Why NPEs Are OMG!
Last time, I explained what patent “trolls”—more politely known as a non-practicing entities or “NPEs”—are and why their business model is such a good one. Mostly they benefit from asymmetrical warfare: a defendant must spend much more money, time and other resources than NPEs, to say nothing of potential liability. Also, the types of patents that NPEs like—broad patents in popular fields with early priority dates—are undervalued and easy to pick up on the cheap. All you need are some good patents and a law firm experienced in enforcing patents willing to take the case on a contingency, and what’s not to like?
Since my main goal is not to condemn (or, for that matter, praise) NPEs, but to explain how got to be where they are, I’m discussing ten patent rules and truisms that make NPEs lucrative. Last time, I discussed three of them: (1) you don’t need to practice the patent to enforce it; (2) you can buy and sell patents like property; and (3) once issued, patents are difficult to invalidate. This time, I’ll discuss the other seven and provide some thoughts about reform (assuming you think reform is a good idea).
Well-timed injunctions
It used to be that if you succeeded in enforcing your patent, you could count on the court to enjoin (i.e., prohibit) the defendant from practicing your patent. Damages might be scary, but the prospect of being deprived of your bread and butter is terrifying. That’s some stick to wield at the negotiations for licensing.
Ideally, though, you don’t want to spring the injunction threat too soon. You want to wait until the manufacturers are fully committed. Threaten too soon, and they may either design around your patent or just abandon the whole thing.
This threat, however, isn’t what is used to be after the Supreme Court’s decision in eBay v. MercExchange, which held that injunctions are no more automatic in patent cases than in other cases. There is some empirical evidence that NPEs are not obtaining injunctions any longer because they have difficulty proving irreparable harm—a downside of not producing anything.*
* If you have a software patent, one way around this is to actually go to the trouble of producing a software product that practices the patent and market it. It’s not free, but it’s a lot easier than the headache of actually manufacturing something.
Complex devices are often subject to multiple patents
We tend to imagine inventors pointing to some device and claiming, “I have a patent on that.” But it’s much more common, especially these days, for an invention to cover only a part, often a small part, of a larger, complex device. If a patent claimed a particular type of processor, you could sue whoever makes such processors without your permission. But you could also sue someone who put those processors in, say, smartphones. In fact, a typical patent is probably subject to thousands of patents because it has so many complex components: the processors, the memory chips, the touch screen, the speakers, the cell-phone and WiFi receivers, to say nothing of soft features such as how you access or lock the phone—yet none of them necessarily is a patent on a “smart phone.”
We’ve been here before.* In the nineteenth century, the Singer Sewing Machine Company combined, in essence, four patented inventions, with innovations of its own to create the first really popular sewing machine. The problem was that Singer was also necessarily infringing those other patents—the fact that it had its own patent was irrelevant because a patent isn’t a right to do something but a right to keep others from doing something. The litigation was breathtaking, and included at least one honest-to-goodness NPE.** In the end, the major players developed the “patent pool,” whereby a separate entity held their patents and cross-licensed them to the others.
* The article I link to seems overly-optimistic to me. Just because an industry avoided being destroyed by conflicting patents doesn’t mean that none has been. Indeed, if conflicting patents had strangled an industry in its infancy or delayed it by 20 years, it’d hardly be common knowledge.
** It is said that Mr. Singer was infuriated by the NPE: “[He] is a perfect humbug. He knowns quite well he never invented anything of value.”
Something similar happened to the airplane industry in the early 20th century. The Wright Brothers (you may have heard of them) were the first to develop “airlerons,” crucial components for roll control, but they weren’t especially good at developing a business that could manufacture airlerons-equipped planes. But they were good at suing the socks off people. The Wright Brothers (well, Wright Brother, by that time) were generally the victors, but the fighting went on until World War I, when the U.S. government forced the parties to dump their patents into a patent pool, so it could, you know, build airplanes to fight in the war. Until then, the airplane industry was, at a minimum, severely hampered.*
* These folks beg to differ, arguing essentially that the aviation industry was progressing just fine, but that the government seized on the lawsuits as an excuse to extract lower royalties for planes that it was the main market for. I suppose that, in the end, it’s difficult to tell how the industry would have developed without the patents for comparison. I can’t find a published version of the paper for some reason.
