« The Sequester Is a Failure—Posner | Main | Patent Trolls—Posner »



Feed You can follow this conversation by subscribing to the comment feed for this post.

Douglas Bridges

While I disagree with several of the issues, your statement regarding patent length shows a fundamental misunderstanding of the patent system.

"Even pharmaceutical and biotech companies, the main examples where patents are clearly necessary to encourage innovation, usually do not need more than about a decade of monopoly power to encourage their very large investments in new drugs."

Patents do not grant monopoly power until after that are granted, but the length of a patent term begins when they are filed. Patents can take anywhere from 3 to 7 years to prosecute through the patent office. Thus, cutting the patent term from 20 years from filing to 10 years from filing cuts the effective monopoly protection from 13-17 year down to 3-7 years. Thus is a cut of 75% of effective patent term.

Brendan Cheong

And that hasn't even taken into account the time that pharmaceutical products need to get FDA approval before being able to be sold.

Brendan Cheong

The statement of why Einstein's and Darwin's theories can't be patented is also not accurate.


I posit that all software patents should be invalidated due to the fundamental underpinning that mathematical formula cannot be patented. All software, that is to say programming code, can be represented as one or more, in other words 'a collection', of mathematical formula. Just as you cannot patent f(x) = x + 1 you should not be able to patent y = x + 1. While my example is purposefully simplistic, the patents for such items as Lempel–Ziv–Welch file compression, H.264 video compression, along with various data encryption mechanisms (e.g. AES, DES, CSS, etc.) should never have been patented as they can all be perfectly described mathematically. In fact, the reason that all of these patented bits of code work so well is because their algorithms have been passed through the rigor of mathematics and proven to work.

"Look and Feel" should never have been patented, either. Trademarked yes. The prime example is the Coke a Cola bottle. Not patented. There are limits to even that... in that the look and feel should be limited to the whole object - not just some minor tweak (e.g. the bevel on the iPhone 5 case).

Copyright of software and it's subsequent system generated object modules (e.g. compiled DLLs)? Sure. Discussions about the never-expiring copyright fiasco are for another thread.

While the exact arguments of the speaker may be slightly off, it should not detract from their overall objective - that being that software patents are placing an overburdenous and undue drain on our economy. When the premise of what the idea (in this case software patents) are economically counterproductive one should indeed take a cold hard look at them and see if they are doing what they should be.




I am glad to see that software patents are being called out as a problem by judges familiar with the problems that software patents pose.

One issue that doesn't seem to get much attention in patent discussions is the presumption of validity. The author begins by saying that the patent system is too loose. That is to say, there is no check on the power of the USPTO. The reason for this is a very high evidence burden on defendants to challenge validity.

Some studies have demonstrated a 92% loss rate for patent trolls in court. Another way of looking at this is that 92% of the patents contested in court by patent trolls were invalid and should never have been issued.

Patent trolls are counting on the lack of invalidity defenses available to their targets. Lowering the bar for defendants will have a few obvious effects:

* If weak patents are easier to defeat in court because invalidity defense burdens are lower, trolls will use much more caution, or maybe they will consider something more productive to do.
* The USPTO, needs the law to back them up when they want to reject a patent application. The evidence burden as it is now, gives the patent office every excuse to approve a patent application and leave it to the courts. Lowering the bar for invalidity defenses will help the patent office to reject more patent applications, thinning the herd, so to speak.

Making it easier to challenge patent validity will create a check on the power of the USPTO and allows the public to nail the bad patents when they see them.


At some point, we need to take the bull by the horns and reform the US Patent Office in a way that will reduce the number of poor patents. I believe that we can accomplish meaningful reform, if we can somehow muster and sustain the political will to admit we have made some important mistakes. I suggest that these mistakes include:

1) More patents are not better than fewer patents. Patents are not Innovation. Patents are not Progress. Patents are simply grounds to file a lawsuit against an industry. More Patents are simply more grounds for more lawsuits. An occasional lawsuit might spur innovation. BUT LAWSUITS DO NOT PRODUCE. Lawsuits are parasitic on innovation and production. Reform must recognize that patents are dangerous monopolies. Reform must place hard limits on the number of patents.

2) Running the US Patent Office as a cost-recovery operation is a mistake. The US Patent Office is a very small, but critical component of the US economy. It's purpose was "..to promote the Progress of Science and useful Arts.." (US Constitution Article One, Section 8(8).) But, once the USPTO became completely cost recovery (Omnibus Budget Reconciliation Act of 1990, Title X, Subtitle B), that primary goal became overshadowed by the more pressing goal of securing funding via patent fees. The primary effect of cost recovery has been to promote the collection of patent fees. Reform is painful, but simple. Admit cost recovery is a failed experiment. Revert the funding model to the model used for the first 200 years. The USPTO must be centrally funded by the US government. Any collected fees should be returned to the US Government.

