It’s More Complex Than You Might Think
With several bills pending in Congress and executive action by the current presidential administration, with the accompanying rhetoric, you just might start to wonder about—or maybe have already formed an opinion on—“patent trolls.” The issue is actually fairly complex. While I think some reform is a good idea*, I’m eager (as always) that we not throw babies out with bathwater. So you can make up your own mind, what I thought I’d do is set out the patent laws and principles most important to the phenomena of the “patent troll.”
* Something more nuanced than either “patents suck; they always interfere with innovation” and “patents are awesome; they always lead to innovation.”
The term “patent troll” is a pejorative term that means slightly different things to different people. It could mean:
- Simply any patentee (owner of a patent) that does not practice the invention disclosed in the patent, politely known as a non-practicing entity (“NPE”).
- A patentee that doesn’t practice anything except enforcing the patents it owns, politely known as a “patent holding entity” (“PHE”) or, more positively, a “patent licensing entity.”
- A patentee that not only doesn’t practice the disclosed invention, but enforces its patents in a morally objectionable way. Mostly, what people have in mind are the folks (whoever they are) behind the scanner-connected-to-a-network enforcement.
I’m going to be using the first definition, and since I’m polite, I’ll be exchanging the term “patent troll” for “NPE.”
Lucrative Business Model
My own experience with NPEs is illuminating. From 2000 to 2003, I was a patent litigation associate, and of the six major patent cases to which I was assigned, four involved NPEs:
- The first case involved a fairly notorious patent for sampling music over the internet. The patentee had originally planned to make and market devices embodying the invention, but when that fell through, it took a look at the patent’s breathtaking breadth and sued a number of companies in 2000 that were using sampling to promote CDs they were selling. It is not uncommon, even then, for failing companies to try to “monetize” their patent portfolio to get cash, perhaps turning into a PHE by default. By 2002, there wasn’t much left of the patent.
- In the second case, a large, well-established company sued a couple of well-known but new companies over a technology that the large, well-established company didn’t practice: personal digital assistants. This was probably a case of a large flabby company trying to squeeze some money out of its patent assets without having to research and develop products that embodied the invention.
- In the third case, the patentee was a company that had plans to develop a product but never got off the ground, and it eventually became a PHE with some dreams of becoming something more. The patentee was trying to develop a pen-like device that reads text or bar-codes, but ended up suing , on a contingency basis, manufacturers of touch-screen-enabled personal digital assistants.
- In the fourth case, the patentee was a straight-up PHE that had acquired its patents strategically by purchasing especially good ones from bankrupt companies, where they may be purchased cheaply. This one involved cell-phone switching technology, and the patentee hired a firm to sue a bunch of cell-phone companies in Chicago on a contingency.
It’s not as though my firm didn’t also work on competitor-vs.-competitor patent suits—there were plenty of them*—but even by 2001, patent suits by NPEs were clearly a trend. We associates discussed whether we’d be better off quitting our jobs and trolling for patents: scouting bankruptcy filings for cheap patents. The idea patent would, obviously, relate to an emerging and lucrative technology; has very early “priority date” (the date before which prior art is, well, “prior”); and can plausibly be interpreted to cover a range of popular devices. What doesn’t matter is how much time is left on the patent. If the covered devices are popular now, even a year’s disruption would be good bargaining leverage—five years’ disruption tremendous leverage.**
* It’s worth describing one of my other two patent cases: a very large competitor brought barely-colorable claims against a much smaller but more technologically innovative rival. The larger company drove the smaller company into bankruptcy just on the legal costs alone.
** Arguably, you start getting diminishing returns from more than five years, since the target companies would have completely changed the technology by then.
NPEs of all sorts have tremendous advantages over those they sue. Because the NPE doesn’t do anything, it’s very easy for it respond to discovery requests—it doesn’t even have that much to produce. In fact, it has been structured for the sole purpose of litigating efficiently. By contrast, the defendant is often caught off-guard, has structured its business to make and sell things, not to respond to patent lawsuits. Patent lawsuits are very, very, very expensive (maybe $500,000 to $1 million in legal fees, for a lawsuit with a manageable number of claims), but much more so for the defendant than a NPE.
Further, a defendant’s main defense—the patent counterclaim*—obviously doesn’t work against NPEs because, again, they don’t do anything. In a competitor-vs.-competitor lawsuit, the defendant is likely to have its own patents that the plaintiff might be infringing on—after all, they both in the same industry doing very similar things.
* The idea that the best defense to a patent lawsuit is a good offense is still very much in vogue. More so, nowadays, which has led to a kind of patent arms race among several leading technology companies.
Ten Eleven Easy (?) Pieces
So, how did we get here, and is it really that big a problem? I can answer the first question much more easily than the second question. To understand how NPEs got be such a good business, you have to look at certain patent laws that they take advantage of:
- You don’t need to practice the patent to enforce it.
- You can buy and sell patents like property.
- Once issued, patents are difficult to invalidate; and the older their priority dates, the harder they are to invalidate (but the broader they are, the easier).
- Injunctions can be devastating if timed right (but aren’t what they used to be).
- A complex device might be subject to many patents, and if a patent covers just one part of the complex device, an injunction would block the entire device.
- Patents have “strict liability”; i.e., you don’t need even to be aware of the patent to be liable for infringing it (though you might need to be aware before you can start being liable for damages).
- Intent on the inventor’s part is irrelevant; it’s what the patent’s claims say that determines a patent’s scope.
- You can infringe a patent not only by making or selling products that “read on” the patent, but by merely using such products.
- Patent damages are determined by a huge number of factors and are almost always really huge; and they aren’t limited to royalty rates.
