Patent Thickets Reduce Innovation

In a report for the UK Intellectual Property Office, Bronwyn Hall et al., find that patent thickets exist in a number of technological fields and that thickets reduce innovation.

We find overwhelming evidence in the literature that patent thickets arise in
specific technology areas….

Our main contribution in this study consists of an empirical analysis of the
effects of patent thickets at the European Patent Office on entry into patenting by
UK firms….Our results suggest a substantial and statistically significant negative association between the density
of thickets and the propensity to patent for the first time in a given technology

As we find thickets to affect entry negatively, there is a strong indication that
thickets represent some kind of barrier to entry in those technology areas in
which they are present. However, we must emphasize that the simple finding of a
barrier to entry created by patent thickets is not proof positive that reducing that
barrier and increasing entry would lead to welfare improvements in the
innovation/competition space. Rather it is the existence of evidence that the
presence of thickets reduces entry combined with the large literature we have
reviewed that shows that currently patent systems do not work as well as they
should. This literature documents quality issues with patents in technology areas
affected by patent thickets, a large decline in the relationship between R&D
spending and patenting in some sectors and a substantial increase in resources
devoted to patent litigation leading to the partial or complete revocation of
patents in areas identified as prone to thickets.

I like their understated conclusion:

All of this may lead one to the conclusion that the operation of the patent system could use some improvement.

In other words, see the Tabarrok Curve.


You might want to consider that countries with a small portfolio of intellectual property assets, or those which are now seeking to expand into technology fields where there are established firms in other countries, would complain about intellectual property assets protecting the property of foreigners.

One man's thicket is another man's carefully tended garden earned with years of blood, sweat and tears threatened by a neighbor eyeing those tomatoes in the garden.

Except that your analogy is for property. You can still buy the land next door and copy the garden exactly. Patents are an artificial monopoly of infinite scope. It would be like a guy who grows tomatoes getting the government to say that noone can grow tomatoes even if they replicate all the blood sweat and tears.

This professor, who teaches at Berkeley (liberal) CA, also makes a distinction in his papers between a patent thicket and a patent fence (see his paper at and further makes the interesting points:

1) if you have a big lead over your competition, no need to patent--just rely on trade secrets. How does Alex T like that? In analogous fashion, the reason the M-16 was never patented is that the way it fits together is such that it's trade secret, since it's hard to do and patenting this would give the enemy (literally) ammunition to replicate that. Again, Alex T must like the fact that Roman concrete was lost to the Middle Ages, and rediscovered, since it was trade secret. Ditto "Greek Fire" a naval chemical weapon--lost forever (though various theories exist on what it was).

2) From the paper cited above, B. Hall et al state: "The evidence available from various firm‐level surveys, which is reviewed below, suggests that on average, firms rely more on informal than formal IP to protect their inventions, and that most firms use no IP protection at all. " - and give, as one reason why, the fact that non-innovators have no need to patent. If you copy innovation naturally there's no need to patent it, as it is not yours.

My feeling is: patents are like traffic lights that promote faster and better travel times on roads. We would like to not obey traffic lights, which can be a hassle to wait for, as they do not obey in the Philippines (I've yet to see a single traffic light here--wait, did I see one in Manilla? no, I think not--rather they use traffic cops or more commonly it's chaos), but the result is chaos and paradoxically slower travel times. Likewise, it would be cool if we can get all our technology and copyrighted materials for free--as I often do at the result is, long term, like rent control in NYC--you end up stymieing innovation. Unless you believe the idea that people invent out of love (which does have historical basis, for example nearly every Nobel prize winning discovery), and, more importantly and controversially, the idea that nobody ever is given incentive by patents (including employers). I think there's room for smarter patent policy to encourage innovation. I honestly feel that with better patent policy we'd be way ahead now on the technology curve. It pays more in the USA to be a 'manager' (chief) rather than 'worker' (indian). A lot more. And ditto to be a rent seeker, second mover, and pie-divider rather than inventor, first mover, and pie-expander. Then, given this, economists wonder why there's a Great Stagnation. Duh.

Is your 'good' patent system dependent on competent government?

Andrew': "Is your ‘good’ patent system dependent on competent government?" - yes, or a competent court, even private court (special magistrate) to adjudicate whether a granted patent is really good or not. I'm sure there's a public choice argument out there that this is unrealistic, but I rather think that somehow government can get it right.

