This paper criticises the longstanding debate about patent holdup – and particularly about whether it is a systemic issue. In short, the paper argues that the ongoing hunt for empirical evidence of systemic patent hold-up in standardised product markets, or lack thereof, is a fruitless academic exercise.
The paper can be found here and is structured as follows:
Part I offers some essential background on standard setting and standards-essential patents. As I am sure we are now all fed up with this, I will skip it.
Part II explores the interrelated questions that form the core of the current hold-up debate: how is hold-up defined, and what can empirical evidence tell us about hold-up in today’s technology-driven markets?
The notion of economic hold-up originated with Oliver Williamson’s leading work on transaction costs and information asymmetry in the 1980s. The owners of specific assets are vulnerable to opportunistic behaviour by potential transaction partners who act dishonestly (e.g., by using deceptive means to argue for a lower price). However, the term “hold-up” has taken on a different and more straightforward meaning in the context of standard-setting, where authors emphasise the manufacturer’s sunk costs and lock-in to a particular technical solution, and the inappropriate leverage that SEP holders could obtain by threatening to obtain judicial injunctions to prevent manufacturers from producing standardised products, usually after lock-in has occurred.
In short, Williamson and transaction cost economists, on one hand, and courts, agencies and commentators who are considering conduct relating to standardization, on the other hand, are using the term hold-up to refer to different types of market behaviour.
Examples of the standard-setting definition of hold-up can be found throughout the case law dealing with disputes over the licensing of standards-essential patents. In addition to individual cases, researchers have sought evidence demonstrating (or refuting) the existence of patent hold-up at a systemic level. The author provides an overview of this research, particularly of how: (i) it has focused on telecommunication markets, where prices have been falling and profit margins for implementers are high. The conclusion is thus that there is no empirical evidence of systemic patent hold-up in wireless telecom or other markets characterised by SEPs and standards; (ii) it has dealt with patent holdout, i.e. the practice of companies routinely ignoring patents and resisting patent demands because the odds of getting caught are small.
Part III challenges the premise that evidence of systemic market hold-up matters, either in assessing the liability of individual firms that have engaged in abusive conduct, or in formulating meaningful policy reform.
Though most commentators seem to agree that hold-up could occur in markets characterised by patented standards, there is significant disagreement over the extent to which hold-up actually does occur in such markets – something already recognised by the FTC in 2002.
In response to the threat of hold-up, scholars including Carl Shapiro, A. Douglas Melamed, Fiona Scott-Morton, Joseph Farrell, Michael Carrier, Mark Lemley and others (including the author) have urged positive intervention by policy makers. Critics, including Stephen Haber, Richard Epstein, Anne Layne-Farrar, David Teece, Joshua Wright, F. Scott Kieff,, J. Gregory Sidak and two then-sitting commissioners of the FTC, Maureen Ohlhausen and Joshua Wright, have argued that that there is little, if any, empirical evidence that hold-up is a pervasive or even a real problem in modern technology markets.
This purported lack of evidence has led some commentators to dismiss individual firms’ complaints regarding hold-up as anecdotal and to conclude that, if hold-up occurs at all in the market, it is sporadic. As a result, these commentators argue, policy initiatives focused on preventing hold-up are unnecessary at best and harmful at worst.
However, it does not follow from lack of evidence of systemic hold-up that hold-up does not represent a threat that justifies policy intervention. There are numerous examples of anticompetitive conduct by individual firms in markets that are not otherwise overrun by anticompetitive behaviour. Examples include not only reverse settlement agreements and merger control, but also patent holdout itself. After all, the threat of patent hold-up was a primary motivation factor for many SDOs to adopt policies requiring the disclosure and licensing of SEPs; and has underpinned a number of enforcement actions by competition authorities on both sides of the Atlantic.
On the other hand, hold-up itself is not a cognisable legal wrong. That is, even if patent hold-up is undesirable for the efficient operation of markets, or hinders the broad adoption of technical interoperability standards, or effects wealth transfers from some market participants to others, or impedes market entry and innovation, these effects alone do not demonstrate that illegal conduct has occurred. What matters is merely whether anticompetitive behaviour has occurred. Antitrust offences may overlap with the exercise of patent hold-up, but in other cases they may not. In order for a violation to occur, a defendant must be shown to have engaged in legally prohibited conduct using established antitrust standards, not the ill-defined economic concept of hold-up.
This is a paper by an author that I always enjoy reading. I particularly enjoyed the discussion of the different meanings of holdup and how they affect competition policy, together with the paper’s overview of the antitrust debates regarding patent holdup.
The paper left me with some lingering questions, though. I agree that it is not clear why evidence of patent holdup being a systemic problem should be relevant to the assessment of individual conduct – but, on the other hand, it is a relevant question for discussing what the enforcement priorities of competition authorities should be, and what tools are best suited to address problems related to holdup.
On the other hand, saying that holdup is only anticompetitive when it infringes competition law seems to me to beg the question of ‘what is the anticompetitive harm here’? According to the author’s reading of holdup, it cannot be the practice of holdup itself – but in that case, what is the anticompetitive harm that holdup creates?
This, in turn, links to a wider issue that has concerned me of late: how do we distinguish holdup from market power, and the exploitation of locked-in implementers/consumers from anticompetitive leveraging? I can see how theories of tying or vertical foreclosure might be relevant, but I am not sure they lend themselves to useful (and clear) guidance in this area. This is something that I would have liked to see discussed by the author. Maybe he will do so – or has already done – in another paper.