This piece – which can be found here – is a rather long , but very comprehensive book chapter that surveys and integrates the economic, business strategy and legal literatures on IP, competition and network effects. It is structured as follows:
Part I introduces network effects. I have done this to death in the past, so I’m not going to repeat it here. Suffice it to say that the author looks mainly at demand side network effects, and what its implications are for IP and competition policy: ‘In a static economic model (i.e., one without innovation), consumers benefit from robust competition within product standards. Open access to product standards encourages realization of network externalities. Although bandwagon effects can enhance consumer welfare in a static context, they can also make it more difficult for developers of improved platforms to enter the market. Consumers and suppliers of complementary products can face significant switching costs in migrating from one platform to another.’
Like economies of scale on the supply-side of a market, the value of a network generally increases with widespread adoption. However, whereas economies of scale typically fall off at some point due to technical or organizational limits, the strength of network effects can typically continue to increase with the size of the installed base.
This means that networks that benefit from such effects and become legal or de facto standards have high value for consumers. On the other hand, such standardisation can make it particularly difficult for improved products to break into the market; bandwagon effects like those present in standards that benefit from demand side network effects can stifle the development and diffusion of improved technology platforms. As a result of this, the recent proliferation of industries where network effects are at work poses new challenges for both IP and competition law.
Part II describes the functioning of network markets. The paper begins by noting that network effects have long been central to human civilization and market economies – and include the adoption of such standards as languages, measurement systems (metric versus imperial), electrical equipment standards (alternating current versus direct current), computer networking protocols, driving conventions (left side versus right side) and railroad gauges. As the importance of these standards demonstrate, ‘The economic and social value of network effects can be substantial’. In practice, ‘the magnitude of network effects depends on several considerations: interdependencies of consumer utility functions, range of complementary products or services, availability of alternative platforms, switching costs, business strategies, and legal limits (such as intellectual property protection and competition policy) on leveraging network markets.’ Furthermore, the design determinants of a network market – such as interoperability or compatibility standards – are shaped by the type and degree of ownership, sponsorship, and governance of network access (including the level of openness of the network).
Part III examines the interplay of business strategy, contract law, standard setting organizations, intellectual property, and competition policy. Firms typically choose among three strategies when competing in network markets: (1) market dominance through establishing and controlling a new proprietary standard; (2) adopting an existing standard either through imitation (where it is legally permissible) or licensing; or (3) working with other firms in the industry to develop an open or quasi-open standard. Firms often pursue Strategy 2 (adopting an existing standard) and Strategy 3 (collaborating with other firms in establishing a standard) in tandem. Both strategies create a more traditional market setting in which firms compete over price, quality, and services to win market share on a common platform.
Intellectual property protection for network technologies can significantly influence the development of standards, follow-on innovation, and market competition. Patents in the information and communication technology fields (e.g. semiconductors, computers, and mobile phones) have presented the most salient concerns. As an example of this, the author reviews how hold-up theory has been used to identify antitrust problems relating to FRAND and the academic debate on the topic.
Part IV looks at the implications of network effects for IP and competition law. Network features of system technologies, like any other technology, are subject to the classic appropriability problem. Without intellectual property protection, inventors of more advanced platform technologies will be subject to being undercut by new entrants who imitate the innovations without bearing R&D costs, which may lead to underinvestment in innovation. This is particularly the case given the risks that challengers of established networks – which are protected by a locked-in base and high switching costs even if the standard is obsolete – face. As a result, IP rights can have two contradictory effects. On the one hand, intellectual property protection for the network features of a technology can discourage the realisation of positive network externalities by limiting access to network technologies. On the other, intellectual property protection can provide valuable incentives for overcoming bandwagon effects that entrench obsolete standards.
Given this, the author presents three principles for tailoring intellectual property regimes and competition policy for network technologies:
- The parsimony principle – Intellectual property regimes should discourage platform adoption choices that undermine realisation of network externalities unless there is a large countervailing benefit, such as a substantial technological advance. Affording meaningful intellectual property protection for network technologies without requiring a significant technological advance encourages wasteful differentiation and increases the risk of undeserved monopoly power.
- The proportionality principle – This principle seeks to ensure that firms can appropriate a fair return on technological advances in platform innovation sufficient to overcome the excess inertia of network markets, but not so large as to stunt network externalities. Robust protection for substantial technological advances may be necessary for entrants to overcome the strong inertial forces driving network markets. At the same time, the proportionality principle ensures that platform innovators who choose proprietary strategies (as opposed to more collaborative approaches) have the potential to reap significant rewards if they prevail in a competition between standards, but that their ability to control the platform (and charge monopoly prices) declines as the network becomes entrenched. Such a regime creates optimal conditions for overcoming excess inertia while promoting the realization of network benefits.
