Patent holdup, patent remedies, and antitrust responses.

• Author: Cotter, Thomas F.

1. INTRODUCTION II. THREE PRINCIPLES III. WHAT PATENT HOLDUP IS A. Patent Holdup as a Type of Market Failure B. Other Necessary Elements of Patent Holdup IV. PATENT REMEDIES V. ANTITRUST RESPONSES A. Patent Ambush as an Antitrust Offense B. Collective Bargaining as to Patent Terms and Conditions VI. CONCLUSION I. INTRODUCTION

   In recent years, influential scholars, (1) practicing lawyers, (2) government officials, (3) government commissions, (4) enforcement agencies, (5) and courts (6) have all identified the phenomenon of "patent holdup" as a serious problem that may require various reforms to both patent and antitrust law. Within the last year or so, however, critics of this view have become increasingly vocal. In two recent papers, for example, Damien Geradin and his coauthors argue that, as an empirical matter, the frequency and magnitude of patent holdup costs are exaggerated. (7) A second line of attack, taken up in recent work by scholars including Einer Elhauge, (8) John Golden, (9) and J. Gregory Sidak, (10) focuses more on perceived theoretical vulnerabilities of the patent holdup literature--arguing, for example, that holdup is not necessarily inefficient, (11) and that neither patent law nor economic theory provides a baseline from which to evaluate whether patentees' royalty demands are so excessive as to constitute holdups. (12) Third, some of these same critics (and others) argue that the reforms sometimes proposed to remedy patent holdup-such as eliminating the presumption of injunctive relief in patent infringement cases, changing the method by which courts calculate reasonable royalties, and permitting standard setting organizations (SSOs) (13) to engage in collective bargaining with member patent owners over proposed licensing terms-threaten worse harms than the harms they would deter. (14)

   In this Article, I address critiques two and three above (leaving the assessment of the empirical case against patent holdup to others, for now at least); 15 and I focus on two sets of policy tools in particular, the law of antitrust (generally) and the law of patent remedies. (16) In particular, I will argue, first, that many of the difficulties the critics claim to have uncovered in patent holdup theory may be due to their failure to appreciate patent holdup as a variation on other types of "holdup" or "holdout" behavior as discussed in the law-and-economics literature. My analysis leads to a proposed definition of patent holdup as occurring (1) when a component patent owner (17) (2) is able to exploit its bargaining power vis-a-vis downstream users (3) due to the possibility that the patent owner will be able to enjoin the manufacture, use, or sale of an end product that incorporates the patented invention, (4) in such a way as to threaten either (a) static deadweight losses far out of proportion to any likely increases in dynamic efficiency, or (b) dynamic efficiency losses due to downstream users' reduced incentives to invest in standard-specific technology or to engage in follow-up innovation. (18) Second, I will argue that patent law is generally more appropriate than antitrust law to deal with patent holdup, when it occurs, subject to two possible exceptions. (19) Third, potential error costs should play a large role in deciding how aggressively courts should police patent holdup. Courts' abilities to discern the presence of serious static or dynamic efficiency losses, and to respond appropriately, may be subject to reasonable dispute-though, I will argue, these limitations are not so profound as to counsel against any reforms.

   Part II centers my analysis as part of a larger project that seeks to delimit the proper relationship between patent and antitrust based on three interrelated principles: first, that the ideal patent law would be structured so as to maximize the surplus of cognizable social benefits over cognizable social harms; second, that antitrust law should condemn the willful acquisition or maintenance of monopoly power but not its mere exercise; and third, that courts applying patent remedies or antitrust doctrine should take the substance of patent law as a given, and not fashion patent remedies or the law of antitrust to correct for errors that they perceive in that substance. (At the same time, patentees surely enjoy no immunity from compliance with antitrust standards; and courts legitimately can apply the law of remedies to prevent patentees from garnering more than that to which the substantive patent law and policy, reasonably construed, entitles them.) Part III argues for an understanding of patent holdup along the lines summarized above. Parts IV and V then discuss possible responses of the law of patent remedies and of antitrust, respectively, to allegations of patent holdup. As we shall see, application of Part II's three proposed principles hardly leaves courts powerless to deal with holdup when it occurs, although the principles do counsel that serious attention be paid to error costs. Part VI concludes.

