The second and final day of Ideas Into Action: Implementing Reform of the Patent System contained morning panels on Non-obviousness, Post-Grant Review (a.k.a. Opposition), and Litigation, while the afternoon was spent primarily on an Industry/Institutional Issues panel. There is a summary by journalist Susan Kuckinskas of internetnews.com. What follows here is merely one biased person's review of a day full of speakers.
As we learned yesterday in Profs. Merges and Mennell's tutorial on Patent Law, patent applications are scrutinized under, among other things, an "obviousness" test. To represent an innovation worthy of receiving a patent, the innovation should be non-obvious. Of course, one of the biggest complaints about patents today is that too many of them seem absurdly obvious. The non-obviousness panel tackled this obvious problem.
Law Prof. Rebecca Eisenberg of the Univ. of Michigan wants to invigorate the non-obviousness standard through peer review of patent applications by "persons having ordinary skill in the art." (That last phrase is in the patent statute and gets pronounced in acronymized form by patent geeks as "phosita." Amaze your friends with that.) Allowing the USPTO to consult with these outside practitioners would have several benefits, says Eisenberg. Patent examiners now tend to focus on the "prior art", that is, everything publicly known before the invention, as illustrated by earlier patents or other published material. Eisenberg points out that phosita have tacit knowledge not represented within prior art such as intuitions about whether equipment is working properly. When they bring these intuitions to bear on a patent application, they are in a better position to tell whether a given invention is "obvious."
If I could interject an idle observer's thought at this point, I couldn't help thinking that if a little peer review is a good thing, then a whole lot of peer review is even better. Why do we expect a single patent examiner to sit down with all the prior art and a patent application and figure out whether the application deserves to be issued? I think it might be because we are smitten with the mythical image of the lone inventor, despite our knowing that innovation inevitably builds on the work of others. Lone inventors churning out impressive work are hard to come by, but it seems like we expect lone patent examiners to be a dime a dozen and to churn out impressive evaluations of purported inventions. Why not consider a collaborative approach to patent examination? What one examiner doesn't think of or find through research, one of the other four assigned to look at a patent might. Or for the truly democratic, apply a Slashdot-like approach to patent examination (or at least opposition) through a public website that takes input from self-selected individuals. Those familiar with Slashdot may be shrieking in horror at the thought, but perhaps a better example is the similar success of GrokLaw, showing that a large group can collaborate on the legal discovery process and produce impressive and money-saving results. The Slashdot crowd regularly finds prior art for the laughable software patents that are occasionally covered there. Why not harness this collaborative research power? Just a thought.
Law Prof. Rochelle Dreyfuss of N.Y.U. read my mind and came up with something more reasonable. She calls it a "thosita" or a team of people having ordinary skill in the art. While Dreyfuss thinks something along these lines might be a good idea, she acknowledges that trying to figure out who the experts are and how they will operate is a total mess. Maybe we know who the six best people are to ask about some technologies, but how does one assemble the team of experts for a "new" rat trap? Keep in mind that the patent office gets all kinds of patent applications. (The largest category is golf equipment!) But in the short time each panelist had, Dreyfuss made a point to focus on something that needed a lot more airtime at this conference. She doesn't like to hear the patent office talking about "user support." Such talk promotes the false idea that the patent office is there to serve the patent applicants rather than the public and the public domain. One brief comment from the end of the day also asked who the "customers" of the patent office are. The implied answer was that if the patent office properly understood its duty, the word "customer" would be a dirty word in its halls. Keep your eyes on the public good.
Todd Dickinson is a former director of the USPTO, though he has more recently been partner in a law firm and is about to take up a position with General Electric. We had heard hints Thursday that the USPTO sometimes has sharp elbows and looks askance at outsiders who would deign to tell it how to operate. Dickinson seemed determined to confirm these rumors by showing he hadn't lost that USPTO contrariness. He said the FTC report (you remember the one condensed from 5,000 pages of transcripts and pretty much the focus of this whole conference?) was "based on anecdotal evidence, not empirical studies." He doesn't like the FTC's recommendation on lowering the burden of proof in patent litigation (FTC Rec. #2) because it would universally devalue intellectual property. (Yup. That's the idea. And?) He doesn't like changes to the obviousness standard, especially not if they are driven by complaints about business-method patents or genomics patents which make up just 3-4% of patents granted. He also points out that the much-maligned business-method patents are only granted 17% of the time, so stop yer whining. He's quick to defend the patent examiners, who contrary to an earlier comment, are not suffering from atrophying skills and knowledge since they left industry, but rather are intimately familiar with the cutting edge of research. (They presumably acquire these cutting-edge skills during those seventeen hours that they devote to each patent application.) And the PTO is, of course, under attack from the judiciary, who far from allowing them to apply general knowledge when examining applications, holds that they cannot even apply common sense. (I'm not sure what citation he had in mind, but I wonder if that wasn't a descriptive, rather than proscriptive, passage. --Joking!)
