Software Patents

Excellent column by Gordon Crovitz in the WSJ on patents and the prospects for reform:

Today’s patent mess can be traced to a miscalculation by Jimmy Carter, who thought granting more patents would help overcome economic stagnation. In 1979, his Domestic Policy Review on Industrial Innovation proposed a new Federal Circuit Court of Appeals, which Congress created in 1982. Its first judge explained: “The court was formed for one need, to recover the value of the patent system as an incentive to industry.”

The country got more patents—at what has turned out to be a huge cost. The number of patents has quadrupled, to more than 275,000 a year. But the Federal Circuit approved patents for software, which now account for most of the patents granted in the U.S.—and for most of the litigation. Patent trolls buy up vague software patents and demand legal settlements from technology companies. Instead of encouraging innovation, patent law has become a burden on entrepreneurs, especially startups without teams of patent lawyers.

…A system of property rights is flawed if no one can know what’s protected. That’s what happens when the government grants 20-year patents for vague software ideas in exchange for making the innovation public. In a recent academic paper (pdf), George Mason researchers Eli Dourado and Alex Tabarrok argued that the system of “broad and fuzzy” software patents “reduces the potency of search and defeats one of the key arguments for patents, the dissemination of information about innovation.”

…For now, the best prospect for real reform is in the Supreme Court, which earlier this month agreed to hear CLS Bank v. Alice Corp., a case about whether a bank’s computerized process for settling transactions via an escrow can be patented. A judge on the appeals court noted this idea was “literally ancient,” developed during the Roman Empire, and should not get a patent now just because a computer is involved.

I think it is too early to call CLS Bank v. Alice Corp. an obituary for software patents as The Economist does but real patent reform is stronger than I thought it would be even 6 months ago.

Addendum: Here is my 2 minute video on some of the problems with patents.


Hasn't the 1980 to present been one of explosive growth in software, technology and engineering? How much faster could innovation move?

Yes but most of the growth has come from companies without a patent portfolio. Interest in patents seems to come later when the companies are bigger and the innovation slows. Microsoft managed for nearly 2 decades with almost no significant patents. Then around the turn of the century, with the PC market growth slowing and MS looking for all the world like the big-iron behemoths they helped to defang, they suddenly cast around looking for patents to use against upstart rivals. Apple and Google have recently followed suit in an almost identical way.

In the mythology of patents (and indeed in all intellectual property law) is that without state-sanctioned monopolies none of our modern world would be possible. And yet somehow it was.

To be fair, patents are critical to start-ups, these days anyway

Thats only because of the threat of patent litigation. Historically, start-ups never even looked at patents until recently.

What is your evidence of that? My personal experience is the opposite, but actual data would be best.

Absolutely. If you're not already established, you need at least one patent to raise venture capital. Big companies like Apple, IBM, Microsoft, etc. want to diminish the rights of patent holders so they can take all the good ideas made by small companies without paying for them.

Software patents and business method patents are a special case and perhaps should be reined in. It would be a shame, however, if the effort to tighten up on these cases also limits the rights of holders of traditional patents.

What patent did myspace have that facebook did not have, given myspace was more successful than facebook and thus deemed the winner back before facebook was the winner?

If myspace had a patent, maybe they would still be a contender. They had the advantage of being first, and somehow managed to squander it. Altavista would be another example.

Sometimes being first is enough of an advantage. eBay is an example of that. I don't think eBay has any patents that would prevent a competing service. They've got momentum, and that's enough in that space, for now.

If you wanted to do something new and you wanted venture capital, the first question is how do you prevent me-too competitors. A patent can answer that question. Moving quickly to seize the space (as eBay did) is another way to do it, but then it has to be something you can start on a shoestring budget.

"you need at least one patent to raise venture capital"

That's not what I hear when talking with people on both sides of the software (retail web, mostly SaaS) startup space. What leads you to that statement?

The venture capitalist I talked to about my idea. He was very interested until I found a patent from the 1980's that anticipated the major features of my invention.