2. THREE PRINCIPLES

   In this Part, I present three general principles that inform my analysis of patent holdup. None of the three necessarily commands universal acclaim, and even among observers who agree upon all three at some level of generality there is likely to be some disagreement over particular applications. I will argue nevertheless that all three principles are essential to the optimal accommodation of competition and innovation policies.

   The first principle (Principle 1) is the simple one that the ideal patent law would be structured to maximize the surplus of cognizable social benefits over cognizable social harms. Cognizable benefits would include the familiar benefits of invention, disclosure, innovation, and coordination. (20) Cognizable social harms would include the familiar harms of static deadweight loss, transaction costs, administrative costs, and error costs, as well as dynamic efficiency losses. (21) In applying Principle 1, economic analysis can be useful, though often not dispositive. In particular, it is extremely difficult, if not impossible, to calibrate precisely what effect--or even if the effect is, on net, positive or negative--an additional increment of patent protection will have on social welfare. Courts and other policymakers nevertheless should be wary of reaching decisions based on intuitions derived from "folk economics" or inappropriate analogies of intellectual to real property. For example, although courts at one time expressed concern that certain licensing practices--such as conditioning a patent license upon the licensee's agreement to purchase unpatented products from the patentee, (22) or licensing a patent only on condition that the licensee agree to pay royalties after the expiration of the patent (23)--might enable patent owners to extend or expand their "monopolies," economic analysis suggests that these practices often do not enable patentees to expand their control over well-defined product markets. (24) Contemporary patent law has shed some, but not all, of this baggage. (25) Even today, one occasionally hears concerns (to cite one example) that allowing patentees to recover as damages the lost profits they would have earned on sales of patented products and other complementary products (so-called "convoyed goods") they would have sold along with the patented products, but for the infringement, because critics believe it improperly extends the metes and bounds of the patent to cover those complementary products. (26) Because the recovery of lost profits on convoyed goods generally does not extend the patent owner's market power, however, there is no obvious economic reason to forbid such recoveries, notwithstanding folk economics concerns that recovering lost profits on convoyed goods oversteps the patent boundary. 27

   The second principle (Principle 2) is that antitrust enforcers (1) should concern themselves with the unlawful acquisition or maintenance of monopoly power, where "unlawful" means that the acquisition or maintenance is (a) not excused by other law (such as patent law), (b) not the result of "superior skill, foresight and industry," (28) and (c) not justified by legitimate business reasons; (29) and (2) should not concern themselves with (a) the lawful acquisition or maintenance of monopoly power, or (b) the mere exercise of lawfully obtained monopoly power. Principle 2 itself is based upon two more fundamental underlying principles. The first is that, for antitrust purposes, only harms to competition, not to individual competitors, are cognizable. (30) Harms to competition generate not only the deadweight loss that results from monopoly, but also longer-term harms to dynamic efficiency. (31) The second is the by-now familiar criterion that antitrust law should seek to minimize the aggregate costs of false positives, false negatives, and administrative costs. (32) In combination, these two more fundamental principles suggest several reasons for adopting Principle 2. First, as a general matter, the "acquisition or maintenance" of monopoly power threatens social welfare due to its impact upon static and (possibly) dynamic efficiency. (33) Immunizing monopoly acquisition or maintenance from antitrust scrutiny presumably would give rise to many false negatives. (Indeed, it is difficult to perceive exactly what purpose antitrust would serve if monopoly acquisition or maintenance were per se legal.) Second, however, not all monopoly acquisition or maintenance should be illegal. To render unlawful the otherwise lawful acquisition or maintenance of monopoly power (for example, the acquisition or maintenance of an economically valuable patent) would not only contravene the entire raison d'etre of patent law (as partly established in Principles 1 and 3); (34) it would also generate many false positives, by condemning conduct that (the law presumes) stimulates greater gains in terms of dynamic...