If a smartphone is covered by, say, 1000 patents, then there are 1000 opportunities for infringement, and 1000 opportunities for an injunction that blocks the entire phone. If the invention isn’t very crucial to the phone, the manufacturer can design around it or use a different component, but even minor changes can result in disruptive retooling of the manufacturing process. If the invention is crucial, then, well…. The gold standard, however, is an invention that every smartphone manufacturer habitually uses. Instead of targeting the way one smartphone is put together, you can target an entire industry with a single patent.
Strict liability: Ignorance is no excuse
Almost any legal trouble you can think of getting into has some kind of “intent requirement”: a mental state you must possess before you can get in trouble. The bar can be high—e.g., you knew what you were doing and what you were trying to accomplish—or low, e.g., you just weren’t very careful, or you knew what you were doing but just didn’t realize it was illegal. But there’s some bar. If you pick up someone’s smartphone by accident, you haven’t broken the law because you didn’t mean to. If you give your friend some wrong information because you were too lazy too look it up (but you honestly thought you were being correct), you aren’t liable for fraud because you didn’t mean to deceive, though if you were a lawyer or financial advisor, you might still be in trouble because you should have done more (i.e., double-checked).
But there are a few laws that have no intent requirement at all, most notably products liability. Patent is one of those laws.* The main purpose of strict liability laws is to force actors to be extra careful and to take away the option of looking the other way. A car manufacturer will take extra care to make sure all the components going into its cars will be safe, and far from looking the other way, will carefully inspect and monitor its manufacturing processes.
* Copyright is also strict liability, but it has a requirement that you actually copy or have access to the copyrighted work, which means you can’t be liable for copyright infringement while being wholly ignorant of the copyrighted work. Update: Actually, that last bit goes too far. There are plenty of circumstances in which the infringer could have no idea that she is infringing, despite having access to the copyrighted work. For example, she might honestly (and reasonably) believe she has permission to use the work. For example, if A took B’s song and gave it to C to publicly perform, C would be infringing the public performance right even though she had no idea B is the actual songwriter and no reason to suspect otherwise. So copyright’s strict liability can be a tough business, just not as tough as patent’s.
This reasoning doesn’t really apply to patents because it’s impossible to be aware of every patent that might be relevant to what you’re doing. In a perfect world, you’d know about the patent ahead of time and either design around it or reach out to the patentee for a license.
The strict liability aspect of patent infringement catches accused infringers off guard because it seems counter-intuitive. It would be one thing if you were aware of the patent and copied it because you thought it was a great idea. But in almost every case (in my experience), the patent comes out of nowhere. By the time you receive the demand letter, offering to license the patent to you now that you’re already practicing it, the patent doesn’t add any value to what you’re doing. You’ve already gone to the trouble of re-inventing the wheel, so to speak.*
* This discussion makes the incredibly naïve assumption that patents are very helpful in actually implementing their inventions. In theory, they’re supposed to disclose the best method for carrying out the invention, but in practice, they disclose a way of carrying out the invention, not necessarily the best or even a very good way. There’s a huge difference between describing an invention in the abstract and actually making the invention a practical reality.
There is an important limitation to strict liability: you are liable for money damages only if you have notice of the patent. That makes sense, right? It’s one thing for a patentee to say to you, “Hey, that’s my patent. Stop it.” It’s another if the patentee also says, “Oh, and you owe me $150,000,000” for using my patented invention for the last six* years. Injunctive relief is still available, but as I’ve just said, that’s not usually available to NPEs any longer.
* The statute of limitations for patent action is six years, so damages can’t normally go back further six years.
There are a couple of effective ways around this limitation. First, if the patentee made or authorized the making of devices that use the invention and marked the device with the patent number*, the world (including you, the accused infringer) is deemed to be on notice of the patents so marked. NPEs don’t make anything, but sometimes the patent, in a previous life, was actually used in connection with actual devices. Second, the NPE can send you a friendly-sounding letter offering to license the patent to you, then wait for up to six years**. In any event, past damages may not be where the action is: the NPE may be angling for a running royalty in the future.
