再Standard procedure “reasonable patent royalties”, May 11th 2013 P76. (標準規格特許使用料)
(If companies cannot agree on “reasonable” patent royalties, courts must decide. How?) Almost by definition, people take standards for granted. Their computers connect to the internet via Wi-Fi from home, the office or a coffee shop. They chat happily to friends with different mobile phones on different networks. Somehow, devices and components made by dozens of different companies have to be compatible with each other - and no one must see the join. To make this happen, bodies such as the Institute of Electrical and Electronics Engineers and the international Telecommunications Union painstakingly draw up complex technical standards with which whole industries comply. Standards benefit both consumers (by reducing the cost of switching from one product to another, say) and producers (by allowing greater economies of scale). But they also have a drawback. Once a standard is set, companies making products that meet it have little choice but to use the technology essential to it, which may be covered by patents. In theory, owners of standard-essential patents (SEPS) are thus in a powerful position. They can demand a large royalty, far more than they could have done without the standard. This problem, known as “hold-up”, can be compounded by “royalty-stacking” if lots of companies own SEPS (and lots do：for instance, dozens of them boast more than 2,500 SEPS for H.264, a standard for video). Multiple fees can push up the cost and prices, and limit the standard's use. In practice, standard-setters are wise to this. They expect companies that claim SEPS to license their technology on terms that are “reasonable and non-discriminatory” (RAND, or FRAND, with an otiose F for “fair”). Most of the time this works pretty well. Companies often charge only a few cents per unit of products that rely on their technology. They exchange licences, so they can use each other's know-how；small net payments may change hands. Sometimes they charge nothing at all. However, standard-setters do not define RAND, and in the heat of the information-technology industry's patent wars the number of disputes over what it means has been rising. On May 6th the European Commission said that it suspected Motorola Mobility (bought by Google for $12.5 billion in 2011 largely for its patent portfolio) of abusing its market power by asking a German court to prevent Apple from using know-how covered by a SEP. It is also examining a complaint by Microsoft against Motorola and is investigating Samsung on similar grounds. Jorge Contreras, of the Washington College of Law at American University, notes that whereas only two RAND disputes went to court in the United States in the 1990s, 17 have done so since 2008. Most of these have ended in a settlement. But last month - for the first time - a court estimated a RAND royalty. James Robart, a federal judge in Seattle, put a figure on what Microsoft might reasonably pay Motorola Mobility for Motorola's contributions to H.264 and to 802.11, which governs Wi-Fi. The firm had asked for 2.25% of the price of any product using either, claiming that this was its usual fee. Microsoft sued, arguing that Motorola's demand was unreasonable. A jury is still due to decide on that, but Judge Robert did some sums to help it. More important, he laid out a method. Suppose, he said, that the two sides negotiated in good faith：what might the result be. As a guide, he took a judgment in a patent-infringement case from 197, Georgia-Pacific v United States Plywood. This sets out 15 factors that might be used to calculate reasonable royalties (they have been applied in several disputes, but not until now to SEPS). The list includes royalties already being paid for the patent at issue and for similar patents；the holder's normal licensing policy；and the technology's value to those using it, including licensees. To fit the Georgia-Pacific factors to SEPS, Judge Robart made several amendments to the list. His hypothetical negotiators would not simply look at Motorola's patents in isolation. They would consider RAND royalties already being paid for other SEPS, thus limiting the risk of royalty-stacking. And a reasonable royalty would not be inflated by the existence of the standard. Microsoft should pay only the value of what Motorola had contributed. This limits the risk of hold-up. Once the judge got his calculator out, the results were embarrassing for Motorola, largely because he found that its technology was not of great importance to the standards or to Microsoft. Its H.264 patents, for instance, covered “interlaced” video, which these days is old hat. Its own experts struggled to find examples of its use. Judge Robart proposed that Microsoft pay just 0.555 cents per unit for the H.264 patents, to be applied to its Xbox game console and Windows operating system；and 3.471 cents for the 802.11 patents (for Xbox) or 0.8 cents (other products). He also set a range for each：0.555-16.389 cents for H.264 and 0.8-19.5 cents for 802.11. Motorola's 2.25% was equivalent to a few dollars per unit. Although his calculations look rough and ready, Judge Robart was “lucky” says Mr Contreras, to have as much to go on as he did. The rate for H.264 was based on the royalties in a large patent “pool” , of which Motorola was not a member. (Members of patent pools agree to license their SEPS to their parties and each other in a single package.) For 802.11, which has a much smaller pool, he also drew on a consultants' report commissioned by Motorola in 2003. And there is no reason to think that other SEP-holders will come off as badly as Motorola did：another court may find that another set of SEPS is worth much more. But warring tech companies at least have some idea of what an American judge might think RAND means. That may make them even keener to work it out for themselves before they reach the courtroom.