Without patent protection, what have you got? Moving quickly and taking over the market niche first can give you an edge, so it can be a substitute, but the guy I was talking to wouldn't even consider touching something without a patent.

Mark, doesn't your own anecdote belie your claim.

You failed to get VC capital *because* of some dusty patent lying about somewhere.

Web SaaS startups? Or retail web? That's not my experience.

Patents are certainly critical in many other industry startups, but if you're trying to become the next big web retail win, patents (held by the startup) carry little weight in terms of convincing investors to invest.

Here's a fairly recent pertinent study

"To be fair, patents are critical to start-ups, these days anyway"

It takes three years to even get an initial examination of a patent nowadays, so patents don't offer startups on the usual schedule any benefit at all. By three years in, you already know if you're going to be a success and raising an A round has already either worked or failed.

I think the real question is how to come up with a reform that trumps on patent trolls while protecting the rights of startups where innovation is concerned. Complicated...

Notice how much software is being developed *outside* the USA nowadays, especially when it comes to open-source.

How is this relevant?

The rest of the world was going to be producing more software no matter what the US did. Yet the US still produces vast amounts of software. There has never been a better time to be a software developer in the US. IPhones are assembled in China but the majority of the software on them is created in the US.

I believe that this is relevant. The USA used to have much more edge against the rest of the world in technological innovation, and it still has that edge in some sectors (weapons, aircraft). Not so much with software anymore. There are small (population-wise) countries such as Sweden and Finland which punch significantly above their weight.

After all these years, Silicon Valley is still software's Hollywood. It has very strong cumulative advantage. Nowhere comes close.

Sure, good work can be done elsewhere, but I'm not sure anyplace else can lead (anytime soon).

That still doesn't say anything for or against patents. Do we need more patents to protect against international competition? Do we need less because smart kids are turning down $220,000 jobs in Silicon Valley to work for $100K in Finland?

Notice how much software is being developed *outside* the USA nowadays, especially when it comes to open-source.

I have no idea. Do you have any decent source material?

China certainly does a lot for internal consumption. But does any country rival the US as a software exporter?

'But does any country rival the US as a software exporter?'

How do you measure that, though? Because Linux comes to mind as being a very serious piece of software, but nobody 'exports' it at all.

Such collaborative work is the essence of the Internet - which nobody has 'exported' either.

The funny thing is, the U.S. only thinks it is exporting much in the way of software these days. Nobody is 'exporting' Android - and nobody is buying Windows Phone much (or Windows 8, for that matter). However, Microsoft is earning money in a way that can be measured by 'exports,' while Google is steamrolling Microsoft (and Apple to a major extent, though there is no question that Apple has extensive experience in surviving as a trailing number 2) in a way that is not measured by the concept of software revenue.

It is a big world out there - the U.S. is a fairly unique, but limitid part of it. And due to things like software patents, the world is interpreting it as damage, and routing around it. (I have talked to Alan Cox, a Red Hat employee at the time, about how he planned to never visit the U.S. - he was completely serious about it. And there is nothing to make me think he has changed that perspective since.)

Where does Linus live?

I am very pro international development, I am just amazed that how long Silicon Valley has dominated as a place, maybe even becoming stronger in the last 10 years.

'Where does Linus live?'

Linus has never worked on software that violates American law. And, perhaps surprisingly, felt completely confident in overruling others in this -

'Torvalds shoots down call to yank 'backdoored' Intel RdRand in Linux crypto' -

Does this tangent disprove the economic power or innovation rate of Silicon Valley?

"‘Where does Linus live?’"

I think the point is that Linus Torvald moved to the US in no small part to make a lot of money off of his brains and expertise. It's a pretty strong point in favor of the argument that the US remains the center of the software development world.

There's no evidence that computer innovation occurred at a faster rate post-1980 than before. In fact the rate of growth and innovation seems pretty constant going back to at least 1970. The only difference is that computer innovation growth appears more visible since IT is a larger part of the broader economy.