* Look on almost anything remotely complex that you own, and you’ll see patent numbers. Now you know why. Note that “patent pending” is legally meaningless. Figuring out how to mark a process is difficult.
** Again, the limitations period for patent actions is six years. Waiting can hurt the NPE’s chances for an injunction, but injunctions aren’t as important as they used to be for NPEs. There are other risks of waiting too long, but smart NPEs know how to manage them.
And the inventor’s intent is irrelevant
If you go look back at the four NPEs I once worked against in my early career, you’ll notice that, with two of the patents, the inventors were clearly trying to invent one thing, but the patent was drafted broadly enough to cover an entirely different and unexpected technology.
Consider this patent, and ignore the reexamination material for now: If you just look at the figures and read the introductory paragraphs, you’d think it’s describing a kind of hand-held, pen-shaped device that reads bar codes, which would have been kind of neat in the late 1990’s. And, pretty clearly, that’s what the inventors were getting at.* But it was used in 2001 to sue manufacturers of touchscreen-enabled personal digital assistants, which look nothing like what’s in the diagram.
* In fact, they were able to manufacture and market such devices. They just weren’t very successful.
How? Because the key term that described the bit that did the scanning, “reading sensor,” wasn’t, on its face, limited to just a bar-code scanner, but could arguably be understood to include any sensor that “reads,” including a touchscreen (which senses touch and “reads” the touch inasmuch as it knows where the touch took place). That was in 2001. Nowadays, the patentee has a rather better argument regarding built-in cameras. From their press releases, they’ve done OK.
Users Can Be Targeted
People are usually surprised to learn that it infringes a patent even to use a patented device (or process) without permission.* Considering that there are literally millions of enforceable patents (most forgotten about) out there, you’re probably infringing on one right now. But I would reassure people by saying that no one ever sues the consumer or other downstream user because it’s not worth the effort.
* Note that if the manufacturer is licensed, then the purchaser is protected by the first-sale doctrine. If the manufacturer isn’t licensed, then everyone downstream—middlemen, distributors, retailers, consumers—is vulnerable.
Apparently, that’s no longer true. There is at least one NPE that targets small businesses regarding patents that, it claims, covers the act of scanning a document and sending the resulting file to someone else by email. They ask for license fees in the low $1000’s, which is much less than the cost of defending against the patent.
I wonder if traditional NPEs are happy about this turn of events. The problem with suing consumers and small businesses* is that they start to complain to their elected officials, who are suitably horrified and aren’t usually great experts on patent law. Stopping “patent trolls” at all costs might become a thing, and a law curbing “patent trolls” would encompass many if not all NPEs. If there is some economic benefit of NPEs….
* Have you ever heard a politician say anything bad about small business?
Patent damages are really huge
Remember Apple’s $1 billion* patent victory over Samsung? For patents that included the incredibly mission-critical “rubber-band” scrolling feature, which I totally love but never influenced my decision to buy an iPhone. I believe I have already adequately expressed my amazement. Apple is hardly an NPE, but for purposes of determining most damages, the distinction is legally irrelevant, i.e., awards of this size legitimately make dollar signs appear in NPEs’ eyes.
* Temporarily reduced pending a new trial because the jury got something wrong. It could be lower—or even higher. Everybody say it with me: “One-hundred-beel-yun dollars!”.
In 2009, Microsoft was ordered to pay what might be termed a quasi-NPE about $290,000,000 for infringing on a software patent that had something to do with the way Microsoft Word handled XML code—a feature that wasn’t all that popular.* The patentee was a true NPE because it did produce its own software that practiced the patent, which, in a sense, competed with Microsoft. Then again, since the patent was for software, it’s not that hard to sell the software once you’ve developed it.**
* Microsoft was forced to get rid of the feature. Did you miss it?
** The software-based nature of the patent cut both ways. It wasn’t very hard for Microsoft to cut the feature from its flagship product and keep selling it for the same price. That’s almost certainly why Microsoft was willing to take it to trial (and, ultimately, to the Supreme Court): there was almost no risk beyond the damages (which it could absorb).
