10 Dec 2009

(by Bruce Lyons) The European Commission it has finally accepted undertakings from Rambus to more than halve the royalty rates on its patents embedded in an industry standard for DRAMs (computer memory chips). Two weeks ago, the Commission quietly whose patents are embodied in the 3G mobile phone standard. Was this Neelie Kroes running out of time and tidying up tricky cases before incoming Commissioner Almunia starts rearranging her office furniture? Or were these more significant deals setting out the economic principles underpinning competition law in relation to Intellectual Property Rights (IPRs) in industry standards?

The common thread in these cases is that each firm holds patents embedded in an industry agreed standard that is essential for downstream manufacturers. They both also stood accused of exploitative pricing of these patents. Now, there is wide () agreement on the need for patents (i.e. time-limited monopoly rights) as the least imperfect way to incentivise the type of invention that brings the most economic benefit. The basic argument is that monopoly profits are highly correlated with the social value of an invention and so provide a good incentive for appropriate R&D effort. Also, it is often efficient for an innovation specialist to license a patent to downstream manufacturers who have complementary competence to embody the technology in a marketable product. In such cases, the reward to invention is taken by the monopoly pricing of royalties. It would be a nightmare of litigation if the pricing of all such licenses was subject to antitrust scrutiny. In fact, it would completely undermine the whole purpose of a patent system. So why have these cases caused so much excitement in competition authorities across the globe?

Let’s start with Qualcomm. During the standard setting process, they had apparently promised to license their technology on fair, reasonable and non-discriminatory terms (FRAND). Once part of the standard, however, they were in a powerful position to renegotiate and set a higher than expected royalty – patent ‘hold-up’. Nokia, Eriksson and all the other mobile phone producers complained that this was unfair, would result in higher handset prices and inhibit the development of 3G. Furthermore, it would adversely affect future generations of standards by making agreements more difficult to negotiate.

The economics of this are straight out of the theory developed by , one of this year’s Nobel Prize winners. Even when there are many equally good technologies that could ex ante be incorporated into a standard, once one is adopted, there is a fundamental transformation that gives the adopted technology new monopoly power. If businesses anticipate that this new power will be exploited opportunistically, this will encourage more defensive and less efficient investment. Now the most appropriate solution would have been a sensible enforceable ex ante contract (i.e. pre-standard being set) between Qualcomm and the standard setting organization (SSO). This would have meant there was no need for an investigation into ‘exploitative’ royalties. However, given the apparent naivety of the SSO, it seems quite plausible that royalties were indeed being set too high, but only to the extent of the extra due to incorporation into a standard with apparent deception over the interpretation of FRAND. Unfortunately, because of the way the case was dropped, we may never know what the Commission really felt about this case. But we do know what the Koreans and Japanese felt, because they fined Qualcomm $200m for abusive conduct.

An interesting question is why Rambus should be different (both cases were opened by Brussels in 2007). Rambus has patents embedded in an industry standard DRAM (temporary memory chips used in all PCs) with 95% of the market. However, the key difference seems to have been that Rambus kept very quiet about its IPRs during the standard setting process and only revealed them once the standard was adopted (i.e. ‘patent ambush’). Unsurprisingly, the downstream manufacturers screamed unfair. Thus, the essential principles seem very similar to Qualcomm but with the added layer of deception about the existence of the patents so the SSO could not even think about ex ante royalty agreements. The US FTC found against Rambus but lost on appeal and was disappointed that the Supreme Court refused to take a further appeal. The European Commission also found Rambus in breach of competition law (Article 102 abuse of its dominant position). However, unlike in the USA, we hear confirmed today that Rambus does not fancy its chances in the European Courts. The includes a cap on royalties at less than half their current level, but no fine for abusive conduct.

Time to get back to my original questions: were these two decisions simple tidying up or a clarification of the limits to legal exploitation of IPRs? I think they are a bit of both. A clear line has been drawn against patent ambush. This is good. The artificial extension of patent rights hidden like trolls under a bridge who jump out to collect tolls from unwary travellers is not something that is necessary for incentives to invent and it creates inefficiencies if it results in wariness in user investment and in adopting helpful standards. However, the line is less clear in relation to the narrower patent hold-up issue in Qualcomm. The Commission does not seem to have been able to reach a timely conclusion on this, and it appears to have simply tidied the case up by closing the file.