I suspect that software innovation has increased, if for no other reason, than the hardware is more prevalent and cheaper now. Thus making if far more accessible and bringing a much larger pool of potential programmers into the mix.

"The only difference is that computer innovation growth appears more visible since IT is a larger part of the broader economy."

That seems reasonable with regards to IT hardware, at least as far as the traditional computing areas. But does it cover the explosion in cameras, phones, tablets, toys, games, etc? I don't really know, I'm really just asking the question.

Software has always had lower startup costs, and with pretty close to ubiquitous computing, nothing can come close. (Zero delivery cost helps too.)

The most serious problems can be solved by properly enforcing the requirement that patents be "non-obvious". For that to happen the USPTO needs to ensure that its examiners are qualified to evaluate the technology that the submitter is attempting to patent. Currently in my experience that examiner is competent about 50% of the time.

If software patents were to cease to exist, the effects would be that: a) many tech innovations would be quickly copied by the Chinese b) companies would reduce cooperation, and c) big companies like Google, Apple etc. would try even harder to force consumers into walled gardens with closed proprietary technologies.

What are some software "tech innovations" of the last decade? Just curious.

While I agree with your other points, software patents really aren't stopping the Chinese from ripping off whatever they can.

Musk, for example, has said that he patents little to nothing at SpaceX, because their biggest competitor is the Chinese and the patents won't stop them. (Their stuff also either explodes or leaves the planet, making it hard to competitors to analyze and reverse-engineer, so patents don't offer much protection even if their competitors would be legally bound by them.)

Lots of companies are also moving to Software-as-a-Service which isn't affected as much by piracy.

I disagree. The biggest problem with patents, by far, is the expansive interpretation of patents that allows filers to claim ownership of things they never invented. This is not a new problem; you can see a prime example of it in the Wright brothers, whose patent on wing warping was interpreted to cover practically any means of controlling an airplane in flight. The resulting litigation stymied aircraft development in the US for more than a decade. These broad interpretations of patent rights pave the way for patents that are more aspirational than inventive. A would-be inventor asks the question, "Wouldn't it be great if we could . . . ?" and patents a half-assed, not-truly-functional solution. Then, if anybody ever manages to devise a solution that actually works, the holder of the patent, typically a non-practicing entity by this point, claims ownership and sues. The "podcast patent" recently discussed on Planet Money and other places is a perfect example of this phenomenon in recent years.

I agree this is the biggest problem, but it is a very thorny and intractable problem.

Given the incredible amount of trivial software patents, I am not sure whether a single competent examiner (as far as software is concerned) actually exists. Things like blinking cursor should never have been patented.

Id say it would be a big win if the SC just nullified the '..on a computer' or '..over the internet' type patents where the idea being patented isnt new at all, just implemented on a computer or over the internet. Im not saying there arent loads of other bothersome patents, but getting rid of that small subset would be a big step in the right direction.

SC has already nullified such patents in Bilski

The Wikipedia entry on that is hard to parse, what all did the SC nullify in that decision?

Well, one of the things it held is that there has to be a physical machine (or transformation, but I won't get into that) for patentability and that machine has to be integral to the patentability of the invention. The machine cannot be mere insignificant post-solution activity. It also reaffirmed that an abstract idea, for example a process that could be carried out mentally, cannot be made patentable simply by implementing it in a machine.

Taking a known method and practicing it over the Internet or on a computer is not going to be patentable, unless there is something nonobvious (patentable) about how it is adapted to the computer/Internet. I.e. if there was a very serious problem implementing a certain known process over the Internet and you came up with a solution for implementing it on the Internet that was nonobvious, you could probably get a patent. But I don't think that is what you were describing.

Now they just need to get the Federal Circuit court to actually apply said rulings in the spirit they were intended. This has been a much more difficult task...

