I'm at PatCon 3 at Chicago-Kent (follow live tweets at #PatCon3), where so far we have heard Judge Linn expressing skepticism about the Supreme Court's role in patent law, a panel of in-house counsel describing their views on the patent system's biggest problems, and a plenary session with Mark Lemley (arguing against jury trials for patent validity), James Malackowski (on IP markets), and David Abrams (on patent value and citations). Below are (very) rough notes on one of the highlights of the day, a debate between Judge Posner and Professor Epstein:
Posner: I have a longstanding academic interest in patents and have volunteered to hear many patent cases as a district judge (though I haven't tried a patent case to a jury, which is disappointing, since I would love to see if a jury can understand a patent case). I think we all agree that patents should not be granted unless there is a significant invention, which, unless patented, would not be developed at all or would be developed much later. If that condition isn’t satisfied, it is a mistake to grant a patent. Patent is a monopoly that causes economic distortions. Beyond that, administration of patent system is very costly; patent is unlike standard property right. Pharmaceutical system is an exception. Most often, the inventor doesn't need patent protection because invention is incremental and not costly and the costs of duplication are high due to learning curve. And there are lots of problems with patent litigation: jurors like inventors and dislike copyists; patent trolls (I like the patent trolls, they're buccaneers!); defensive abuse of patents by accumulation of large inventory; patent races. I have decided that in every jury trial I will appoint neutral experts that jury can feel comfortable relying on. Other issue is jury instructions: need to get rid of patent jargon.
Epstein: I've never tried a case or argued a case; perspective matters a great deal. Litigation is the tip of the iceberg, and the essential portions of the system are outside litigation. There is a fair bit of agreement between the two of us: no one thinks patent system is easier to operate than system for land, though land systems are more difficult than you would suppose. How you design the property system matters. Patentable subject matter can be used to make categorical exclusions; obviousness can be used to weed out patents that would have been invented without patents. But don't go too far: availability of first-mover advantage, brand-name status, etc. is argument for patents, not against—patent system is available for those inventions that need it. Improvement patents such as pharmaceutical follow-on patents have been shown to improve consumer surplus. Software patent thicket can occur but is severely overrated: new patents can be substitutes for others and can thin out the thicket by rendering others obsolete, and there is no evidence that the number of new patents has diminished, as you would expect if a patent blockade were occurring. When I think about patent system reform, I think about ways in which the patent system has become too lax in its enforcement of property rights. Stronger injunctions (which need not be absolute) will reduce damages actions and reduce pressure on administrative side of the system. Without strong injunctions, people will flee the licensing system and dare somebody to sue them. If you can't prove advantage of new system, stick with the old system.
Posner: I'll just mention one point where we disagree. Professor Epstein said that in a situation where it is possible to protect your invention without a patent, companies will not apply for a patent. I don't think that's right: a patent is a way to protect your property right, but it's also a competitive weapon.
Epstein: If you're talking about patents of sufficiently low value that you can protect them without patents, they are unlikely to be stumbling blocks to others.
Posner: That's not true. To grant a patent where you don't need the patent in order to create the incentives is to create property rights without social function.
Epstein: No, and in many cases, the ideal way to defend something is not to patent it, which is expensive, but to publish it and put it in the prior art.
David Abrams: For Epstein, what about differences in term by technology? For Posner, why don't other district judges use neutral experts?
Epstein: Expected utility for a software patent is relatively short, so the longer you go the greater it is defensive use not offensive use, but defensive use depreciates relatively quickly as well. The one area where we do need longer patents is pharmaceuticals.
Posner: I will appoint neutral experts in any jury case. I cannot envisage what a jury does when it has competing expert witnesses.
Ted Sichelman: Judge Posner, you glossed over a large debate in the literature about pre-invention and post-invention activity. What role should this distinction play?
Posner: To the extent post-invention activities are essential to get the product to market, it strengthens the case for patent protection (and for elongating patent protection).
Epstein: I have a slightly different view. The real difficulty is figuring out whether you want to enter the race when you don't know when the race starts, people try to conceal their activities. All things considered, I would make patents slightly longer to get people to enter the system sooner.
Mark Lemley: Recent data by Colleen Chien suggests that over half of all patent lawsuits filed in 2012 were filed by patent-assertion entities and that the suits overwhelmingly end up getting resolved for relatively small amounts of money, a substantial chunk of which goes to pay lawyers' fees. How do you both feel about fee-shifting in cases that seem weak or seem designed to be targeted to extract litigation costs?
Epstein: I think the German system gets it right in terms of incentives.
Posner: Problem with fee-shifting in general is that the incentive effects are ambiguous. One effect is that if you think you have a good chance of winning, you'll actually win more, so if the trolls select their cases carefully, it may make trolldom even more attractive.
Epstein: Social benefit of assertion entities is in maintaining integrity of their voluntary licenses, so we want to know how these suits compare to the total license fees they gather
Chris Buccafusco: I'd like to hear you reflect on remedies in the patent context. Difficulty of calculating damages, benefits and costs of triple damages, etc.
Posner: I think there are serious problems both with damage remedies and with equitable ones. In damage remedies, many patents are on components, not the whole device, and valuing a component is hard. But an injunction can be a disaster. If you infringed a component and now you're enjoined, even if component is very unimportant, you have to throw away huge inventory.
Epstein: No question that it's an important difficulty, but I would put a more optimistic spin. If one component turns out to be infringing, it is an advantage to have component, only gets small royalty.
Colleen Chien: A couple of weeks ago, the House Judiciary Committee held hearings on tort and patents. In tort testimony, small business trade group testified about certain person receiving hundreds of settlements of thousand dollars or more. In patent context, we've recently seen PAE using scanner patents to go after small and large companies for $1000 per employee. What can patents learn from tort?
Epstein: I don't think you can learn from tort system; you rarely have remedial set of choices. Future flow and sale stuff is not relevant.
Posner: What you learn is that litigation is Darwinian. These companies and their lawyers are trying to kill each other. There is a lot of socially wasteful activity.