- Patent cases are about the most expensive cases to defend, and insurance hardly ever covers it.
- Patents are an exception to anti-trust.
Add these together, and you have fertile ground for NPEs. Obviously, they don’t practice their own inventions. You can also create a special-purpose entity just for enforcement, in case you don’t want the public knowing who is really behind it. The most effective ones search around for cheap but effective patents, often out of bankruptcy. Don’t worry about what the inventor was trying to do; just look at what the inventor’s lawyer was able to achieve. Invalidation of the patent is the worst-case scenario, but it’s also really hard for a defendant to pull off. The patent doesn’t need to cover the entire target product, just enough of it. You don’t need to bother with any R&D—you just look for someone who has already gone to that trouble. If you don’t feel like fighting someone who can fight back, you can go after consumers and businesses that merely use the (allegedly) infringing device or system. You know the damages are going to be big because juries will assume that, without your patent, the infringing devices would have failed, and interruptions to popular products are a big part of every manufacturer’s nightmares. And your potential opponents aren’t ready for you, and might not even have enough money to last a year.
It stands to reason that any reform relating to NPEs would involve tweaking these rules. But to do that, you need to know the policy reasons behind these rules.
Why don’t you need to practice what you enforce?
The theory is that an inventor has invested time and resources into the invention, and that sort of effort is something we want to encourage. Thus, even if the inventor can’t pull together enough investors to turn the invention into a viable product, or the inventor just isn’t good enough at business to manufacture, market and sell products that embody the invention, the inventor should be compensated.
Why do we let patents be transferred like property?
The theory is that, even if the the inventor can’t even pull together the resources to enforce a patent, or simply finds licensing and litigating distasteful, the inventor shouldn’t be stuck with a worthless patent. He or she should be able to sell the patent to someone willing to license and enforce it, the price being determined by the likely royalties. The inventor is thus further encouraged to expend time and resources in inventing new things.
One reason that NPEs are so profitable might be that the market for patents is very inefficient. Unless there are actual royalty payments being made on the patent, it’s very difficult to value a patent. Most have almost no value. But, in theory, with the advent of NPEs, they should be worth more now than before because NPEs can extract more value out of them. This should cause the price of the really good patents (for NPEs) to go up to the point where NPE’s profits are squeezed. But nothing like that appears to be happening: patents are still a bargain.
Why is it so hard to invalidate patents?
Mostly because a patent has already been vetted to some extent. If a patent could be easily invalidated, why go the bother (and expense) of the vetting process? It has to count for something.
When you apply for a patent, the examiner searches for prior art that either anticipates the invention or renders it obvious. The examiner’s resources are fairly limited, however. Mostly he or she looks at previous patents, which are readily available and relatively* easy to search through. He or she might also have access to certain publications, but not to a great deal of art “out there.” Work being done by obscure academics, companies in quasi-secrecy, and researchers who abandon their work is difficult to locate easily. In theory, a single publicly-accessible product or even published papers describing a device could derail an entire application. But, you know, needle, haystack.
The system works well because patents over time are a pretty good proxy of the state of the art. But it breaks down for emerging technologies for the simple reason that there aren’t (yet) any relevant patents, and the literature is too undeveloped to be helpful—all the art is in guys’ garages, personal laptops and scribbled napkins.*
* Software presents a special version of this. Obviously software is hardly an emerging industry, but before 1998’s State Street decision, it couldn’t be patented. When the software applications started, there obviously could be no prior patents for examiners to reference. Worse, the USPTO had a prejudice against “soft” engineering and science, so very few of the examiners knew anything about software. The result of was a lot of very, very bad “soft” patents. The puzzle of how to protect software is far from resolved, but this was a pretty big part of the problem.
So, when a defendant is confronted with a broad patent that probably covers its products (and probably shouldn’t have been issued), it is left with the arduous task of invalidating the patent at trial. Why is it so arduous? Let’s count the ways:
- Because of the prior vetting, patents get something that almost nothing else gets in civil litigation: a judicial thumb on the scale. To invalidate a patent at trial, you have to prove it by “clear and convincing” evidence, which is higher than the usual civil standard (“preponderance of the evidence”) but lower than the criminal standard (“beyond a reasonable doubt”).
- Finding prior art is very difficult and expensive. There are professional searchers, who cost a lot and are limited to patents (including foreign patents) and publications. Lots of needles, lots of haystacks (often foreign haystacks). For emerging technologies, that might not be good enough. You might need to track down actual researchers, who might be discussed in publications but don’t themselves publish, and investigate them.
- Then you have to get that prior art into evidence. If it’s in the form of a patent or a publication, this isn’t too difficult. But if it’s in some other form—often the case for emerging technology—you’ll need to line up witnesses and take their testimony the hard way, which is expensive, even if they feel like being cooperative.
Another reason why NPE’s preferred patents are so hard to invalidate is that they have early priority dates. Defendants are often confronted with patents based on a series of applications dating back ten years or more. Trying to find prior art more than ten years old in a form admissible at trial is … difficult. As one potential once told me (about six-year old prior art): “That was three technologies ago. I don’t remember what I was doing then!”
Oh, one more thing. You might not have enough time. Many courts (most notoriously the Eastern District of Texas) have “fast track” litigation schedules and special patent rules that make it hard to use prior art, and guess where NPEs prefer to file their cases?*
* Even in a court whose patent rules are friendly to the defendant, Sanyo had much of its prior art thrown out because it disclosed it too late. Maybe Sanyo’s lawyers were lazy, but I suspect it was because it just takes a long time to track down prior art in emerging technologies.
I’ll try to finish the policy discussion, and provide some suggestions for reform, next time.
Thanks for reading!