3) It is a mistake to organize the US Patent Office to create economic incentives to grant poor patents. Currently most of the revenue of the US Patent Office comes from GRANTING patents. See the USPTO FY 2013 President's Budget page 37: www.uspto.gov/about/stratplan/budget/fy13pbr.pdf "..More than half of all patent fee collections are from issue and maintenance fees, which essentially subsidize examination activities." A recent study by the Richmond School of Law found that the USPTO's actual grant rate is currently running at about 89%. In 2001, it was as high as 99%. See http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2225781 page 9. In 2001, it didn't matter if an application was overbroad, obvious, trivial, a duplicate, or unreasonable, they ALL got granted. Things haven't improved much since then. Reform could come in many forms, but the simplest and most reliable would be to eliminate and unify the Patent office fees into a single filing fee. This fee would provide no guarantee of receiving a patent, only a guarantee that your patent would be considered. This would free the Patent Office to be able to deny poor patents. The filing fee should be high enough to discourage spurious patent applications.

4) Scaling up the Patent Office to produce more poor quality patents is a mistake. Currently, we expand the number of patent examiners based on demand. See the USPTO FY 2013 President's Budget, page 60, Gap Assessment: "Meeting this commitment assumes efficiency improvements brought about by reengineering many USPTO management and operational processes (e.g., the patent examination process) and systems, and hiring about 3,000 patent examiners in the two-year period FY 2012 and FY 2013 (including examiners for Three-Track Examination)." Again, the assumption is, more patents are better, even if it means decreasing examination, and increasing the number of untrained examiners. Poor quality is an inevitable result of this patent process. Reform must tightly control and limit the number of patent examiners.

5) It is a mistake to grant all patents that meet minimum standards. A review of the last couple decades changes in the patent approval criteria will reveal that the minimum standard for granting a patent has consistently shifted downwards. We must abandon the idea that any patent that meets minimum standards is granted. Over time, the standard always degrades. Reform is easy. We rank Patent Applications according to an agreed measure of quality, and only grant the top few percent. Over time, the pressure will be to improve the quality of patent applications, instead of degrade them.

6) Finally, I suggest that it is a mistake to allow patent applicants to modify or extend their patents after submission. This complicates the patent pipeline. It facilitates ‘submarine’ patents. It enables capturing Standards. It also enables gaming the patent system. Reform must simplify and reduce the patent process. Patents should be quickly evaluated. Most should be denied. If an applicant wishes to modify a denied patent, they should alter it, resubmit, and pay a new filing fee.


As common with economists, everything is reduced into a cost/benefit analysis, without any understanding of the actual nature of the domain in question.

All patent litigation, not just software patent litigation is expensive. That problem arises from the intersection of complex technology with hyper-complex rules of claim construction, infringement and validity. The litigation cost analysis of Bessen has been shown to faulty. http://goo.gl/gW5xHL.

Second, there is no solid evidence that eliminating software patents would encourage innovations. I've never had a software client who decides not to innovate because of fear of patents (yes, a biased anecdotal sample). Those folks that assert that such concerns stopped them are far outnumbered in real life by those who POR: Press On Regardless. And that's because it's the latter who are actual innovators and risk takers, not "If only..." talkers.

Third, there is no principled distinction between software as an artifact of human creativity and any other form of innovation. Further, there is no principled distinction between hardware and software. Indeed, the vast majority of hardware, from smartphones to TVs to the computers that run your car are first designed in software, and then implemented in hardware. The boundaries between hardware and software implementations are highly permeable.

The one remark that really demonstrates the Becker does not understand software or technology in general is this: "creators of new software may then try to protect their innovations through secrecy."

The vast majority of value in software comes not from some deeply embedded algorithm that can be protected by trade secret. Rather, it comes from the creation (invention) of new functionality that has immediate and apparent value to the end user, whether that's a consumer like you and me or an enterprise. It's just as much an invention to enable users to use their smartphone in a new way or a new application as it is to invent a new algorithm for processing cell tower signals. This is true of new machines as well: as soon as it leaves the manufacturer it can be easily reverse engineered and copied. A modern technology company with a consumer facing product or software simply cannot rely on trade secret protection: if they were smart enough to figure out the solution to a problem, their competition is likely to just as smart.

But then again, I don't expect an economist to know much about the real world. After all, for an economist, the solution to any problem is to assume a can opener.

Chuck Crane

As an engineer I am constantly amazed at the ignorance of lawyers and judges in this realm -- like a clutch of re-animated medieval schoolmen trying to reason their way to understanding things far beyond their ken by quoting Aristotle and making analogies from their trivial, superficial knowledge. Pathetic. You don't have a clue, so butt out. I'll take economists any day. At least they understand advanced mathematics and are capable of constructing models that actually work and can be empirically verified.

The major issue with software patents can easily be illustrated with a reductio ad absurdam. This involves simple math, however, so I doubt any lawyer will understand it. Here we go anyway. Assume that the population of potential inventors is very large. If the probability that a randomly-selected member of this set of potential inventors will come up with an invention is some finite number, however small (say one in google), then the number of potential inventors who will produce the invention approaches infinity as the population of inventors grows indefinitely large. Under current patent law, the first inventor of the gazillion to file would get the patent. However, given the Planck time unit, it would not be possible to distinguish who was first to file, so you would have a gazillion inventors filing at the same time. Robed dufus in the Galactic Court would have to sort it out, I guess.

Practically, this scenario already exists. With 50 million software developers in the world, something obvious to only one in 10,000 is obvious to 5,000 people. Does it make sense to give the first one of these people to file a patent application a monopoly for 20 years? Of course not. (It will be readily apparent that a test of "ordinary skill in the art" is especially ludicrous under these conditions.)

The comments to this entry are closed.

Become a Fan

May 2014

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30 31