Law Prof. Emeritus John Barton of Stanford explains that "obviousness to try" used to be (over 40 years ago) a reason why you wouldn't get a patent. We would look at where the state of the art was at the time and ask whether what you tried to do was an obvious next step. If so, no patent. But Judge Rich came along and said, No. there is obviousness-to-try in most inventions.(Application of Tomlinson, 363 F. 2d 928 (1966) at 931.) So nowadays, it is "obvious" how to relate genes to diseases, it simply takes a lot of money and time to accomplish it. But the results can be patented, despite being produced through this mechanical process. The PTO claims it's not obvious what the result would be, so we'll provide patents. A large amount of money is being expended on diagnostic techniques because patents are available, so this is a case where the patent office seems to be achieving its goal of serving the public good. I apparently spaced out a little at this point, because I had to read Susan Kuckinskas' article to learn that Barton thinks we might do better to set the bar a little higher. We could require patent applicants to include information on the routine level of innovation in their fields and only those with true innovations that went beyond the routine would be awarded patents. I was thinking along the same lines, formulating the following question I didn't get an opportunity to ask:
[QuestionBrianDidn'tAsk]Professor Barton, you described a situation where an initial innovation in linking genes to diseases paved the way for a lot of expensive and time-consuming mechanical work to find these links in thousands of other cases. Allowing patents for this mechanical work waters down the obvious-to-try standard, but also serves the important public goal of working towards treatments for disease. Don't we have an alternative to giving away patents here? We could disallow patents for things that are obvious to try in a given field of research, and simultaneously increase government spending on research grants earmarked for that specified area. The research could then still get done, and the public would reap the double benefit of cures for disease and a consistent patent policy. One last concern: our current path of allowing such patents may produce a subtle harm by occupying the time and resources of those knowledgeable in this field with this routine work, when a lack of patents would motivate them to invest their capital elsewhere, where it is might be needed even more. Such true innovations would remain patentable, and in the absence of low-hanging fruit would likely be pursued more vigorously.[/QuestionBrianDidn'tAsk]So while I day-dreamed, Barton apparently said something close to that anyway.
My notes indicate I was still pondering the problem Prof. Barton raised while Ron Laurie of Inflexion Point Strategy wrapped up this panel.
Opposition and Post-Grant Review Panel
This panel started off with a bang as UC Berkeley Prof. of Economics, Bronwyn Hall, couldn't contain a response to Mr. Dickinson's comments in the prior panel. She made the point you love to hear an economist make, when she suggested that Dickinson's claim that just 3-4% of all patents were business-method or genomics patents was misleading if that 3-4% happened to contain extremely valuable or otherwise important patents. Lesson: Don't toy with percentages with economists in the room. They call you on it. Time was short for each panelist, so with what remained Hall provided some numbers on how the patent grant rates at the European Patent Office have started differing from the grant rates at the USPTO. The story behind the numbers? Around 20 years ago, the USPTO and the EPO would come to similar conclusions in similar situations. But since 1983-84 a difference in grant rates between the two offices is splitting open like your uncle's pants after Thanksgiving dinner. Somebody's changed, and it just might be us.
Hall's comments provided a great setup for Dietmar Harhoff, of the University of Munich, to teach us Yanks about the European Patent Office's Opposition system. Dietmar Harhoff had the numbers: 7.9% of approved patents are opposed at the EPO, with an appeal rate of 31.7%. The opposition process takes about 1.9 years, and appeals take 2.1 years. The outcomes: 33.2% of opposed patents are revoked, 32.6% are amended (narrowed), 27.4% come through unscathed, and 6.8% of the oppositions just end due to non-payment of fees or non-opposition. The Opposition process costs between 15,000 and 25,000 Euros, compared to the roughly $800,000 companies spend (on less crucial patents!) in U.S. patent litigation just to get through what is called a Markman hearing. (A Markman hearing is where a U.S. district court judge decides what your patent's claims mean and hence what the scope of your patent is. Almost all patent litigation will settle or be disposed of on a summary judgment motion after this hearing because it typically decides the central issues.) And when compared to fully litigating a patent suit which can run more like $6-8 million, the inexpensive European Opposition process looks really nice. Harhoff points out that NGOs such as Greenpeace participate in patent opposition in Europe because it is so inexpensive. He also makes a critical point that despite his bringing it up again later, seemed to be ignored by the afternoon Industry panel, that is, that we might want to consider allowing only a very short time-frame for the opposition process, around three months, because if you give two companies involved any longer, they will start negotiating and just settle. That may sound good at first, but settlement between two companies does nothing to serve the public purpose of cutting out unworthy patents. We don't want companies colluding to allow low-quality patents through. Rather, the public is best served by an honest opposition process that really investigates the patent-worthiness of the opposed patent.
Robert Blackburn, IP Counsel at pharmaceutical company Chiron, left no doubt about what Chiron wants from the world of patents: predictability. They want to know who owns what in a fast and inexpensive way so that they can negotiate a business arrangement and move on to allocating their capital rationally. They are keenly aware that money spent on litigation is not available for Research & Development. He had several very specific suggestions for improving the opposition process that bump up against this law student's understanding of the world of patent litigation. But since it is easy for these conferences to be heavy on platitudes and light on practical specifics, Blackburn's very clear and specific recommendations were welcome.
UC Berkeley Economics Prof. Joseph Farrell, also head of the Competition Policy Center, a co-sponsor of the conference, is one of these economists, you know the type, that truly and sincerely believes that people make decisions rationally. I feel like some economists must have never dated in college. Anyway, Farrell's primary point is apparently one that can be found in a forthcoming joint paper with Prof. Merges, and amounts to: there are reasons to believe that the incentives to challenge and defend patents are wildly skewed. Farrell suggests that in the average case, the patentee cares much more than the infringers for two reasons: 1) When there are multiple infringers, a validity challenge is a public good among them and so they each can afford to care a bit less and allow their fellow infringers to pick up some of the caring slack. 2) When multiple infringers are not independent, but compete downstream, things are even worse because a successful patent challenge for one will yield lower costs for rivals too and so the incentive for this infringer to challenge the patent is decreased. I kept thinking that it seems on the face of it that this only covers cases where there are multiple infringers. What about the mano y mano showdown? But, alas, each presenter's time was severely limited.
Douglas Norman of Eli Lilly was like Robert Blackburn in having several very specific recommendations for how to run a good patent opposition system. He recommended something that he thought of as in-between the Japanese and European systems. He'd prefer that challenges not be allowed for the entire life of the patent, again echoing Blackburn's desire that this process yield some predictability at some point, and to keep the costs down believes that discovery should be limited. Norman points out that the vast majority of that massive expense of a patent litigation case comes from discovery. (Discovery is the part of litigation prior to trial where lawyers request documents from each other and ask interrogatories, and so on. Some patent cases create truly insane truckloads of documents.)
The question and answer period for this panel featured a highlight of the conference when Jason Shultz of the Electronic Frontier Foundation said that EFF would announce on Monday that it will be getting into the patent-challenging business. He asked the speakers what advice or thoughts they had on what non-profits and open-source projects who lack significant resources to challenge a bad patent can do. The panel seemed a bit dumbfounded and could only muster that the goal of the recommended opposition system would be to greatly lower these costs. I couldn't help thinking that even if we miraculously found a way to be as efficient as the Germans, we would still be looking at 15-25k Euros per challenge and while an open source project could potentially rally the troops to do this once, the problem we are dealing with is that there is an absolute torrent of these absurd software patents flying out the PTO's door. Given a third day of conferencing, this topic would have been panel-worthy.
As you might guess, one obviously starts a Litigation panel with a Law Professor. Mark Janis of The University of Iowa College of Law actually has six years of patent litigation experience. He focused his comments on the FTC's recommendation to change the burden of proof that a plaintiff challenging the validity of a patent has to meet from the higher "clear and convincing evidence" standard to the "preponderance of the evidence" standard. As you can tell, lawyers love words. This distinction is supposed to consist in the following: a preponderance of the evidence is something like a 51% standard. More evidence than not. While the clear and convincing evidence standard is somewhere in the ether between 51% and 100%, maybe something more like 2/3 of the evidence. No matter what definitions you read for these terms though, they are intended to be subjective standards. Currently by placing the higher "clear and convincing" burden of proof on the plaintiff challenging a patent we give the patent a stronger presumption of validity. The FTC's thought was that since so many poor-quality patents seem to get through the examination process, patents don't deserve and shouldn't get this bonus presumption of validity. Rather, a preponderance of the evidence standard would better suit the actual state of things in the world of patents. Janis joined the many in attendance at the conference who are not so taken with this recommendation. He pointed out two functions that the presumption of validity serves. 1) The Expressive Function: The presumption of validity indicates the relative importance that society should place on the decisions made. 2) The Instrumental Function: The presumption of validity appropriately allocates the risk of error between the litigants. Janis' bottom line was that changing the standard might not make a difference and the message it sends takes on an unnecessarily adversarial tone. The idea is that making this suggestion is an admission that the patent system is broken and that rather than repair the damage we're going to tinker with a burden of proof standard that plays a role long after the inevitable mistakes in the patenting process have occurred. We'll learn later that lots of people are not thrilled with this idea for reform.
Duke University Prof. of Law, Arti Rai suggested a friendly amendment to the FTC's burden of proof recommendation. Rai suggests we should have "an asymmetric response" to the PTO. If we implement a robust and inexpensive opposition system then she believes that almost all of the most important patents will be challenged, even the most solid and deserving patents. Given that, we could continue to apply the "clear and convincing" standard for patents that survive the opposition process, and apply the lower "preponderance of the evidence" standard only to those patents that were not challenged in the formal opposition process.
Next Fenwick & West partner (and Boalt grad) Lynn Pasahow thanked the FTC for its recommendation to change the burden of proof standard, which Pasahow explained would increase patent litigation. Patent-litigator jokes aside, he couldn't see how it would play an important role in solving the problems with patent litigation. Pasahow reported that they have done their own empirical studies and found that no juror has ever read a patent front to back. Firms regularly try to have their cases heard in the Eastern District of Texas so that they can get the least educated jury they can find. What the attorneys are allowed to tell the jury is, in Pasahow's view, far more important than the short jury instruction on burden of proof that the judge will give just before the jury deliberates. While acknowledging this is not the case in San Francisco, he says they've found that most jurors have the opinion that "if the government does something, it's probably right." So, this wide-held belief already gives the patent a strong presumption of validity that won't be altered by lowering the official burden of proof standard. As an aside, he suggested that the video jurors are often shown that explains the patent system has more to do with a juror's opinions on the validity of patents than the burden of proof standard. A more balanced video might therefore be a more productive avenue to pursue.
Edward R. Reines, a partner in Weil Gotshal & Manges' Technology Litigation Group made the good point that there are a number of aspects to patents that are rarely, if ever, objected to, such as "best mode" or "inventor's name". Consequently, since there's been no evaluation of these points, it doesn't make sense for at least these things to receive a presumption of validity. I couldn't tell if he would go so far as to advocate a hybrid system such as the one Prof. Rai suggested, but his comments were in that same vein.
UC Berekely Law Prof. Mark Lemley gave a rapid-fire defense of the FTC's recommendation on burden of proof, based mainly on the total inadequacy of the curent PTO to establish validity. Perhaps the best way to convey Lemley's delivery method would be to simply provide my notes nearly un-edited (read this paragraph as fast as you can): Lemley explained that the presumption in the patent office is that you get the patent. The applicant never has to show anything if the examiner can't show prior art. The office is overworked. Examiners get rewarded for first office action and for disposing of the patent. The long complex patents (the ones that get litigated) get less scrutiny because their incentives are set up for quantity. There's a culture in the office aimed at issuing, not denying applications. Continuations are another problem. You can never reject a determined applicant. 85% of all applications result in at least one patent. The PTO is doing what it is supposed to do: A quick once over. Because we've had 17 hours of scrutiny in the PTO we must give that deference!? By the way, Pasahow, it is defendants not plaintiffs that want to go to Texas. So, we never can get a serious look at whether the patent is valid. Consequently, there's no reason to give patents a presumption of validity. It does not take statutory reform to fix this. The courts can do it themselves by just adopting the lower standard. And another thing regarding Section 105 relevancy statements. Todd Dickinson says he got the PTO the power to declare the relevancy of different documents. Lemley says if you really start relying on these relevancy statements then he can win an Inequitable Conduct defense every single time because there will always be something that they left out. So what are attorneys going to do? All you can do is over-disclose. Whammo: PTO's burden goes up! Willfulness: the FTC wants this changed. Lawyers tell their clients: Whatever you do, don't read patents. The FTC's idea is a necessary but not a sufficient fix for the willfulness problem. (Whew!)
James Pooley, partner at Milbank, Tweed, Hadley & McCloy let us in on a secret. He helped write the script for that patent video Pasahow mentioned! He explains that they received just as many complaints that it was biased towards those challenging patents as they received from those defending patents. Pooley echoed what we heard from Chiron and Eli Lilly as he explained that the problem in the patent system is in the enforcement system. Businesses are petrified by the unpredictability of the process and that is what motivates them to pay these dizzying sums for litigation. They want certainty/resolution. Pooley participated in the we-promise-to-release-it-on-Monday NAS report and hints that it may include a recommendation to do away with the inequitable conduct defense. Lemley chimes in that this worries him because he wants people to know that lying will be punished, so such a recommendation has to be carefully crafted. We'll see on Monday.
Next came lunch and a short talk that I missed by Judge Ronald Whyte of the Northern District of California. (Judges aren't generally allowed to take positions on reforms.)
Industry/Institutional Issues Panel
This panel was sub-titled "12 White Guys and the Dude from Google." All joking aside, this was an issue that several participants and I could not help discussing after the conference. The conference organizers obviously can't control who CISCO, Microsoft, Genentech, Symantec, Intel, lobbying groups, etc. etc. send to this panel, but it provided a fairly frightening look into hiring/promotion practices in the industry. In the post-bubble economy women and people of color apparently do not get the jobs that allow you to take Friday off to go to a conference. Or perhaps we should just hope it was a bizarre anomaly.
I came in as Bart Eppenauer of Microsoft was finishing up. He seemed to be agreeing with Prof. Lemley's earlier remarks that the willfulness issue in patent litigation deserves scrutiny. Basically, if one willfully infringes a patent, the patentee can receive treble (that's lawyer-talk for "triple") damages. The problem is with how we currently define willful infringement. The unclear standard leads to numerous problems, such as the one mentioned earlier where attorneys believe the best thing for their clients is that they not read patents. But given that innovation inevitably builds on the past, we're locking away the potential of a world of innovative ideas. Bummer.
Google attorney, Kulpreet Rana, made it clear that patent litigation sucks. You got the distinct impression from hearing Rana talk that Google would much prefer to spend time and money on innovating technologies rather than responding to DMCA take-down notices from Scientologists or investigating the worthiness of somebody's patents. I was thinking: When the leading innovators in business, from pharmaceutical companies to Silicon Valley survivors all tell you that our Exclusive Rights Regime is broken, why do we keep getting laws in Congress that just push us further into the abyss?
Michael Shallop of Symantec Corp. explained that they make software on a rapid development cycle. There is very incremental innovation and a short lifespan for their products. Nonetheless, they find that they have to play the patent game in order to defend themselves. It seemed like they too would prefer to just work on keeping your computers free of virii.
David Simon of Intel (read his testimony before the House Subcommittee on Courts, the Internet, and IP—it covers similar concerns) joined the chorus of businesses who hate the uncertainty that patent litigation presents. I fell out of my chair when I heard the size of his patent filing budget compared to Intel's research budget, and he said that he wasn't even counting the litigation budget. Simon presented the picture that an inordinate amount of time is being spent on these issues every year, siphoning away funds that could be going into giving your laptop a 40GHz processor that won't set fire to your pants. (OK, those are my words, but you get the idea.)
Then the panel moderator, Carl Shapiro, of the Haas School of Business, UC Berkeley ran through almost all of the FTC's ten proposals (find 'em in the executive summary) and solicited reactions. (Many of the below are how I would refer to the different FTC recommendations, not their actual titles!)
FTC Commissioner Mozelle Thompson then wrapped up the conference by announcing that a group of companies (basically the companies present) had committed to further meetings on this subject.
My overall impression: In a day and a half, you could hardly have a better Patent Reform Conference. But if we had gone to two full days or a third day and had worked in more public interest issues and gotten our hands a little dirtier tackling the hardest problems, we might have done even better. Hopefully the numerous upcoming reports and continuing meetings on this issue will fine tune what occurred here. I do believe this conference pushed the thinking forward on a difficult subject.Posted by Brian W. Carver at April 17, 2004 01:56 AM | TrackBack