Why are patent damages so high? To answer that, we must resort to speculation (because we can’t interview enough jurors). First, damages in U.S. courts are always high. Second, the main form of patent damages—reasonable royalty—is determined by pretending you’re back at that mythical place where the manufacturer learns about the patent ahead of time and negotiates with the patentee, and guessing what the result of those negotiations would be. Only, it’s impossible to ignore the fact that the defendant has been practicing the patent and is pretty committed to the product, so the jury assumes that the defendant would have been really, really desperate, because without the patent, the defendant couldn’t have done anything and would shriveled up and died or something.
Third, juries don’t appreciate how hard it is to actually make something that people want to buy. Patents cover every conceivable version of things that practice that patents. Not just the popular version that’s the subject of the suit, but also the unsuccessful, clunky, ugly, over-priced, poorly-designed, poorly-manufactured, mis-marketed ones. Or the super-awesome ones that would have been produced except the company’s funding fell through, or was torn about at an early stage by internal dissension. If you’ve ever started a business that produces something, you know how hard it is. The basic technology—and that’s all a patent is, at best—is just the start. But juries read that story differently: since the basic technology is the start, it’s also the sine qua non. Without it, you never would have had the chance to screw up everything else.
Oh, and there’s “willfulness.” If you did have notice of the patent and practiced it anyway without a good-faith, reasonable belief that you were avoiding the patent, you are liable for up to triple the damages. It used to be that willfulness findings, and treble damages, were commonplace, but not so much any more. Still, the threat of treble damages is significant.
Patent cases are expensive.
The American Intellectual Property Law Association (“AIPLA”) conducts a survey of how much defendants spent in patent lawsuits. In 2011, it said the average cost of defending a suit worth less than $1 million, through but not including trial, was $490,000,000. If you want a trial, too, that’ll be $916,000,000 total. That’s if you win. Bigger cases cost even more.
Patents are unbelievably complex beasts, with multiple claims, written in incomprehensible language that’s hard even for seasoned patent litigators to get their heads around, with an extensive patent file to go through, and often complex or obscure technologies involved. Experts are required (usually two: one for the technology, one for the economics), and they’re expensive. There are lots and lots of relevant documents (literally millions of pages if the defendant is pretty big), lots and lots of witnesses (I’ve never seen fewer than a dozen deposed). As I’ve discussed, the process of finding and authenticating prior art is expensive and arduous.
Take a look at your insurance policy. You’ll notice that it always excludes patent damages. Why? Strict liability. Insurance companies prefer to insure risks that can be controlled, but patent damages come out of nowhere, there’s little you can do to avoid them, and are nearly impossible to predict. You can get patent insurance, but it’s going to be very, very expensive.
Patents are the anti-anti-trust.
When the targets of NPEs try to strike back legally, they almost always reach for some theory based on anti-trust after all. After all, someone is trying to restrain their trade, and even have the power over who may use a crucial technology.
But patents are designed to be a restraint on trade. That’s how they work. Therefore, they are an exception to anti-trust and similar laws (including RICO). The only exception–patent misuse—isn’t really an exception. It occurs when you try to use the patent to get a monopoly beyond what the patent would allow. The classic case would be a licensing agreement that went beyond the term of patent and restrained the licensee from using other providers. Attempts to expand this doctrine into the scope of what the patent covers have all failed, except perhaps where the patentee’s interpretation of the patent’s scope is patently (ahem) absurd, but I’ve never seen that theory succeed.
America is build on free commerce and innovation. With patents, those two values collide head on.
Is There a Problem? What Is it?
There’s been something of a crisis of faith in the patent system, and the patent system is difficult to defend because requires a bit of faith. If you look at the patent system case-by-case, it just looks awful. With some exceptions, like pharmaceuticals (which are expensive to develop but easy to reverse-engineer), it’s very difficult to directly link a patent to an innovation that we value. This goes double for NPEs’ patents, which are usually the flotsam of businesses that failed or never were. Since the company that ends up being successful with the claimed invention almost always developed the technology independently on its own, it sure doesn’t look like the patent helped anyone.
So what we end up seeing is a number of individual cases where patents seem to do nothing but interfere with commerce. Worse, when a patent is asserted against a company that independently developed the technology and seems to be making a successful go at producing and marketing devices that use the technology, the patent looks ironic: innovation appears to be stifled in the name of innovation.
