TABLE OF CONTENTS I. INTRODUCTION 123 II. HISTORICAL AND RESIDUAL SKEPTICISM TOWARD LICENSING 125 A. Doctrine of Indivisibility 126 B. Exhaustion (or First Sale) Doctrine 127 C. Naked Licensing; Assignments in Gross 128 D. Patent and Copyright Misuse 128 E. Summing Up: Why the Law Thinks It Should Restrict Licensing 129 III. A NEW VIEW OF IP LICENSING 129 A. Licensing as Design Tools 130 B. Licensing and Supply Chain Design 131 C. Licensing and Risk Diversification 134 1. Diversification and Extreme Skew 135 2. Illustrations: Diversification and Organizational Form 136 3. How Licensing Promotes Diversification 137 D. Licensing and Fractionalization 138 1. Fractionalization Strategies 139 2. Why Fractionalization is (Mostly) Good 140 IV. REEVALUATING IP LICENSING LAW 142 A. The Sylvania Precedent: Antitrust Lessons for IP Law 142 B. The IP Licensing Time Warp 143 V. CONCLUSION 144 I. INTRODUCTION
Legal scholars have tended to view the licensing of intellectual property rights (1) with skepticism, calling for judicial and legislative intervention to protect the public domain against encroachment by licensors. (2) The Supreme Court now appears to agree. Out of four Supreme Court decisions pertaining to licensing since 2006, three affirmed precedents that bolster constraints on licensing and one reversed a precedent that had relaxed those constraints. (3) Scholars' and the Court's skeptical view relies on the concern that licensing can be used to expand the IP grant crafted by Congress, the Patent & Trademark Office, and the courts. The renewed drive to enhance constraints on licensing runs counter to historical trends that have tended in the opposite direction. IP law has progressively attenuated doctrines that limit licensing; (4) modern antitrust law emphasizes the efficiencies generated by licensing; (5) and lower courts have largely resisted calls to invalidate mass-market software licenses. (6)
Skepticism toward licensing overlooks the fundamental role it plays in the commercialization process that delivers an innovation to market. Specifically, licensing promotes three core objectives that lie at the heart of real-world content and technology markets. First, licensing permits firms to customize supply chains so as to allocate commercialization functions to the least-cost provider of each function. Second, licensing permits firms to devise diversification strategies that spread the extreme risk of innovation projects. In particular, licensing supports the hub-and-spoke formations that recur across innovation markets, in which larger firms bear production and distribution risk and smaller technology and content originators bear development risk. Third, licensing enables firms to divide innovation assets into sub-assets that can be deployed across multiple parameters in space and time. These "fractionalization" strategies yield efficiency and distributive gains by expanding access across the full spectrum of budget constraints and valuation intensities in the user population.
The skeptical view approaches licensing as a threat to the social bargain that underlies IP rights. This is misguided--licensing implements that social bargain through the commercialization process without which an innovation could not reach market. Failure to take this point seriously has perpetuated formalistic rules that frustrate efficient transactions, compelling the market to engineer second-best arrangements or leading courts to create ad hoc exceptions to avoid economically absurd outcomes. By contrast, antitrust law has essentially recognized this same point for several decades, largely rejecting per se prohibitions on certain licensing practices in favor of "rule of reason" approaches that assess competitive effects on a case-specific basis. It is time for IP law to consider a similar nuanced approach. Academic commentators have long called for courts and legislatures to protect the public domain against overreaching by licensors. The Court now seems to be listening. It may be wiser to consider how to protect the market against overreaching by courts and legislatures.
The organization of this Article is as follows. Part II describes legal doctrines that have historically restricted licensing freedom in IP markets. Part III proposes a new approach to licensing as a tool for structuring supply chains in content and technology markets. Part IV applies that framework to revisit legal restrictions on certain licensing practices, especially as compared with antitrust treatment of those same practices. Part V concludes.
HISTORICAL AND RESIDUAL SKEPTICISM TOWARD LICENSING
The skeptical view tracks long-standing legal doctrines that restrict licensing. While courts have relaxed the application of some of these doctrines, they continue to impede IP transactions. The principal limitations are discussed below, namely: (A) copyright law's doctrine of indivisibility; (B) patent law and copyright law's versions of the exhaustion doctrine; (C) trademark law's prohibitions against "naked" licensing and "assignments in gross"; and (D) patent law and copyright law's versions of the misuse doctrine.