That depends. We can 'get invading tinhorn backwaters right' even though we shouldn't be invading them in the first place. What does "more IP" mean? Do more examiners yield more or fewer patents? Do more or fewer patents and of what type increase or decrease innovation and in what industries, respectively? Unfortunately, with the division-of-labor-FAIL that is government, we have to know what we want before we cut loose the locusts.

I'd like to see free or even paid (by fees) defensive patents.

I believe that the Coase theorem says that if you set up a law, and assume no transaction costs, or low ones, then it does not matter how the gains are apportioned insofar as society goes. I could for example envision a US Patent Office something like the German or Japanese "laid-open" patent applications, which are unexamined until such time there is litigation (akin to how copyright applications are handled in the USA--which are not filed until such time litigation is foreseen). The point is: more and better patents would IMO encourage more and better innovation. As for defensive patents, Google "SIR patent" and yes we could make costs of publication lower.

What about $10,000+ transaction costs per right to hire a lawyer and get stonewalled by the lawyers/courts? What does the Coase theorem say about that?

Applications for copyright registration are and should be filed well before risk of litigation becomes apparent. Also you can do a defensive publication for free.

One man’s thicket is another man’s carefully tended garden earned with years of blood, sweat and tears threatened by a neighbor eyeing those tomatoes in the garden.

Yes exactly.

The "patent thicket" is acknowledged to encourage innovation in pharmaceuticals. Why would a company invest millions to develop and test a drug if it is trivial to create knockoffs?

But the same thing is true in tech. It took Samsung 5 yrs to catch up to the iphone. But for many innovations it takes much less time, so that companies will have no motivation to invest in R&D due to the low ROI.

It does often seem that some people both decry the very concept of patents and lack of investment in R&D.

Personally, the US patent system if far from perfect, but we should prioritize fixing the copyright system. Which, in my opinion, is almost a joke at this point.

Pharmaceuticals is a special case as IP is surely trivial relative to the requirement to disclose during FDA approval. I'd like to see the cost of synthesizing a small molecule compared to the costs of the medical testing. I also wonder how much of the synthesis is amped up because that is how molecules are monetized. For example, it will be hard to monetize resveratrol, but a synthetic analog could be patented.

The things I'm concerned about are the inventions that are the equivalent of Oklahoma Sooners just getting to the undiscovered territory first and not really contributing much and then stifling innovation with roadblocks. The goal of innovation is to promote the innovation. Since there isn't really an inalienable right to not be copied, the government's primary job is to determine if what they are doing is promoting or hobbling innovation with its efforts. The IPod for example was nice, but technically it was mostly just a copy of all the preceding technology in MP3 players. And that is fine since it did bring the innovation to more people by getting the design right. It's just that they have some nerve trying to set up patent thickets around some of the same things they borrowed from others.

The same thing is absolutely not true in tech. It is inconceivable that you could make such a claim with a straight face.

"The same thing is absolutely NOT true in tech."

The development of tiny communications and battery technology that is within every smart phone device was developed through billions of dollars of R&D, and consumers are willing to pay for small, well working cell phones, even before smart phones came along.

Apple and Google came around with a great device design to lay on top of that technology. They are now both really raking it in. Is it a huge surprise that someone that finds themselves in such a position would try to de-emphasize the importance of the patent system?

I find it ironic that Apple would sue Samsung for infringement of Apple's design of Samsung's technology.

I find much of the criticism of the idea of patents to be short sighted.

Considering that the purpose of IP policy is almost entirely pragmatic (it is not a real property), the first order of business for the government should be determining the precise shape of the Tabarrok curve for various industries.

I think that's almost impossible in practice. How'd you go about it?

Experiments would be one possibility. But if it can't be done then we'd better pack it in. By rights, until we figure that out any time someone were to attempt to lecture me on science or data I'd get to poke them in the eyes.

Experiments? Maybe. But I don't know how you'd design them when Alex's axes are quantitatively ill defined entities: "innovation" and "patent strength".

Maybe patent-duration would be a surrogate for strength.

Andrew-poke-the-eye arguments are generically silly. Each time you don't agree with something you propose a convoluted measurement and then the world must stop till you see your statistically significant, randomized controlled, double blind study?

PS. Though in this case I agree with you on the underlying IP issues. Current Patent law sucks.

Well it should be pretty easy qualitatively speaking when Ray thinks innovation will increase with patent strength.

No, my poke their eye argument isn't silly. As Ray Lopez has correctly pointed out, innovation is more important than anything the government does, especially when it is using national defense as special interest offense. We should really be deposing and lynching. The "Science" that The Left keeps harping on is utterly insignificant relative to innovation.

" the world must stop till you see your statistically significant, randomized controlled, double blind study? "

No, keep up, Rahul. The Left makes such statements. Not I. They for example do a single shoddy study on daycare and then feel they have the authority to mandate universal daycare. You only need to actually figure out what you are doing if you are going to screw everyone over with laws that potentially screw up the source of half or more of the world's innovation, mandate daycare for hundreds of millions, or eliminate federalism and trial-and-error in healthcare markets. And that these people claim they have science on their side deserves a lot more than my ridicule.

Here's an easy indicator: The job of the government should be 'general welfare' and not funneling business to special interest. What is just the NIH budget that funnels into pharmaceuticals compared to attempts to determine the effects of the patent system on innovation. This post is an example of such an attempt. I don't have great expectations. Attempts are good enough for me.

By the way, your previous comment seems to place some equivalence between other people relying on the "science" of a single study to justify coercion of millions and my request for ANY evidence to justify coercion of millions. I respectfully disagree!

"So it is striking that there seems to be no empirical studies or analyses providing conclusive evidence that an IP system is indeed worth the cost."

If I and the author are wrong, then that's fine. I don't claim to be an expert in every goddamn thing the government is supposed to handle.

It's an excellent paper. In case it helps for teaching, i refer my students to this link which tries to summarize the paper,

What's unfortunate about this paper is that it does not distinguish between the acquisition of a patent by your own reasearch and development, and the acquisition of a patent from a third party to add to your patent portfolio.

The antitrust laws cover the acquisition of patents from third parties: they are simply the acquisiton of an "asset" subject to Section 7 of the Clayton Act which forbids acquisitions which may substantially lessen competition. In addition, some licensing restrictions and acquisitions via licensing are likewise covered under Sections 1 and 2 of the Sherman Act, and the Clayton Act.

But, development from your own efforts are not: even if they create a thicket for someone else. Wonder why?

Can you cite a case where Antitrust laws were used to prevent a company from acquiring a patent?

Not doubting you but curious.

Rahul: please see, for example, "The Particular Antitrust Concerns with Patent Acquisitions," Competition Law International August 2012, pp. 30-38 (pdf link).

There may not be many cases where DOJ or a court has outright stopped a company from acquiring a patent (or, in a lot of the cases, a portfolio of patents, for example from a company entering bankruptcy), but there apparently have been cases where a deal is announced, and DOJ conducts a study and "recommends" that the contract, in its announced form, is "troubling." So, for example (from 2001):

"DOJ’s suit to prevent 3D Systems, Incorporated (3D Systems) from acquiring DTM Corporation (DTM). Absent the merger, neither 3D Systems nor DTM individually had the ability to block a competitor from the market but when the patents were combined, 3D Systems would hold a blocking position and have
the unilateral ability to foreclose competition. The DOJ therefore required 3D Systems and DTM to license their rapid prototyping patents to a third-party competitor that would compete in the US market as part of the settlement."

Sure, this happens as part of relief in a merger case, particularly in the drug industry, where co A acquires company B and they have competing patents. See Pfizer case (you can look up), some cases involving biologics relating to arthritis meds, etc. In addition, the government may require compulsory licensing where there is a threat of foreclosure. See, merger cases involving cable and broadcast companies (Time Warner). In addition, if the patent is acquired, and its fair market value is greater than around $50 million, the acquisition of the patent has to be reported under the Hart-Scott-Rodino Act. Finally, if a company, particularly in a patent infringement case, is claimed to be an illegal monopolist, the infringer would raise prior patent acquisitions as acts in furtherance of a monopoly or monopolistic acts, and challenge that way as well.

Actually, rather than block an acquisition, the mechanism I've seen several times is to require the acquiring firm to license out the acquired technology.

FWIW, my broad suggestion for fixing patent problems is this:

Will he get an economics "nobel" prize for having a curve named after him?

Comments for this post are closed