The popular conception of the patent ecosystem goes like this. X invents, X patents; X needs a business partner to make the invention a reality; X licenses the patent to Y, which has the ability to manufacture, market and sell things. Or Y sees a market opportunity but lacks the technology; Y looks for relevant patent and finds X; X licenses the patent to Y. It certainly doesn’t work like that today. I seriously doubt it’s ever worked like that.
If the patent system works at all (again, outside of obvious cases like pharmaceuticals), it works on a very large scale. Surely, the promise of patent royalties—or sale of the patent—motivates people to innovate: if you work hard, you’ll end up with something of value, even if you fail at turning your invention into a viable business. And although hardly anyone uses others’ patents as their basic technology (because the written description is too idiosyncratic and usually doesn’t give away any more information than it needs), information about innovation gets disseminated.* Maybe you can’t connect the dots from lone inventor to popular or crucial device, but there are a lot of dots and they flow the same way.
* To be sure, there are patents held by highly-secretive inventors, but most inventors are proud and eager to share. Patents encourages that because the inventor no longer must keep the technology secret because he or she feels that it’s “protected.”
But that patent’s value takes the form of a tax that we all must pay. Royalties are passed onto consumers, or make staying in business more difficult. Valuable products might be delayed until some business can reach terms with the patentee. Patents might help innovation, but they’re hell on commerce.
It seems to me that if you think NPEs are a problem, then you think the patent system is a problem, and I can accept that. Getting the balance between innovation and commerce right can be difficult, and reasonable minds can disagree* about (1) what the balance should be and (2) how close we are to achieving it.
* Not that the debate has been very nuanced, reasonable or productive. Have you ever read the comments to the excellent Patently-O blog? It’s like the Usenet in 1997.
Whither Reform?
The question comes down to: is the tax too high for what we get? If you think the answer is yes, then reform must focus on both the tax and “what we get” sides of the equation. I don’t know about you, but I’m pretty satisfied with “what we get.” There has been no shortage of innovation. If there’s a problem, it’s that the tax is too high. The poster child for that is Apple’s rubber-band scrolling patent*, which I sort of enjoy, but is not worth even $1 per unit. More like $.01.
* Which, by the way, is in some danger of being invalidated, but for reasons that aren’t relevant to this discussion.
If we got patent damages under control*, a lot of other things would fall into place. That would make patents—especially lame patents—worth less, which would diminish NPEs’ bargaining power and force them to search more diligently for more valuable patents. NPEs’ margins would decrease, and they’d have to get better at what they do. With the stakes reduced, legal fees should come down as well,** though probably not enough. Bad patents would still be bad, but they wouldn’t do as much harm. New innovative companies would actually be able to budget for and pay for patent licenses (whereas, right now, if a patent rears its ugly head, it’s usually over).
* You might have noticed that I have a thing about damages.
** Common sense tells us that, the more money is at stake, the more you’ll be willing to pay to avoid paying that money or to obtain that money.
How do we get patent damages under control? First, jettison the “hypothetical negotiations” test for calculating reasonable royalty. It’s unrealistic and circular. Second, replace it with a new test the focuses on the value of the basic invention as just a component (rather than the component) of the value of the infringing devices.*
* OK, I can see I need to work on step 2.
To reduce the value of nuisance patents, we could implement a kind of fee-shifting arrangement.* Fee-shifting forces parties to become realistic about their cases earlier, before the other side starts racking up fees. Patentees will need to be more cautious and do more homework before filing suit, but infringers—even big ones—will need come to terms earlier and more realistically. Lots of devils in these details.
* Fee-shifting is always controversial in the United States because we’re the only common-law country (i.e., England, Canada, Australia, etc.) that doesn’t do it on a regular basis. In fact, the rule that parties pay their own legal fees is known as “the American Rule” (and the rule that the loser pays everyone’s legal fees as “the English Rule”). It might be a total coincidence that we see many more lawsuits here and they’re way more expensive.
If you think bad patents are the problem, I think it would be best to wait and see how the new procedures under the AIA work out. These procedures might also affect the value of patents, especially nuisance patents. It may turn out that these procedures aren’t quite good enough, or are too easily to abuse, but at this point, we’ll need to watch carefully.
That’s enough on this.
Thanks for reading!
