Mark A. Lemley (BA ’88), William H. Neukom Professor of Law at Stanford Law School
If invention is the engine of the U.S. economy, driving the start of new companies—then intellectual property law is the fuel, helping to spur or impede the innovation that a well-designed patent system should produce. And at the center of it all is Mark Lemley.
An expert on patent, trade-secret, antitrust, and constitutional law matters, he is one of the true thought leaders in the field of intellectual property, with the vantage point of multiple worlds. He is a practicing IP attorney and a founding partner of Durie Tangri LLP, one of the nation’s top IP firms, where he has represented Comcast, Genentech, Google, Grokster, Impax, Intel, Netflix, Palm, and TiVo, to name a few clients. He is also a legal scholar and teacher, inspiring the next generation of IP lawyers and publishing some of the most important papers on the subject. But prolific doesn’t quite describe him—Lemley is a scholarship machine. Since joining the legal academy in 1993, first at Berkeley Law School, where he earned his JD in 1991, and then here at Stanford Law, where he joined the faculty in 2004, he has published some 137 articles and seven books. According to Lemley, “a decent amount of inspiration comes from disconnects you see between different worlds,” the perspectives offering friction to spark yet more questions and debates.
That creative tension has led Lemley to become an innovator himself. Frustrated by what he calls the “anecdata” and “weather prognostication” of IP law, with various competing claims about IP data batted around, in 2006 he introduced some metrics to the field and launched the Stanford Intellectual Property Litigation Clearinghouse. A now massive database, the academic research project captured every electronically available patent litigation event and outcome since 2000, with the aim of bringing openness and transparency to IP law. In 2010, the project was spun out into a legal-tech company and renamed Lex Machina. Today, the company is in the middle of the big data boom that is aiding IP lawyers by giving them access to the information they need to do their jobs well. And the project is still faithful to its original goal—providing the information for free to academics and policymakers.
While firmly grounded in Silicon Valley and the local tech world, Lemley is known nationally and globally too. His works have been reprinted throughout the world and translated into Chinese, Japanese and Korean as well as Danish, Italian and Spanish. And his expertise is sought both inside and outside of the Stanford Law classroom. He has taught intellectual property law to federal and state judges at many ABA and Federal Judicial Center programs and has testified on multiple occasions before Congress, the California Legislature, the Federal Trade Commission, and the Antitrust Modernization Commission. He has also filed numerous amicus briefs before the U.S. Supreme Court.
While much has been written lately about the troubles in the U.S. IP system, with complaints of an overgrown bureaucracy and corporate opportunism leading to its downfall, Lemley is optimistic. “We’re still the most innovative country around, with tremendous entrepreneurialism. So it’s far from doom and gloom.”
A. Douglas Melamed, Herman Phleger Visiting Professor
Douglas Melamed also has his feet firmly placed in both the scholarly and practice worlds, having offered groundbreaking legal analysis in each. Over the course of a career spanning more than 40 years, Melamed has been at the center of precedent-setting antitrust cases, helping to shape law in the digital era. And from 2009 until this past summer, he was senior vice president and general counsel of Intel Corporation, helping to guide the company through its antitrust challenges and a rapidly changing information technology world.
After graduating from Harvard Law School and clerking for Judge Charles M. Merrill of the U.S. Court of Appeals for the Ninth Circuit, he began his legal career at Wilmer, Cutler & Pickering, where he rose to chair of the antitrust and competition practice group. He left the firm in 1996 when he was tapped to be the principal deputy to Joel Klein, who headed the U.S. Department of Justice’s Antitrust Division. There, Melamed helped develop the government’s legal theories for the case that accused Microsoft of monopolizing personal computer operating systems, while also looking at the broader issue of how antitrust laws could constructively address the new challenges of the rapidly growing technology sector. When Klein left the DOJ, Melamed succeeded him as acting head of the Antitrust Division until the inauguration of President Bush in 2001.
When he returned to what became WilmerHale LLP, Melamed again took on what would become seminal cases at the intersection of antitrust and intellectual property law, including Rambus v. the Federal Trade Commission (FTC) in the U.S. Court of Appeals for the D.C. Circuit and U.S. Philips v. the International Trade Commission (ITC) in the Federal Circuit. All required challenging and creative analysis of legal problems—an exercise Melamed enjoys.
“We took positions in those cases that were novel, that people didn’t expect to prevail, and we succeeded,” he says. “We made new law, and I think we made good law.”
Taking a fresh look at legal issues has been central to Melamed’s career, starting at the very beginning with publication of “Property Rules, Liability Rules and Inalienability: One View of the Cathedral,” which he began writing as a 3L at Harvard Law with his professor Guido Calabresi (now a judge on the U.S. Court of Appeals for the Second Circuit). The famed Harvard Law Review article set forth a novel taxonomy of legal entitlement that became a framework for the nascent field of law and economics. It began as a paper Melamed wrote for a third-year seminar and is now one of the most frequently cited law review articles ever written.
Melamed continues to straddle the worlds of practice and scholarship. This year, he is visiting at Stanford Law School co-teaching Going Global: Advising Clients in a Global Economy. And he continues to write important articles, most recently co-authoring “Missing the Forest for the Trolls,” 113 Colum. L. Rev. 2117 (December 2013), with Mark Lemley. The article was chosen as the Best Academic Intellectual Property Article—2014 Antitrust Writing Awards by the Institute of Competition Law and George Washington Law School. In the conversation that follows, Lemley and Melamed discuss the big issues facing intellectual property today.
– By Sharon Driscoll
MELAMED Let’s start with the last term at the Supreme Court, where there were several—many—important questions before the Court [ten IP decisions, six in patent law]. First, why is the Court so interested in IP issues? Is that unusual?
LEMLEY It has been at least a century since the Supreme Court had so many IP cases at once. In part, this reflects the growing importance of IP to the U.S. economy. But I also think the Supreme Court has been dissatisfied with how the Federal Circuit has been handling patent cases. When you look at the last decade, almost all of the Supreme Court’s patent decisions have reversed the Federal Circuit, and they have almost all done so in the direction of restricting rather than expanding patent rights.
MELAMED In an interview earlier this year, you said that the Supreme Court’s Alice decision was “a real sea change” and that “a majority of software patents are invalid under Alice.” Can you explain that?
LEMLEY The patentable subject matter doctrine, which is the subject of Alice, has been a sleeper doctrine for many years. We basically eliminated it for all intents and purposes in 1998 in a Federal Circuit case. It went away for 10 years and then it started to rear its head again in 2008. The U.S. Supreme Court had four patentable subject matter cases in four years, and in every one of those cases, the justices have unanimously agreed that the invention in front of them was not patentable subject matter. And in Alice, while they don’t say “software is not patentable,” what they say is that if your invention, whether it’s in the form of a system or a method or anything else, is connected to an idea, and the only things you do to implement that idea are conventional computer steps, or conventional computer hardware, your idea is unpatentable.
A lot of the cases that are being litigated in the country right now in the software world are patents filed in the late 1990s and early 2000s, during a period when there was no limit on patentable subject matter, and certainly not under this new test. And so they have exactly that characteristic. They say, “I claim any computer programmed in any way to achieve this result, to solve this problem, to reach this particular outcome.” And I think all of those claims are invalid under Alice.
We’ve seen now 15 post-Alice decisions in the software world; 13 of them have invalidated the patents in front of them and 2 have said, “Well, I’m not going to invalidate it under a motion to dismiss, so come back in summary judgment.” And that’s a pretty remarkable record for the four months since Alice.
I don’t think it follows from that software is unpatentable, but I do think that the kinds of claims that people have been asserting and writing for some time, which are extremely broad and not tied to a particular program or particular algorithm, are quite vulnerable.
MELAMED I would think the approach the Court took in Alice could be important outside the software context.
LEMLEY I think that’s right. One of the things the Court says in the Alice decision is that this is a general test. It’s not just about abstract ideas. You can do the same thing to products of nature, or to natural phenomena, facts about the world. In doing so, the Court adopts its 2012 decision in Mayo v. Prometheus, which most patent lawyers had hoped was an aberration, as the definitive test for patentability.
One other area it’s going to significantly impact is medical diagnostics, which was the subject of Mayo. If you discover a correlation between an elevated level of a chemical in the body and a disease, or a vitamin deficiency, that discovery itself is not patentable. So understanding that discovery is not itself patentable, and if all I do after that is routine application of medicine, that also can’t be patented. I’ve got to have an inventive concept that’s different from the abstract concept or idea.
That’s a tough test. And I think it’s a tough test not just in software or medical diagnostics, but you can imagine, conceptually, applying that test in a wide variety of other areas. One of the cases involves DNA sequences; so discovering that this DNA sequence is associated with an increased risk of breast cancer, the Court says, “may be a really valuable scientific discovery, but it’s not patentable. And you don’t get to own the gene sequence if you’ve only done routine modifications to it.”
MELAMED When I was at the Justice Department, I came to think that the biggest unsolved antitrust problem was what to do with patents. Antitrust’s conceptual framework was capable of dealing with all the problems set before it, except the patent problem because, as you and Carl Shapiro have taught the world, the scope and existence of the patent rights are almost always uncertain, at least until they are litigated. By contrast, I’ve never seen a merger case that turned on uncertainty about who owns the factory. I don’t think antitrust has yet to figure out what to do about the uncertainty of patents, which affects a wide range of antitrust issues.
LEMLEY The Supreme Court’s decision last year on Actavis is notable. I think it reaches what seems to me is clearly the right conclusion on the merits, which is you can’t pay your competitors to stay out of the market, even if you have a possibly valid patent. But in the course of doing so, it says something very interesting—that the scope of the patent monopoly is a policy conclusion that’s not just going to be determined by patent law—it’s going to be determined by patent law and antitrust law interacting together. And that’s a very different conception of the rule of antitrust vis-à-vis patents than we’ve had in certainly the last 40 years, which has been that patent law decides what its scope is and then antitrust polices the boundaries to make sure you don’t extend that scope. So I think it may well be that patent and antitrust are headed for a much more dynamic interaction, shall we say, in trying to decide what the proper scope of a patent right is.
Patent law has certainly had an influence in other intellectual property areas too, so we’ve seen patent law doctrines like inducement drawn into copyright law, for instance.
MELAMED Misuse too?
LEMLEY Misuse, yes, gets drawn into copyright. The eBay case is an example where having changed a rule in a patent case, we’ve now applied it to copyright and, more problematically, to trademark cases. But remedies scholars will also tell you it has changed the rule of remedies more generally. That the traditional “four factor test” that the Supreme Court spoke of wasn’t so traditional in equity; it was drawing from a bunch of different doctrines. But we’re now applying the patent-created four factor test in remedial circumstances.
MELAMED Teva Pharmaceuticals USA v. Sandoz is before the Court now, oral argument was this month. How big a deal is this case?
LEMLEY Teva concerns whether appeals courts should defer to the interpretation of patent claims by district judges. District judges have been angling for deference for some time, but so far the Federal Circuit has said that because patent claim construction is a question of law, it is something appellate judges can do just as well as district judges. One of the things notable about the argument is that the tech industry mostly lined up against deference. It seems to be worried about concentrating too much power in the hands of district courts in the Eastern District of Texas and the District of Delaware, the two most popular courts with patent plaintiffs.
MELAMED There’s a lot in the news about “patent trolls,” something you and I have written about. And there’s increasing pressure from tech business leaders for patent reform. Proposed legislation would require demand letters that companies send to accuse others of patent infringement be more transparent and precise in their language and would codify authority given to the Federal Trade Commission and state attorneys general to police the delivery of abusive letters. Would such legislation be important? Will we see patent reform from Congress?
LEMLEY Patent reform directed at patent trolls—companies that enforce patents as a business and don’t actually make products—has been on the agenda in Congress for a decade now, but very little has actually happened. In significant part, that’s because courts have been tackling the same problems Congress was hoping to address. With the changes that have happened in patent decisions in the last year, it may once again be time to wait on legislative reform and see if the courts can solve the problem. Two exceptions may be the threat-letter business model you mention, which has been addressed by a number of state legislatures already, and the cost of discovery in litigation. If we can reduce the cost of litigation, we can reduce the business model of coercing nuisance-value settlements based on bad patents.
MELAMED We’re going to have a new director of the patent office. What advice would you give her?
LEMLEY Thank God, it’s not me who’s got that job! I think the patent office has real challenges that are systemic. Some of those challenges are due to internal management. There’s a culture in the patent office that’s driven in part by the reward system, but also in part just inculcated in the employees, that encourages granting rather than denying patents. It’s extremely hard, in part because of legal rules and in part because of behavior, to actually say no to a patent application finally. I published a paper one or two years ago called “Fixing the Patent Office” that talks about some of those structural limits.
So, we’re seeing more and more patents issued. In this past fiscal year we saw the PTO issue 300,000 patents. That is five times the number issued per year in the 1980s. And that is true despite the fact that those patents were filed during the Great Recession. So it’s not as though it’s a sign of a great business boom and lots of new opportunity. I think it’s a long-term change and the problem for the patent office is that as it gets more and more applications—the system is set up to mostly grant and not to say, “This is not patentable.”
It’s conceivable that Alice starts to change that, in software in particular. Alice might be the kind of tool that makes it easier for the examiners to say, “No. This entire swath of claims is unacceptable and I don’t have to track down individual pieces of prior art to do it.” But given the extreme resource restraints and the fact that the applications keep growing, I think the patent office is hard pressed to do the kind of searching analysis that might weed out some of the invalid patents.
MELAMED That’s not so bad, if there’s an efficient way to take a hard look after the fact at those relatively few patents that matter enough to warrant a close look.
LEMLEY That’s right. And, in fact, because most patents, once they’re issued are never heard from again, it’s probably rational not to spend the money up front weeding out the silly patents of the world—the patents on twisting swings and painting with a baby’s bottom and all of the other things that get issued.
We could probably find prior art on those things if we tried hard enough, but it would be pointless to do so.
But if we’re not going to do that weeding out at the front end, we’ve got to have an efficient and effective mechanism for catching the bad patents at the back end. Patent litigation has been really expensive and really uncertain. One of the challenges the PTO director faces is integrating all of the new administrative revocation systems. I think they offer promise—inner-parties review and post-grant opposition and covered-business method review offer the potential to have that searching opportunity of patent validity only for the patents that are important enough to warrant it and at a cost that is substantially cheaper than litigation.
But I think the real challenge for the patent office, as those things start to flood in, is that examiners are going to be backlogged—even worse than they already are. And fixing that problem is going to be a test for the new director.
MELAMED Let’s talk about patent law heterogeneity. Is there a one-size-fits-all patent system that can work well for all industries?
LEMLEY I think there can be a one-size-fits-all statute that can work for all industries, but that would work precisely because it leaves open enough flexibility for the tailoring of circumstances of particular inventions. An example is the eBay case, which we discussed earlier, where the Supreme Court said the prior rule—which is, you always get injunctions when you win your patent case—is out. You’ve got to have a four-factor, case-by-case test in order to decide whether you get an injunction. And that’s a unitary rule, but it’s a unitary rule that has had radically different impact in different companies. If you’re a pharmaceutical company and you’re suing your generic competitor, you’re going to get an injunction. If you’re a patent troll and you’re suing somebody who’s making a product, you’re not going to get an injunction.
Now, you could look at that and say, “Well, let’s create a rule for patent trolls, a rule for software, and a rule for pharmaceuticals.” But I think what we’ve done is create a test that’s flexible enough that we can treat different circumstances differently.
MELAMED From my vantage point, coming into the patent world from the outside, it seems that the important division is not between industries; it’s between patents covering technologies that are relatively small components of commercializable products, on the one hand, obviously common in tech but not unique to tech, and on the other hand, patents that cover most or all of the commercial value of a product.
LEMLEY That’s one distinction, and you’re right—it is associated with the IT versus pharmaceutical divide, but by no means limited to it. So medical devices might be an example, for instance, of a thing where the design is covered by a single patent, or a couple of patents, and captures a lot of the value. But there are other industry-specific, or at least invention-specific, differences that ought to play into this—one is the speed with which innovation moves. The patent office has a lot of different applications and it takes a lot of time to process them. Alexander Graham Bell’s patent on the phone was issued three weeks after the patent was filed. Now patents take, on average, something over four years from first filing to when a patent issues.
In the pharmaceutical industry, four years in, you’re still waiting for FDA approval. And the real importance of a patent is at the end of its life. People fight about the end of its term and see if they can get more life. But in software, it’s quite often the case that no matter how cutting-edge my technology was when I invented it, by the time I hired a patent lawyer and got the thing on file and then waited four years for the application for a patent and waited a couple more years—even if I filed my lawsuit right away—it’s almost always the case in those fast-moving industries that patent litigation is about obsolete technology. It’s “pay me past damages for something that nobody uses anymore,” or it’s an effort to apply an invention that is old to a new kind of technology that is three generations removed. So those two things interact in different ways. When we look at patent litigation in the software world, we’re quite often looking at efforts to assert a patent on one small component of a product that was developed many years ago, against the entirety of a new product that’s in a different generation.
Even in circumstances where we move fairly quickly, you can still see that. The example to me is Apple v. Samsung, where no matter how quickly they try to process the cases to trial, the products are obsolete by the time the trial comes around. So we had a trial about a set of already-obsolete products and by that point we’d already filed another lawsuit. We had a second trial about another set of already-obsolete products, and by then we’d already filed a third lawsuit. And you can imagine an endless parade, because the patents and the patent litigation system are just always going to run more slowly than the product cycles.
MELAMED It has a flavor of World War I trench warfare—with nothing being accomplished.
LEMLEY I do think that’s right. And it means that the basic theory of patents—that they give exclusivity in the marketplace and therefore supra-competitive returns—works in the pharmaceutical industry but just doesn’t work in software.
MELAMED There seem to be two competing paradigms in conversations about patent law and policy. There’s the property rights notion, in which the rules are deduced from the concept of patents and property. And there is the idea of patents as instruments of government policy, in which the rules are chosen to further policy objectives. But those strike me as rhetorical framing devices that advocate use to advance their cause. What’s really at stake in your opinion? Why are these battles so intense?
LEMLEY One of the reasons is that rhetoric has power. I find myself fairly strongly in the “patents are government regulation camp,” not in the “patents are property camp.” I wouldn’t be so concerned about the patents are property story if I thought that everyone who said it understood and internalized that they are “kind of like property rights in some respects, but here are all the ways that they are different.” They have no certain boundaries, they are contingent, they’re granted right now instead of 300 years ago, so we haven’t internalized them; there are new ones being granted every day, they interfere with what you can do with your own private property—all of these differences.
But we use these labels precisely because they serve to elide those distinctions and they allow us to sort of reason by analogy, even when that analogy is not a very good one.
MELAMED I think that is a sensible explanation of why these framing devices are forceful. What you didn’t say—and I think it’s a striking omission—you didn’t say that these battles are so pitched because there is a lot at stake and they are essentially battles between haves and have-nots.
LEMLEY Maybe, though I’m not sure. One of the things in the patent business is that it’s not that obvious who the haves and the have-nots are. In general, the people who have patents might be more inclined to view them as property, because I think in doing so that usually makes us treat those rights more powerfully. So if by have, you mean have-patents, that might be right. But we’ve also got this interesting world in which the people who are most vocal in asserting patent rights are not the largest patent owners. The companies who actually invest the most money in the patent system, by and large, are not patent plaintiffs. There are exceptions—I think Apple is one. But Intel, Google, Microsoft have not classically been patent plaintiffs.
MELAMED Let’s look at the question of the haves and have-nots in the national or international context—not just a company-by-company context. You have China and other emerging economies with their indigenous innovation policies that are thought by most in the North Atlantic community not to give sufficient respect to intellectual property rights. And you have a history of the United States in the 19th century often just ignoring property rights. And then there is the TRIPS Agreement in 1994 where the emerging economies thought, to some extent, that patent rights were being crammed down by the haves against them as the have-nots. So you can look at that and say, “Intellectual property rights are respected and protected when people or nations own them.” Is that a fair characterization?
LEMLEY It’s a fair characterization. There is a parallel to the phenomenon we see, primarily in the IT industries, of companies acquiring patents for defensive purposes. Really what that company wants is freedom to operate. The ways in which patents in particular look like a form of government regulation is “I’ve got to participate in the patent system in order to try to simply build from things that I myself invented, without any interference from the legal system.” Even then, it’s only going to work to a limited extent—I’m still going to get sued, not withstanding the fact that I invented that product on my own, because it turns out that somebody else also invented it.
One of the papers I’ve done recently, “The Myth of the Sole Inventor,” focuses on just how much independent invention and creation there is in the world. It’s really quite remarkable and not just in industries like IT—where Chris Cotropia and I have shown that somewhere between 95 and 99 percent of all patent lawsuits are filed, not against anyone who’s accused of copying from the patentee, but against independent inventors.
But it’s even true when you go back and look at great inventions throughout history. There are often lots of different people working on fixing a problem, who all come up with a similar solution at around the same time. And the patent system reflects a judgment that says, “You stole this from me and therefore I should be able to stop you.” That’s something that sounds more like property. So when two people independently developed the same thing at about the same time with their own physical property and with no knowledge of each other, the government is going to choose one of those people and say the other one has to stop.
MELAMED Is the notion that, if the government through the patent system gives big rewards to the winners, the races are run faster?
LEMLEY Yes, absolutely. And the fact that it’s government regulation doesn’t mean it’s a bad thing. It could well be government regulation in a good cause—and I believe that the IP system is a good cause. But we think about it differently if we view it as government regulation rather than as property. Regulation requires justification for its existence. So it’s not to say we shouldn’t have an IP system; it’s to say we should look at this government regulation, as we should with any government regulation, and do a cost-benefit analysis. Is it working? Is it giving us innovation?
If you call it property, that’s now “rights rhetoric,” so this is something I’m entitled to—whether or not it’s benefiting society—because it’s mine. And that worries me just because there’s increasing evidence that it’s not always good for society.
MELAMED You’ve had a 20-year academic career. You’ve been unbelievably prolific—and not just in quantity, but in quality. But I’m struck by the sheer quantity. How do you manage, not just to get the work done but to have the range of ideas to write that kind of volume?
LEMLEY A part of it is just keeping up with what’s going on in the world. New cases are decided, and I’m fortunate to be in an area where things change all the time and a lot of the basic issues and principles haven’t been long settled, so there’s a lot of work to be done. But it’s also helpful to have a bridge between teaching and practice, and I think a decent amount of inspiration comes from disconnects you see between those different worlds. You know—from seeing the fact that scholars all assume one thing about the way a patent doctrine works, but the way it’s actually being practiced is very different. There are often very good questions to ask in those kinds of disconnects.
MELAMED The connection between you and practice, I find, is emblematic of what I see as an evolution in universities. When I was a student, universities were thought of as retreats for reflection, the ivory tower. Now, I think of them as evolving toward a model in which the university is a hub—none more so than Stanford—a connection between scholarly endeavor and certain aspects of the outside world.
LEMLEY Yes. I think that’s absolutely right. I think there are various factors driving it—some good, some maybe less good. There’s been a real explosion of legal-tech startups coming out of the law school. And part of that, I think, is just the Silicon Valley mindset. I like to say that you know you’re in Silicon Valley when the law students are starting companies. But it’s also—I suspect you’re right—a more general move in the university toward relevance, toward being engaged in the world around you. I guess I’d say in general that it’s a good thing. I think that kind of cross-fertilization can be useful. The only hesitation I have is that I do think there is something lost in the absence of the ivory tower. There’s value in pure research for research’s sake. And universities have to make sure that, in the increasing connection they have to the real world, they don’t eliminate the possibility of pure research, because pure research is often what turns out to have the biggest long-term scientific benefit.
MELAMED You anticipated a question I was going to ask. The term that came to my mind was the tyranny of the immediate.
LEMLEY I write a number of articles that I would call immediate. Maybe they’re about a specific case or they’re about a particular patent doctrine existing at a particular moment in time, and you might put empirical work in that category too.
You know: “Here’s a snapshot of the world as it is.” But I also think you want some kind of broader pieces; pieces that step back and say, “Hey, what’s really going on here?” And I think our article together is, in some sense, about that—it’s a broader look that says, “Let’s not focus on patent trolls, per se, which are the cause célèbre of patent reform today, but what do they tell us about the broader patent system.” I think the key is to find a balance between the long-term theoretical stuff and the immediate pieces. If you get it right, the connection with the world can actually inspire the more abstract, the more philosophical or the more long-term things. And they, in turn, can help shape the more immediate matters.
MELAMED It’s been a pleasure speaking with you.
LEMLEY Thank you. I’ve enjoyed it.