Monday, December 9, 2013

Android device makers must tread carefully when colluding to bring patent prices down

Last Tuesday an interesting order addressing the intersection of patent and antitrust laws came down in the Northern District of California in a case I didn't watch (hence the delay here). Judge Yvonne Gonzalez Rogers dismissed motions to dismiss that Samsung, Google's Motorola Mobility, HTC and patent aggregator RPX had brought against an (amended) antitrust lawsuit by "patent troll" Cascades Computer Innovation alleging anticompetitive cooperation between those defendants in licensing and purchasing negotiations involving U.S. Patent No. 7,065,750 on a "method and apparatus for preserving precise exceptions in binary translated code", which allegedly optimizes Google's Android mobile operating system.

The standard for a complaint to survive a motion to dismiss is not extremely high: there must be a legal theory, and the alleged facts must support it should they be true, which is a separate question. All well-pleaded allegations of material fact must be taken as true and construed in the light most favorable to the plaintiff, but they don't count if they're merely conclusory, unwarranted deductions of fact, or unreasonable inferences. So the hurdle is low, but there is a hurdle, and while I personally feel that the judge could and perhaps should have granted the motions to dismiss, the important lesson to learn from this is that companies colluding to bring patent prices and license fees down -- in this case, three leading Android device makers (which indirectly includes Google itself because it owns and controls Motorola) -- can be accused of and sued for antitrust violations.

Most antitrust cases addressed individual monopolies, the acquisition of monopoly power by illegal means, or alleged conspiracies for the purpose of inflating prices or keeping them artificially high by leveraging a collective monopoly. There is, however, also the notion of a monopsony: a monopoly on the buying side of a market (again, by one player or collectively by a group). Cascades alleges that the three leading Android device makers, all of them clients of patent pool firm RPX, conspired to monopsonize the market for Cascades' patent licenses at least with respect to the aforementioned '750 patent. Simply put, they told the patent holder that there would be a deal with all of them (through RPX) or none of them. At some point, LG Electronics (which allegedly has a 4% Android market share) and Philips (which is just about to enter the Android market) "broke from the RPX-driven conspiracy [after the filing of this antitrust lawsuit] and independently negotiated settlements with Cascades", says Cascades in its amended complaint.

The court proposes to stay the case (unless Cascades still dissuades it from that idea) pending an infringement proceeding in the Northern District of Illinois since the outcome of the infringement case would provide useful information as to the actual use and validity of the '750 patent. If this case gets stayed, the likelihood of a settlement prior to its resolution increases significantly. So we may never find out who wins the case. It would probably be desirable for the alleged conspirers to work out a settlement just to avoid an adverse ruling on the antitrust side; but they might also believe that this case has now reached a stage at which it's actually necessary for them -- at least for RPX -- to be cleared of any allegations of wrongdoing.

This is not the first time for RPX's business model to face criticism. In May 2011, the Gametime IP blog reported on a letter by security software firm Kaspersky to the FBI accusing RPX of racketeering. Three months ago PandoDaily published a lengthy article on RPX's business model that is anything but consistently flattering. Both these articles mention dealings between RPX and patent trolls. While RPX likes to portray itself as the antithesis to Intellectual Ventures, it interestingly faces the same allegation of divesting patents to trolls with the allegedly intended consequence of convincing third parties that signing up as clients of those firms is a smart choice for avoiding those attacks (though there are still plenty of other patents in the hands of trolls anyway).

I don't mean to engage in RPX-bashing here. I just wanted to point out that this firm portrays itself as a White Knight but doesn't have a clean white shirt in some people's eyes.

Cascades may be a typical troll, but despite the Federal Trade Commission's inquiry into deceptive practices by patent assertion entities, the rule of law (regardless of the political environment) is that it also deserves to be protected by competition law. It's unlikely to get help from antitrust authorities, but it can bring antitrust lawsuits as it did here, and a court of law is not swayed by lobbying: it simply looks at the merits of a case. Here's the order (this post continues below the document with further thoughts and some information concerning the new publishing rhythm of this blog):

13-12-03 Order Denying Motion to Dismiss Cascades Antitrust Complaint by Florian Mueller

Cascades' allegations -- again, at this stage it doesn't matter whether true or not -- are a plausible conspiracy theory to the judge, and the conspiracy part per se also makes sense to me, taking the allegations as true. Allegedly some parties, such as Google's Motorola, told Cascades that they weren't going to take a license (or, alternatively, buy the patent) individually but only as part of a group, through RPX. RPX approached Cascades and allegedly said that key RPX clients wanted "a global solution", and offered a price that it indicated was the best offer its clients would make.

