The Stanford Law professor and board member at Lex Machina discussed his thoughts on intellectual property trends and patent-reform measures.

September 01, 2016

Mark A. Lemley is an intellectual property thought leader who wears many hats. Mark is the William H. Neukom Professor of Law at Stanford Law School, a founding partner of the Durie Tangri LLP law firm, and co-founder and Board of Directors member of the data analytics company Lex Machina. Mark has authored numerous books, law review articles, and amicus briefs. He has argued many federal appellate cases and has won numerous awards for his work in the field of intellectual property law. Mark recently joined Stout Managing Directors Michele Riley and David Haas to discuss his thoughts on various issues regarding intellectual property trends and the current state of affairs in U.S. courts and Congress involving patent reform measures.

STOUT: Where did you grow up?

LEMLEY: I grew up in St. Louis, Missouri.

STOUT: When you were growing up, were there any signals that would have predicted your interest in intellectual property law?

LEMLEY: I did debate in high school and college, so I did like to argue. I had some interest in astrophysics and astronomy when I was growing up, and I was also very interested in politics. I actually went to the presidential nominating convention my senior year in high school in Missouri to speak for a candidate — Gary Hart.

STOUT: You’ve had quite a long and diverse set of experiences and accomplishments, particularly considering that you’re not yet 50 years old. You have taught law, and have practiced as an attorney. You’ve authored books, law review articles, and amicus briefs. You co-founded the data analytics company Lex Machina and clerked for a Ninth Circuit Court of Appeals judge. Which of those activities have you enjoyed the most and why?

LEMLEY: You know, it’s tough. I like the mix of all these activities, but I would have to say my first love is teaching. I have been a law professor for 22 years, and I love both sharing knowledge with people and also discovering new things; teaching law is an opportunity to do both. I enjoy practicing law very much as well. Practicing law one day a week is probably about the perfect amount of time to practice law, because you can do the interesting bits but not the annoying bits.

STOUT: How are today’s students different from the students of 20 years ago?

LEMLEY: The students, especially since the legal recession, are much more sophisticated about the world outside of law school, and about what it means to be a lawyer. Law school is less a default path for people who don’t know what they want to do with their life than it used to be, and that’s probably a good thing. Students today are coming to law school with their eyes open about what a law degree will or will not get them.

STOUT: What future activities do you have on your IP bucket list?

LEMLEY: I think there are a number of interesting legal issues around autonomous vehicles and artificial intelligence that are worth looking into. But I expect my teaching job will be largely the same 10 years from now as it is today, except that the topics I am focused on are probably going to be different.

In terms of practicing law, I’ve had a very successful career. I really enjoy working with my colleagues here at Durie Tangri. I’ve gotten to prepare quite a number of appellate arguments, and I would like to argue a case in the U.S. Supreme Court at some point, so that’s probably a bucket list item.

STOUT: As a legal scholar, do you have a view on IP law in China and how it is evolving?

LEMLEY: It’s been very interesting to watch. I think 20 years ago when I was starting out in teaching law, the perceived wisdom on China was that it was a pirate nation. They copy stuff, they counterfeit stuff, and IP law between the U.S. and China was a matter of the U.S. trying to coerce China to stop taking our stuff.

But one of the things we’ve seen in the last five or 10 years, especially on the patent side but also increasingly on the copyright side, is that China has decided that they want to be an innovator nation. They have embraced patenting to a really surprising degree, to the extent that there are more patents now issued in China every year than anywhere else in the world. And I think we’re starting to see a turnaround in terms of attitudes about intellectual property in the Chinese government and in the Chinese business community.

I do think there’s still a significant concern about how fair a shake foreign companies get against Chinese companies in the Chinese legal system. But I can imagine that in the not-too-distant future we’re going to be thinking of Chinese companies as patent plaintiffs in China and elsewhere in a way that wouldn’t have seemed even remotely plausible 20 years ago.

STOUT: There’s been quite a bit of legislative and judicial activity in the IP world over the past several years, from the America Invents Act and associated inter partes and post grant reviews to the new Defend Trade Secrets Act to various rulings regarding patent eligibility, damages, and standard essential patents. Based on this activity, do you think Congress and the courts are heading in the right direction?

LEMLEY: I think they have been heading in the right direction, and the question now is what happens next? If you look back at the long arc of IP history, you see a whole bunch of pendulum swings, and we often tend to cycle between over-protection and under-protection, and I think in the two decades after the creation of the Federal Circuit through about the mid-2000s, we kept strengthening patent protection more and more; I think we arguably went too far.

Many of the corrections we’ve made in the last decade have been necessary to try to get things back on the right track. The question now becomes, do we settle into a happy medium or do we go too far in the other direction?

I think most of the things Congress and the courts have done in the last 10 years to stabilize the patent system and cut back on some of the excesses have been good things. Not all, but most. I do worry that we might continue blindly toward an anti-patent mindset, which could be just as destructive as too strong a patent protection mindset.

STOUT: With where things stand right now, which IP law policies, practices, or interpretations still need fixing and how would you suggest fixing them?

LEMLEY: I think Congress could still do something about venue reform. We’re now in a situation where half of the patent lawsuits in the country are filed in the Eastern District of Texas — and the Eastern District of Texas ends up deciding cases in a way that’s at odds with much of the rest of the country. I think we’d be better served by a system that spreads the cases across a wider variety of districts where there are real technology businesses as well. I know there’s an effort by John Duffy to challenge the venue rules in the U.S. Supreme Court, and there is a bill pending in Congress that would change the venue rules. That might be the single thing Congress could do that would most improve the patent litigation system.

In terms of courts, obviously there are good decisions and there are bad decisions. I think the biggest wild card out there right now in the courts is 101.

I am not, as a general matter, a fan of 101 as the be-all and end-all of patent validity litigation. I think we’re starting to see indications that the Federal Circuit might navigate a pretty reasonable road. The Federal Circuit seems to be doing a pretty good job of distinguishing between cases that look like they involve real technology and survive the 101 threshold, and cases that are really about taking an offline concept and just saying do this in a computer or do this on the Internet, which are the ones that are being invalidated.

STOUT: What about establishing special courts like the current pilot program for dealing with patent litigation issues? Do you see that as a possible solution to some of the problems regarding consistency in deciding patent litigation?

LEMLEY: I think it can help. The experience has been very different in different districts in part based on how many judges don’t want patent cases and pass them along. If you look at the Central District of California where there are a few judges who signed on to the patent pilot project and lots of judges who are happy to give up their patent cases, we really do see a form of specialization within the district.

By contrast, in the Northern District of California, there are a number of patent pilot judges, but even the non-patent pilot judges like their patent cases and have mostly held on to them, so there’s not really much difference between the two. And I think the project might be a kind of nice compromise between a truly specialized patent court on the one hand and random assignment to district court judges across the country on the other. It means that you know the judge who’s going to hear your patent case is someone who actually wants to hear it and has some expertise, but it’s not like the Eastern District of Texas in which you can basically pick your judge by deciding where in particular in the Eastern District of Texas to file your lawsuit.

STOUT: You recently filed a brief with the U.S. Supreme Court regarding the question of design patent damages in the Apple v. Samsung matter. What are the main issues in this case that you’d like to see the Supreme Court address?

LEMLEY: There were two issues in that case. I wish the Supreme Court had taken the functionality question, but it didn’t.

The Federal Circuit had both trade dress and design patent issues in front of it and it concluded that the very same design elements in the Apple phones and iPads were functional under trade dress law but were not functional under design patent law. And that’s an indication that functionality as a doctrine is a mess in IP generally. So I wish the Supreme Court had taken that question on, but it didn’t.

The other issue is the damages piece. I think the damages regime for design patents makes no sense, because it says if you infringe a design patent, your entire profits from the sale of the product go to the design patent owner, regardless of anything else that went into the product, whether it is other patents, your own contributions, et cetera.

And that’s not the way we do it in any other area of IP law. So I think the Supreme Court will look at that and they will say, boy, this doesn’t make a lot of sense. The question is whether it will conclude that it is required by the statute or whether there’s a way to read the statute to allow some sort of apportionment or at least a limiting to the circumstances where the design really is the basis for the demand.

STOUT: If the Supreme Court did that, it would be more consistent with the world of utility patents where there presently is a requirement, at least for reasonable royalty purposes, that the invention covered by the patent was the basis of demand for whatever components are to be included in the royalty base, correct?

LEMLEY: Yes, exactly. What we do in utility patent cases is engage in some kind of apportionment, because we want to know what this patent contributed to the value of this product, as distinguished from the other contributors. And the only exception we have is a pretty narrow exception in which I can say, you know what, my technology is so important that it is the reason people bought the product in the first place, and I think doing something similar in design patents would make a lot of sense.

STOUT: What suggestions do you have as to how litigants and the courts can best measure the value of a single element of a multi-component product?

LEMLEY: This is a difficult problem and it’s something that experts in the utility patent cases have been dealing with for a long time. But I think there are ways you can structure the measurement.

Daralyn Durie and I did a paper regarding a structured approach to reasonable royalties a couple of years back. And I think you start by asking, what does the “but for” world look like in a market where this invention doesn’t exist? What does the product look like? How much more or less attractive is it? How much more expensive is it to manufacture?

Once we have captured the value of the technology, we then ask, what else besides this patent needs to go into making that technology? Are there other patents you’ve got to license? Are there other expenditures you have to make? I think narrowing it down that way is the right conceptual way to do it.

People have come up with some relatively innovative ways to try to get to the answer to that question, and a good form of conjoint analysis can sometimes get you where you want to be, because we’re trying to offer people different menu options and seeing how they value different pieces in combination with each other. But it is an art, not a science question.

STOUT: Please describe how you decided to start Lex Machina.

LEMLEY: This started as an academic project for me in 2006, when patent reform was just starting to rear its head in Congress. At that time, we were hearing all of these wild statistics being thrown about — patent trolls are 80% of all lawsuits. No, they’re only 2% of all lawsuits. And I thought, well, you know what, this is an answerable question.

My Stanford colleague Joe Grundfest had created something called the Securities Class Action Clearinghouse where he collected information on every securities class action lawsuit that was filed. And I thought we could do something like that for patent cases, and then run the numbers and see how many people won cases and find out what other results we could learn from the data.

It turned out that things were more complicated. There were a lot more patent cases than securities class action cases, and so I brought in Josh Walker to help put the project together. We quickly realized this wasn’t something you could do by just having some law students code all of the results in the cases. We needed to try to automate as much of the process as possible.

Pretty soon here at the Stanford Law School we had a bunch of world-class computer software experts in machine learning and artificial intelligence trying to figure out issues that are pretty hard for computers to answer, but important to automate, like finding all of the substantive rulings on summary judgment but excluding the ones that just set a hearing date or something of that nature.

And so we wrote some pretty sophisticated programs that automated the collection of this type of data. And that, in turn, took money. We raised money here at Stanford to start the project and we got a prototype together, but it became clear to us that it would take more money to keep it going and to make it everything it could be, and also that there was a real demand for what we had.

Both law firms and companies wanted access to the raw data itself. They could use it to benchmark themselves against their competitors.

Ultimately, in 2010, we decided that we would spin the entity out from Stanford as a private company called Lex Machina. It was the first, although not last, of the legal technology startups that have spun out of Stanford Law School in the last several years.

I like to say you know you’re in Silicon Valley when the law students are all starting companies. We sold Lex Machina to Lexis Nexis in December of last year, and they are running it.

STOUT: Have you noticed any change in the goals of Lex Machina since it was acquired by Lexis Nexis from what your original goals were for the company?

LEMLEY: One of the things that Lex Machina really was aiming to do, and one of the things that the Lexis acquisition will enable, is to move beyond patents into other areas where you could do the same kind of data analytics and provide people with a lot of advantages, like trademark and securities law, as well as other fields of federal law.

STOUT: Do you have any insight into the most interesting or unexpected findings?

LEMLEY: There have been a number of interesting results. One of the things that surprised me is that patentees lose about three cases in four. And that is true today in a climate in which courts have been cutting back on the strength of patent protection, but it was also true in the early 2000s when patent protection was at its height. That finding was really interesting to me. I’m working on a paper now that will be a little bit of a broader picture on kind of how the patent system rebounds and is resilient to change, and no matter what, one of the issues in the system for patent owners is there are a lot of legal issues and you have to win pretty much all of them, or you lose.

STOUT: With inter partes and post grant reviews, it seems like defendants are given even more of a chance for a second bite at the apple on some of these issues.

LEMLEY: Yes, and what’s interesting there is what happens in court when the patent survives the IPR. Because you could imagine that, as a practical matter, if you survive the IPR on validity, courts are going to be reluctant to invalidate the patent even if it is on separate legal grounds. And so it might end up providing equilibrium to the system. We’ll see. It’s an interesting question.

STOUT: Looking into your crystal ball, which industries or market applications are you seeing as most likely to face future intellectual property challenges and what will those challenges be?

LEMLEY: One is definitely 3D printing. I did some writing on this a year or two ago, and I feel like the 3D printing technology is poised to do for patents and design patents what the Internet did for copyrights, which is that it will separate the manufacturing from the design and put the design in a digital file that could be anywhere, which decentralizes the manufacturing. The effect of that on the Internet has been to make it much harder to enforce copyrights in bulk because now, instead of a few counterfeiters who have to run their own manufacturing and distribution networks, you’ve got people all over the world who can engage in small-scale acts of infringement.

I think we’ll increasingly see something around the manufacture of consumer products, but I also think it’s the manufacture of tools and other inputs into commercial products that, instead of just kind of all being manufactured in one place that’s easy to monitor, we’re going to see a shift to made to order on-site. And in that environment it is just much harder to detect and stop infringement.

The other issues I see from where I sit are the legal questions around virtual reality and augmented reality.

There are a bunch of interesting legal issues. They may not be primarily intellectual property issues, but there are a bunch of legal issues around how people intersect with the technology, how they view it, and so forth.

STOUT: We have a few questions for you on some of your writings on patent damages and valuation issues. First, is there any place for citation analysis in assessing the strength or value of a patent?

LEMLEY: I think citation analysis can tell us some things, but it’s definitely not a be-all and end-all in terms of identifying valuable patents. One of the things that it tells us is not so much how important or revolutionary an invention is, but rather whether the invention is in a crowded field where lots of people are working, which might make the patent valuable in a business sense, but it doesn’t mean that the technology is necessarily intrinsically more valuable.

I think it’s something that we might want to pay attention to, particularly if we feel the need to distinguish among different patents, all of which claim to cover or be essential to the same technology.

A patent pool, for instance, might be in a position in which 50 people show up and say, hey, I’ve got a patent, and you have to use it to practice this invention. In that instance, we might want some way to distinguish between a patent that covers a necessary feature, but it’s a small necessary feature, or it is somehow less important.

Citation analysis might be a way to do that, but I’m a little nervous, just based on what I’ve seen, about saying there’s a strong correlation here between forward citation counts and how important the technology is.

STOUT: What other methods do you think would yield better indications of strength or patent value?

LEMLEY: I feel like the best we can do is say, how is the world better because of the technology than if we had the same product without the technology? And to answer that question, you can, of course, look to other existing licensing transactions and so forth, although there’s a certain amount of circularity to that.

STOUT: In a few short words, can you explain the simple approach you advocate for setting reasonable royalties for standard essential patents?

LEMLEY: I would say a FRAND commitment is binding on patentees and anyone who’s willing to pay the reasonable royalty gets a license. If we can’t figure out the right price, we should arbitrate that price.

STOUT: Then who should be involved in the arbitration?

LEMLEY: I think the standard-setting organizations could certainly set it up. It would be reasonable for an organization to basically either contract with an arbitration group or have some arbitrators who know something about standard essential patents.

And I think the key is that we can eliminate all of this collateral litigation around standard essential patents. Patent owners want to avoid the FRAND commitment and go to court for patent infringement. The potential beneficiaries of the commitment want to go to court and sue for breach of contract, but they want to do so without committing to actually pay the reasonable royalty. Or they might want to go to court and file an antitrust lawsuit.

I think you can avoid a lot of those unnecessary complexities if you just say, you know what, when I make a FRAND promise, I am promising to license these patents on reasonable terms to anybody who’s willing to pay a reasonable term. If you’re not willing to pay a reasonable term, I don’t think you should get the benefit of the FRAND promise. So then the only question that should remain is, well, what’s a reasonable term? And if we can’t agree on that, we ought to resolve it, but I think you could resolve that in arbitration. You shouldn’t have to go to court and make it a breach of contract claim or an antitrust claim or a patent infringement claim.

STOUT: What makes a patent valuable?

LEMLEY: I think there are two ways to answer that question. One answer to the question of what makes a patent valuable is really the same answer as what makes a technology valuable. It’s valuable if it improves the world in some measurable way. It allows you to produce something people like or it allows you to produce the same thing at a lower cost or with less risk of error or something of that nature.

And so that’s sometimes an easy answer, but I think one of the difficulties that we have in the patent system is that the value of a patent and the value of the technology don’t necessarily always coincide. In some sense, the value of the patent is what people are willing to pay for it. And what people are willing to pay for it may be a function of how much I am going to lose if I get sued and have to go to court. But I also think a lot of the patent marketplace isn’t driven by saying, oh, take a license or I’m going to sue you and here’s how much money I’ll get.

There’s a heck of a lot of patent transactions that go on where the parties don’t seem particularly concerned about is this patent valid or infringed? How broad is its scope? What are the damages? Nobody goes through that analysis in a lot of the deals. We’re treating the patents as kind of a currency that has value and we’re using it as the basis for our transactions. But for that reason it’s really hard to say that it’s got an intrinsic value. It’s got value that people are willing to put on it.