How "Patent Trolling" Taxes Innovation

Applying for a patent is expensive. Fees can exceed $25,000, and most applications require at least a couple years of effort. We might expect that anyone considering applying for a patent would be fairly certain of the merits of their case for one. And yet, of the patents granted by the U.S. Patent and Trademark Office (PTO) that are subsequently litigated, 40% are declared invalid in court.

A court’s declaration that a patent is “invalid” means it should never have been granted in the first place, usually because the invention has been done before, or because it’s obvious to anyone familiar with the patent’s particular scientific or technical field. So why do so many people spend so much time and money filing for patents that are ultimately declared invalid?

Uncertainty is one reason. Applicants can’t always predict whether the patent will later be struck down in court. New evidence may come to light that the “invention” isn’t really new; or that the basic idea underlying it is widely understood by those in the relevant field. So misperception, aided by the well-known phenomenon of “optimism bias,” may explain a portion of this puzzle.

Another answer is that seeking even dubious patents is a gamble worth taking. The PTO receives over 520,000 patent applications a year. That’s about 1,425 applications a day. The agency’s approximately 6,500 patent examiners are overburdened; on average each patent gets only about 16 hours of review. Virtually all experts agree that this is nowhere near enough time to properly assess a patent. More fundamentally, patent examiners are rewarded for processing applications – and the easiest way to clear a file is to approve a patent. The PTO’s examiners don’t get paid to say no. They’re incentivized to approve.

But there’s a third, more interesting reason so many people seek patents that have a high probability of being found invalid. As it turns out, invalid patents may be worthless as innovations, but they can be very nice bargaining chips.

Any patent the PTO grants enjoys a legal presumption of validity. To overturn this presumption, defendants in a patent infringement case must prove the patent’s invalidity by a high standard of “clear and convincing evidence.” It takes a lot of lawyer fees to do that. And the result is that defendants often pay rather than fight, even when they think they could ultimately have the patent invalidated. More than 97% of patent infringement suits are settled before trial.


In this environment, brandishing invalid patents can be a good business strategy. This behavior, called “patent trolling,” gets a lot of attention today, but it’s been going on for more than a century. In 1895, George Selden obtained a U.S. patent for putting a gasoline engine on a chassis to make a car. The patent clearly should never have been granted: the idea was so obvious that many people worldwide thought of it independently as soon as the first workable gasoline engines became available. Nevertheless, Selden brandished his patent, threatened nascent carmakers with suit, and collected hundreds of thousands of dollars in royalties.  Selden didn’t contribute anything meaningful to the development of the automobile, but his patent abuse made cars more expensive for years, until Henry Ford, who refused to license from Selden, finally busted the patent in court in 1911.

The patent system exists to encourage people to invent – to add, in Abraham Lincoln’s words, “the fuel of interest to the fire of genius.” But as this post suggests, there’s a dark side to the patent system. Just as good patents advance innovation, bad patents retard it. And even bad patents can be valuable to a patent troll. Every penny that goes to license an invalid patent, or settle a meritless lawsuit, is a tax on innovators.

Eric M. Jones.

The real reason was that in 1981, President Reagan forced the Patent Office to change their direction and grant patents for virtually everything. This was a real sea change. He said, "Let the courts decide".

Ostensibly, there was a giant backlog and this seemed to clear it up. But suddenly so many dubious patents got passed that I had the task of deciding what patents the company I worked for could just "walk over" --that is-- ignore because they were utterly indefensible.

Lucian Armasu

The more I hear about Ronald Reagan, the more I realize how much damage he has done to USA. Was he just a puppet for others?


I found a patent in which all the claims had been in the public domain for years. All but one were in college text books and the other was in a technical paper pre-dating the patent by 10 years.

The patent was filed by a respectable attorney who should have known better.

Most people just ignored the patent and even used the inventor's trade name for the product. I know of one case where the inventor threatened the infringer. The latter ignored the threat and the matter never got to court.


If the patent is so obviously invalid there is no risk for you to reveal what it was specifically. Since you haven't, your story reads like an urban legend and I assign it as much credibility.

Mike B

An east fix is that if a patent is found to be invalid, the holder of the patent will be liable to return any infringement awards to those who paid them initially.

Michael Peters

That's not really a fix. That's like saying a robber is only liable for what he stole and faces no other penalties. Of course they need to give back what they took, but they need stiffer penalties on top of that. Maybe interest on that money?

David Leppik

While I agree with the gist of this article, the introductory paragraph contains a little statistical nonsense. The fact that 40% of litigated patent cases declare the patent invalid has nothing to do with how many patents overall are invalid-- only the those that make it to litigation. Apples and oranges.


Selection bias -- invalid or questionable patents are much more likely to attract lawsuits in the first place and avoid pre-judgment settlements later in the process.