- The deterrence principle – This principle seeks to discourage deceptive practices and overreach in network markets. The pursuit of this principle requires competition law policy to take network effects into account when assessing monopoly power, scrutinising collaborative arrangements, and fashioning remedies. As regards remedies, their crafting to combat abusive and anti-competitive behaviour in network markets requires careful consideration of effects on consumers and competitors: ‘antitrust authorities and courts should consider remedies that promote the realization of network benefits while promoting enhanced competition and innovation. In some circumstances, these considerations favor compulsory licenses, which can be flexible and adaptable, over injunctive remedies.’
Part V traces the evolution of intellectual property protection for network features of systems and platforms. Although patents have long protected platform technologies (e.g., electric standards such as AC/DC, phonogram, colour television, and telecommunications), more recently the debate on the contours of intellectual property protection for network features of systems and platforms has revolved around software. Given the immense value of controlling access to a platform, platform developers and entrepreneurs have sought to use intellectual property to protect APIs and other means to exclude competitors from their platforms and systems. This section maps out how trade secrecy and contract law provided relatively effective protection for much of the software developed during the mainframe and minicomputer eras. Although advances in computer hardware fell squarely within the patent domain, there were significant doubts about the patentability of computer software into the 1990s. As microcomputers emerged, which spurred retail distribution of computer software, copyright law emerged as the primary battleground for computer software by the mid-1980s. The role that copyright has played in the software industry since then is mapped out. The section ends with a discussion of how patent protection emerged as a more robust and controversial form of protection for computer software in the 1990s.
Part VI discusses the interplay of intellectual property protection and competition policy in the US. It is divided into two sections:
- Section A explores the limitations of theories of improper leveraging of intellectual property rights that arise in private litigation. The discussion is framed by a number of U.S. legal doctrines, such as the patent misuse and the patent exhaustion doctrines which have been reviewed in previous posts (of 19 May 2017 and, more recently, 16 January 2018). The author also reviews concerns related to the standard setting process, namely regarding how the process may be ambushed – the two leading examples of this being cases brought against Rambus and Qualcomm – and how FRAND licensing should operate. Finally, the author provides an overview of additional matters that are particularly relevant to network technology markets: (i) refusal to license patented technologies; (ii) patent thickets; and (iii) leveraging of monopoly power.
- Section B examines the public enforcement of antitrust law in network markets. Network technologies are prone to high concentration levels that can enhance consumer welfare through network effects. Therefore, antitrust authorities have had to shift their focus away from market concentration toward anticompetitive tactics such as leveraging market power into new markets and stifling innovation. This section summarizes the major contours of this shift. First, it discusses the evolution of Department of Justice and Federal Trade Commission guidelines for intellectual property licensing. Secondly, it reviews the most significant network market enforcement actions in the US over the past two decades (including actions against Dell, Intel and Microsoft).
Part VII assesses how the various legal, market, and policy institutions – and, more notably, IP and competition law – have adapted to the emergence of network technologies in the Information Age by reference to the principles outlined in section III. The conclusions are, broadly speaking, that the extent of IP rights’ protection goes beyond what it should under the parsimony principle; and that the level of patent protection for interface design and other network technologies has both fluctuated over the years and been a complex and costly tool for achieving proportional appropriability of network technology innovations.
The deterrence principle is directly concerned with antitrust, so I’ll cite the author: ‘The past several decades has witnessed substantial evolution of antitrust doctrines and enforcement policies toward a balanced innovation and competitive ecosystem. Antitrust enforcers have come to appreciate the economic benefits of high concentration in network technology markets while also focusing on abusive practices, such as failure to disclose essential patents to standard setting organizations. (…) courts have broadened their assessment of antitrust, contract, and patent remedies in view of network effects. The dynamism of network technologies and markets, however, will continue to challenge enforcers, policymakers, and courts (…) there is a subtle line between promoting interoperability and encouraging innovative forking of established standards.’
I think this is a really useful piece for anyone trying to understand the role that IP law (and, to a lesser extent, competition law) have played in the development of the digital economy, and the challenges of regulating it appropriately. Section V, in particular, contains a very good discussion of the role that IP rights played in creating the contemporary digital economy; and section VI’s discussion of US competition law enforcement in network industries is easy to follow and comprehensive.