While I can understand that the judge looked at this alleged behavior as well as alleged meetings between RPX and those companies to discuss a joint negotiation strategy as price-fixing by potential buyers, I'm troubled or at least unconvinced by this antitrust theory for mostly three reasons:

  1. It would actually have been the most logical thing in the world for Google (not just its device-maker subsidiary, Motorola Mobility, as part of a group) to negotiate a global Android license on behalf of its ecosystem. The fact that Google didn't do that is not surprising giving its notorious reluctance to take licenses (though it is apparently willing to pay up for software patents asserted against Android through Motorola Mobility, which makes it appear rather hypocritical). But a company that takes better care of its ecosystem would do so, and in that case there would also be a single licensee. If that single licensee told its device makers that they don't need to take a license individually because Google will deal with this problem one way or the other (and indemnify them if necessary), that would not be an antitrust violation. It would be perfectly above board, but it could have a similar effect on Cascades' business opportunity with the Android ecosystem.

  2. Antitrust matters are centered around a particular market definition. Here, the judge basically said that since Cascades alleged that its patent licensing opportunity is specific to Android, that's the market definition based on which to adjudge the motion to dismiss. But this patent is older than Android. It's a general acceleration technique for a bytecode machine. If it's even valid and actually infringed, then it must also be possible to work around this one without a noticeable degradation of product quality because other bytecode machines don't use this technique (otherwise Cascades would have more targets than Android device makers, resulting in a broader market definition that, in turn, diminishes the collective market share of the defendants). If it's easily worked around, then it's not too valuable anyway, raising doubts about whether those Android companies were really trying to bring the price of this patent down to a subcompetitive level or whether Cascades, as the Android camp claims, simply tried to overcharge.

  3. When I read the order I wasn't sure that the judge had fully considered the differences between how patent licensing works compared to how actual products are sold. If the patent is valid and infringed (and let's assume it's not worked around too easily), Cascades can enforce it through infringement proceedings. It can seek injunctions and damages, and damages alone should be enough of a disincentive for someone who refuses to take a license on reasonable terms -- in theory at least. But you can't sue someone over not buying your product (unless a contractual commitment was made).

The bottom line is that I'm more sympathetic to RPX and those Android companies than the court's analysis, and I'd really like to see this adjudged by an appeals court. But the fact that this complaint has now, in its amended form, survived a motion to dismiss -- in a tech-savvy district -- must be considered by companies devising strategies to leverage their collective purchasing power when trying to license or acquire patents. The Android camp has a particular problem because Google (see bullet point #1 above) doesn't negotiate those global licenses on a basis that no one could reasonably allege to be anticompetitive, and because the platform exposes device makers to more infringement allegations than any other.

This blog's publishing rhythm

For more than three years, starting in early October 2010, I tried hard to provide rapid analysis of new developments in the "smartphone patent wars", not only but also (and most visibly) on this blog. I didn't even take a vacation during those first three years and was on alert 52 weeks a year. I have since taken a week off on two occasions, but apart from that I still reacted very quickly to breaking news and frequently dug up court filings before anyone else did, making some people wonder whether I ever sleep or whether I have cloned myself. This level of activity, however, required a degree of availability that precluded me from the pursuit of other opportunities, especially an app development project I believe in.

I have been weighing and exploring different options. Starting this week, I am no longer going to make the same effort as before to be the first, or at least among the first, to analyze and comment on new filings, rulings, and announcements, 24 hours a day, seven days a week, 52 weeks a year. Instead, I am now going to work on smartphone patent (and related antitrust) matters only on some days -- typically, Mondays and Thursdays, though this can vary in certain weeks (and there will be weeks off or weeks such as the week between Christmas and New Year's, where I'll spend only one day on this, not two).

Today I happened to be around when the European Commission published a speech by the EU's competition chief that contained some interesting comment on a likely need for Samsung to improve its settlement proposal. So I commented on it immediately. If the same thing had happened tomorrow, Tuesday, I would have waited until Thursday. If, however, something absolutely huge were to happen (say, an Apple-Samsung settlement or a Federal Circuit ruling on a major case), I would still act upon it immediately, any day of the week including the weekend, any time of the day or night as soon as I hear about it.

The benefit to me is that I know I can not just devote time to my own project on certain days of the week but that I can also concentrate on it without being constantly interrupted (see Paul Graham's blog post on "Maker's Schedule, Manager's Schedule").

I know that this will work fine for my consulting business while allowing me to create an app (which will have nothing to do with patents, by the way). And this blog, which is basically a byproduct of and promotional tool for that consulting business, will remain popular, I'm sure. A couple of years ago this blog mostly got attention when it reacted rapidly. By now, the most popular posts are actually Q&A documents and in-depth analysis that doesn't have to come out on the same day as, or the day after, an important event.

This will be my rhythm for the remainder of the year, and for going into next year. I can't look too far into the future, but this is the plan, so please don't be surprised if my commentary is no longer going to be as instant as it used to be for a few years. I'm confident that you'll still find interesting information here. And on certain days of the week, even breaking news.

If you'd like to be updated on the smartphone patent disputes and other intellectual property matters I cover, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents and Google+.

Share with other professionals via LinkedIn: