusal to allow a competitor to experiment with, or on, his invention may delay all these gains.
Yet even concern with maximizing information available to innovators fails to justify a more expansive experimental use doctrine than that established by the Federal Circuit. First, its call is not as compelling as it may initially appear. The vast majority of patented inventions are "self-disclosing" and require little in the way of testing to understand. (42) Moreover, patents are not issued in the dark; applicants are required to disclose an invention's best mode of practice and satisfy the enablement condition. (43) Thus, even a patentee who "unreasonably" holds out makes much information publicly available for cumulative innovation, "inventing around," and other activities. Just as important, there is an underappreciated "quasi-fair use" aspect to patent law, which allows much experimentation to take place free from challenge. Given the extraordinary expense of patent litigation--an expense that is entirely absorbed by the patentee-there is an environment of considerable under enforcement. (44) As a result, the only experimental uses that will likely give rise to infringement proceedings are significant research activities by horizontal competitors.
Second, unrestricted experimental use will inevitably take place when an invention enters the public domain upon expiration of its patent. This consideration is entitled to greater weight within the field of patents than it is in copyright, given the former's considerably shorter duration. (45)
Third, and most importantly, the number of "irrational" holdouts is likely to be less than many fear. Negotiating a license to experiment with a patented technology is, in many ways, like negotiating for a license to sell a product incorporating an improvement patent. Both theory and empirical evidence demonstrate that strong property rights can yield mutually beneficial bargains. These bargains often take the form of simple pecuniary compensation and social convention, but, increasingly, they occur through such sophisticated exchange mechanisms as individual cross-licensing, grant-back provisions, and patent pools. (46) Ultimately, little evidence supports the view that excessively powerful patent rights can create an anticommons. (47)
Part I of the Article discusses the leading case law that has created and molded the experimental use exemption. Given that these cases are well known and widely understood, the discussion will be brief. Part I also considers in greater detail the overarching debate concerning the respective primacy of property and liability rules in the intellectual property setting. Defining the proper reach of the experimental use doctrine constitutes only a limited, albeit an important, subset of this larger controversy.
The heart of the Article follows in Part II, in which the views of those who favor a more substantive experimental use doctrine are critically assessed and found wanting. In particular, the incentive-to-disclose theory is both a vacuous concept and a misnomer. As it is usually construed, it provides no normative support for those seeking an expansion of the exemption. Nor are the theory's proponents justified in arguing that an exemption could be applied without negatively affecting ex ante incentives to engage in the innovative process. Finally, it is a mistake to think that strong property rights will foreclose socially valuable cumulative innovation in the context of experimental use.
The question of whether propertization will foreclose efficient access to information is difficult, and it should not be resolved in favor of full proprietary control of information generally. There is much to be said for a parsimonious position insofar as it is arrived at ex ante through an appropriate revision of patent and copyright duration and breadth, as opposed to indeterminate ex post dilution of inventors' and authors' rights. Liability rules may be justified in some limited situations where bargaining costs are apt to be disproportionate to the social gain sought to be achieved. With respect to experimental use, however, where the end sought (disclosure) is largely satisfied by other means, and where the end itself is subsumed within the larger incentive to innovate, the case for such governmental intervention, and the associated dilution in private property rights, is weak.
I. FRAMING THE ISSUE IN CONTEXT
A. A Brief Overview of the Law
In his famous 1813 decision in Whittemore v. Cutter, Justice Story first articulated the principle that has led the courts to recognize a common law doctrine of experimental use. He proclaimed that "it could never have been the intention of the legislature to punish a man, who constructed ... a machine merely for philosophical experiments, or for the purpose of ascertaining the sufficiency of the machine to produce its described effects." (48)
This original conception of the doctrine is considerably broader than the current version. In particular, the second use Justice Story exempted would grant any person the right to experiment on a patented technology to confirm its efficacy. In order to accomplish this goal, an experimenter would obviously be required to investigate how the invention operates. (49) Accordingly, this reading of the doctrine would grant experimenters considerable freedom to research others' patented inventions.
Justice Story's view on experimental use gradually gave way to a standard that distinguished between commercial and noncommercial use of a patented technology. (50) The Federal Circuit embraced this dichotomy in Roche Products, Inc. v. Bolar Pharmaceutical Co. (51) There, the Federal Circuit held that a generic drug producer's testing of a manufacturer's patented drug formula constituted infringement and was not protected by the experimental use doctrine. (52) The defendant hoped to satisfy the labyrinthine statutory requirements in advance of the expiration of the plaintiff's patent, so as to be able promptly to begin marketing its generic drugs. The court based its decision on the commercial purpose underlying the defendant's experimental use. (53) It explained:
[Defendant's] intended "experimental" use is solely for business reasons and not for amusement, to satisfy idle curiosity, or for strictly philosophical inquiry. [Defendant's] intended use of [the patented information] to derive FDA required test data is thus an infringement of the [patent]. [Defendant] may intend to perform "experiments," but unlicensed experiments conducted with a view to the adaption of the patented invention to the experimentor's business is a violation of the rights of the patentee to exclude others from using his patented invention. It is obvious here that it is a misnomer to call the intended use de minimis. It is no trifle in its economic effect on the parties even if the quantity used is small. It is no dilettante affair such as Justice Story envisioned. We cannot construe the experimental use rule so broadly as to allow a violation of the patent laws in the guise of "scientific inquiry," when that inquiry has definite, cognizable, and not insubstantial commercial purposes. (54)
The practical repercussions of the Roche Products decision for the pharmaceutical industry were quickly rendered moot by congressional enactment of the Hatch-Waxman Act, (55) which granted generic manufacturers free rein to experiment on brand-name patented drugs. (56) However, the commercial/noncommercial divide underlying the decision was subsequently re-affirmed in Embrex Inc. v. Service Engineering Corps In that case, the Federal Circuit cited Roche Products with approval and placed determinative weight on the fact that the defendant "performed the tests expressly for commercial purposes." (58) Notably, the court was unmoved by the fact that the defendant's infringement emanated from its attempts to design around the patent. (59)...
|