(In case it is not clear, I agree with you.)

Todd Sullivan

The premise of this article seems to be, 40% of patent infringement cases end with the patent being invalidated, so 40% of all US patents should not have been issued. Isn't it just as likely that the cases settled because a strong argument for invalidity did not exist? many patent validity determinations are made on motion before trial and, thus, settlement was achieved after the defendants determined their invalidity arguments would not be successful. If 97% of cases settle, then only 1.2% of initiated infringement cases end with the patent being invalidated. Isn't 98.2% of patents being validated by an inter partes system (litigation) indicative of an effective patent office?

It should also be noted that the US Patent Office is allowing less than 50% of applications to issue, at the very least an indication that Patent Examiners are not incentivized to rubber stamp approval for every patent application that hits their desk.

Too many of these Freakonomic blog posts are nothing more than an author's opinion mixed with a couple convenient statistics to rationalize the opinion, rather than the thoughtful analysis developed in the Freakonomics books. These blog posts tarnish the brand.



First, there is a selection bias in the finding that 40% of patents that are litigated are found invalid. Defendants are going to be more willing to litigate to completion over patents they suspect have a reasonable chance of being invalidated (or that they believe will not be found to be infringed). This is especially true where the damages are steep.

A more interesting statistic might be to look at the results of patent reexamination process. Few patents come out of that process unscathed.

Second, patents are not special in this sense. Asserting dubious claims of any type (discrimination, breach of contract, etc.) when available can also be a lucrative strategy to obtain a settlement of the claim just South of the nuisance value.


Could you provide a citation for the 40% declared invalid and the 97% settlement rate, please?

Faye R.

Although I generally agree with the overall argument in this article, the definition of patent trolling provided here is off-base from the standard use of the term. A patent troll is traditionally an entity that enforces patents it does not itself practice -- an entity that does not invent, make, manufacture, or create but rather just buys patents from others so it can sue alleged infringers and extort settlement and licensing fees. Many argue that this pejorative definition itself is too overarching as there can be legitimate activities by such entities, but the definition provided here is off base. The entity here described is related to a patent troll -- often patent trolls buy patents of questionable validity to enforce them against unsuspecting companies that independently developed their product -- but it is not the same. There are certainly legitimate businesses that practice their own patents but still obtain invalid patents in an effort to force costs on competitors. These are not patent trolls by the normal definition.


Pete Griffiths

I believe the writers introduced a new definition of 'patent troll.' When they say " ...brandishing invalid patents can be a good business strategy. This behavior, called “patent trolling,” gets a lot of attention today, but..." this suggests that "patent trolling" is characterized by "brandishing invalid patents." But this is surely mistaken. There may well be patent trolls who behave in such a manner but there is nothing about patent trolling that requires the patents be invalid. They may well brandish perfectly valid patents.

A more accurate definition of a 'patent troll' follows:
"Patent troll is a pejorative term used for a person or company that enforces its patents against one or more alleged infringers in a manner considered (by the party using the term) unduly aggressive or opportunistic, often with no intention to manufacture or market the patented invention."

The above definition includes the key elements that better characterize those so referred to:
a) unduly aggressive or opportunistic
b) no intention of manufacture or market
And these characteristics do indeed result in the term being used pejoratively.

I hope this helps.


Michael Sherrin

While 40% of patents in court are found invalid, the same study you cite finds that on average, these cases cost more than $1 million per patent. For small companies, this is an impossible sum to challenge and because the patent was originally considered valid, there is no reason to award legal fees to the defendant. And this doesn't count the massive amount of patent settlements - since it can be so expensive to defend against a patent case and even more expensive to pay tens of millions of dollars in infringement fees, many patent cases never reach the court.

Consider that Microsoft currently makes more money from Android phones made by several companies than it makes from Windows Mobile phones. This is because of the patent licensing deals it extorted from companies like HTC. Oracle is now pressuring companies to pay it similar licensing fees. This is a huge tax on innovation.

And if 40 percent of patents are found invalid in the court, then why do we make is so hard and expensive to declare them invalid (let alone why are they getting approved in the first place). The Supreme Court, even after KSR v. Teleflex expanded the definition of what is considered obvious and thus not patentable, has stopped reigning in patents. It's most recent ruling in Microsoft v. i4i (funny how Microsoft wants to limit patents here but abuse them again competitor Google), the Court said patents still benefit from assumed validity even when 40% are found invalid when they reach the courts. This means Microsoft still has to pay a company that produces no products $290 million because of a tiny feature in Microsoft Office almost no one uses (how often have you used XML editing in Office).