Doctrine of Indivisibility
In 1891, the Supreme Court adopted the "doctrine of indivisibility," which deemed a license anything other than a complete assignment of the rights covered by a patent. (7) In 1908, a federal court extended this principle to copyright. (8) As observed in the Nimmer treatise on copyright, this precedent, as interpreted in light of the Copyright Act of 1909 (9), effectively made it "impossible to assign anything less than the totality of rights commanded by copyright." (10) A transfer of less than all of a copyright risked being deemed a license rather than an assignment, in which case two adverse consequences could follow. First, the licensee might not have had standing to independently sue infringers. (11) Second, if the licensee published an unregistered work without giving copyright notice in the owner's name, the work could fall into the public domain. (12) Despite practitioners' repeated observations that this formalistic doctrine did not conform to market practice, (13) the development of contractual detours around the doctrine, (14) legislative efforts to amend the statute (15), and some courts' efforts to limit or skirt the doctrine entirely, (16) it was not explicitly abolished until the Copyright Act of 1976. (17) Nonetheless, the doctrine stills casts a shadow over copyright transfers: in 2002, the Ninth Circuit applied what it viewed as a surviving element of the doctrine of indivisibility to bar a copyright licensee from engaging in resale transactions without the licensor's consent. (18)
Exhaustion (or First Sale) Doctrine
The exhaustion doctrine (and its copyright analogue, the first sale doctrine) provide that, upon the first sale of an article embodying a patent or copyright, the IP holder can no longer assert legal control over the article's subsequent use or distribution. Courts adopted the doctrine in patent law in 1852 (19) and copyright law in 1908 (20) (codified by Congress in 1909 (21)). In copyright, owners must engage in contractual fictions that characterize sale transactions as licenses to avoid triggering a "first sale." While courts typically honor this drafting trick, (22) some residual uncertainty persists. In patent, the exhaustion doctrine compels patent holders to engage in contractual detours in order to control use of a patented technology throughout a supply chain. As the Supreme Court suggested in 2008 (effectively reversing a Federal Circuit precedent in place since 1992 (23)), a sale of an article embodying a patent is deemed to have exhausted the patent even if the sale is subject to a contractual condition. (24) Under this reasoning, patent holders that wish to exert downstream control in a supply chain have two imperfect strategies to avoid exhaustion. First, the patentee can eliminate the supply chain and vertically integrate forward to the point of sale--an infeasible proposition in many if not most technology markets. Second, the patentee can maintain the supply chain but distribute the IP-embedded product subject to a license that sets the terms on which the licensee may use the product or distribute it to any sub-licensees. (25) Properly structured, this contracting device approximates, but does not match, the legal certainty that would be achieved through the more potent combination of patent plus contract.
Naked Licensing; Assignments in Gross
Historically, trademark law effectively prohibited the licensing of trademarks except in connection with the transfer of the associated business. (26) While the Lanham Act of 1946 abandoned this approach, (27) trademark licensors still can run afoul of the ban on "naked licenses," which requires that licensors exercise adequate control over licensees. In a 2011 Seventh Circuit decision, application of this doctrine resulted in abandonment of a mark that had been in continuous use by the owner and its predecessors for almost 50 years. (28) Relatedly, transfers of ownership in trademark can run afoul of the "assignment in gross" doctrine, which requires that any transfer include the goodwill or business associated with the mark. (29) While transfer agreements typically avoid this issue by a recital of the underlying goodwill (and courts have been receptive to this practice), this is not foolproof since courts occasionally apply the doctrine to invalidate assignments. (30)
Patent and Copyright Misuse
The misuse doctrine, which courts have implied in patent and copyright law, bars the enforcement of licenses that "improperly" exploit the underlying IP right. This imprecisely defined claim is especially potent since, when successful, it results not only in invalidation of the license but also the underlying IP right until the misuse is corrected, thereby barring (at least temporarily) the patentee from injunctive...