And consider how Lodsys, another company producing nothing, is abusing its patents. It has a patent on in-app purchasing, something completely obvious to anyone knowledgeable in the art along with tons of prior art, already collects licensing frees from Apple but is now suing dozens of small developers, most of whom could not afford the $1 million needed to defend themselves in court.

And let's remember, there is no independent inventor defense for patent infringement. Meaning even if I invent something on my own, if someone patented it first, I would need to pay them for the privilege of using my own invention. Only a tiny percentage of patent cases are considered willful infringement. And the new patent reform legislation makes it so the first person/company to file for a patent will get the patent. Meaning even if I invented something first, someone else can patent the invention.

The patent system has become nothing but a tax on innovation. The risk and threat of litigation limits how companies can offer products and increases the prices for consumers. There are far more options to incentivize innovations, from government and educational grants (universities only started being allowed to patent their research in 1980 and its ended up costing most universities money) to prizes (like the X Prize).



Economists need to push this issue and push it hard. Put the brokenness of patents into the core curriculum. Talk about it on cable news. Make it a mainstream. Patent issues are the number one problem with innovation in America, and it's spreading to Europe, Canada, and many other places.


While I agree that the patent system is screwed up and needs some major overhaul, their statistic implying that 40% of patents are invalid is absurd because it only includes patents that actually go all the way through the court system. People don't give in to patent infringement suits only because it is expensive to fight them, but often because the infringement is real. Because it's so expensive to fight them, companies will only do so if they feel they have a reasonable chance of winning. In fact, a 50/50 split is about what I would expect if experts on both sides feel that it is worth fighting (assuming that the two sides have equal access to expensive lawyers).

If they want to do a real investigation into the ratio of valid patents vs invalid patents, they should do a statistical sampling of all patents and do a more thorough review of them to see if spending more time would turn up new reasons to consider it invalid. I'm not going to pretend to know what that statistic would be. Perhaps it would be higher because most of them are owned by people who are no longer interested in enforcing them (because they are now considered invalid). Perhaps it would be lower because most of them are so obviously valid. Keep in mind that most patents are extremely specific. The ones that you hear about on the news that are broad and ambiguous are the exception rather than the rule (hence being newsworthy). Many patents are meant to protect a very specific product from exact duplication, not to block a competent competitor from entering the field entirely.



It is not a tax on innovation. It is a free market response to protections for innovators. Overall, the post is complaining that patents are issued and that's all. So what? You want them to issue fewer patents. So would I, but that requires a lot more money, more staffing, more people to read and research. You going to fund this?

Much of the problem is court decisions that have defined new, useful and unobvious in ways that don't make intuitive sense. Heck, Amazon's one-click patent is an example of something obvious. But then in court unobvious means something different. That goes back to basic inventions like John Deere and plowing. I'm not kidding. When you add a flange to a plow, it makes the plow more efficient. Obvious, huh? Yes it is, except someone had to invent this obvious thing. Courts don't apply intuitive notions of obviousness.

You don't say how patent examiners are supposed to deal with new technology using ideas developed for plows. One-click happens to be obvious and yet Amazon did it. Why say that's legally obvious as opposed to intuitively but not legally obvious? If examiners start denying patents, they will tend to stifle innovation as well because then small companies won't have the protections of the patent laws, we'll have to rely more trade secrets and more innovations will not come into the public sphere.



It works both ways. In the movie, "Flash of Genius," the inventor of intermittent windshield wipers had to fight against almost insurmountable odds to obtain rights to his invention.

Just as patent trolls rely on high lawyer fees to discourage any contest of their claims, so, too, do large companies rely on their armies of lawyers to delay, confuse, obfuscate, and detour claims by wronged inventors.

That's what happens when people play dirty. You HAVE to rely on the courts...but the lawyers are there, in some cases, to make sure the case never gets to court.


As both an inventor and a patent attorney (now retired) I have seen the system in operation for many years, from both sides of the fence.

To take a step back and see the broader picture, consider the following:

Patent protection is authorized by the US Constitution, and overall has served this country well over the years. I am reminded of the following description of our system of government, which could equally be applied to the patent system: "The worst system in the world, except for all the others". In other words, where would we be without a patent system, and what would be the alternative?

Most patents are not litigated, and patents widely believed to be valid are generally licensed or respected by competitors. There will always be a "tipping point" where reasonable men will differ, and it is not surprising that the outcome of litigation is divided in such cases.

The key point that seems to have not received much attention here is that patents by and large do serve to promote the advancement of science and engineering, and provide property rights to inventors to justify sometimes considerable R&D expenses. Who could justify mega-buck research programs absent some form of protection?

In sum, the system may not be perfect, but it works, and (like our society) is constantly evolving.



I'd love to see the math on hiring extra patent examiners from tax money versus consumer costs of patent trolling. Difficult to measure though.