Ho-hum, boring. An op-ed piece, weakly reasoned, that promotes the blogger's own work. Please. The only people agreeing with this are the self-selecting crowd that uses Linux as their OS.

As liberalarts correctly noted, the CAFC (1982 court) has been a boon to technology, as computers have exploded and helped increase productivity (wages are another matter, still stagnate).

Let's analyze the paper by Alex T and "Eli Dourado", which I thought was a pseudonym until now. Hereinafter the paper is by "Alex"

The first 13 pages of Alex[s paper is devoted to a sort of telescoped history that a real historian would probably find laughable. But Alex makes this point: Disney's Mickey Mouse is copyright but the Brothers Grimm fairy tale characters are not. Further, Disney's Snow White is copyright but other versions are not. Now a question for the reader: which version of Snow White do you prefer for your kindergarten kids, the Grimm version of fairy tales or Disney's? Consider the public domain of these tales: Little Mermaid? The mermaid dies. She drowns herself. Hunchback? Esmeralda was stolen from her mother as an infant. She and her mother are reunited just in time for Esmeralda to be torn from her mother's arms and hanged. She's thrown into the charnel pit, a place where bodies not fit for Christian burial are dumped. The hunchback curls up beside her dead body and dies. Good night little one! Sleep well my angel!

Now let's move to patents. Fig. 1 of Alex's paper shows patents issued from 1900 to 1980 to be constant, at 50k issued patents a year. Clearly, as is well known, the Patent Office was biased against patents during the 1960s to 80s, since technology exploded but the patents issued did not. How is this good? But Alex likes this "no patent" stance, presumably Alex is OK with companies, like Musk's SpaceX, keeping things trade secret. Keep in mind the recipe for concrete was lost and reinvented several times in history just because of trade secrets. Other examples exist.

In Fig. 3 Alex draws his Laffer Curve for patents, which is ludicrous since without parameters we don't know which side of the "peak" we are at. Next Alex mentions Wikipedia as a sort of "anti-IP" website. Yet everybody knows that Wikipedia in recent years has been heavily edited and effectively closed to revision; so much for "open source"?! If Wikipedia has closed its website to much further revision, they must understand IP better than Alex. Then Alex still drones on about copyrights and says nothing about patents until page 30 of 44, when he finally mentions the *benefits* of patent pooling. So much for patents being bad? Where's the beef? Moving on, page 33 mentions how government R&D in agriculture is good. Again, if this paper is anti-patent, where's the beef Alex? Alex mentions the Bayh-Dole Act but surprisingly does not criticize it much (I could do a better job) despite giving patents to firms that use public money. Where's the beef Alex? Page 37 suggests abolishing the CAFC, which would encourage what lawyers call "forum shopping" which is filing a case with the judge known to invalidate all patents, if you are looking to invalidate a patent, or vice versa. This was common before the 1982 CAFC. Alex presumably is unaware of this, or maybe thinks this is cool. Alex mentions on page 38 crowdsourcing as an alternative to patents. Yeah right. Crowdsource fusion anybody? I pledge $10 if it can be proven in the next six months. Alex concludes his paper on page 39 without really attacking patents other than using buzzwords like 'patent trolls' (most trolls are minor nuisances to big business btw). Essentially, Alex's paper is summarized as: "it would be cool to have Disney's versions of classic fairy tales in the public domain, along with stuff published post WWII; patent trolls are bad; public choice says various things about IP that could be bad, but I have no real proof". What a joke.

^ IP lawyer?

Ray, TL;DR. With the exception of the first few sentences, the text is chaotic. Especially with the missing structure.

As for your comment re "self-selecting Linux crowd".

First, the condescension isn't OK, because this crowd contains a huge percentage of innovative and original IT minds. If they squawk, they have something to tell and you'd better listen.

Second, I don't believe it. Pretty much every single manager from Big Software corporate world that I met was disgusted by the patent system, and the most frequent comparison was to the bad old arms race of the early 1980s, where both NATO and USSR just had to stockpile huge amounts of very expensive atomic weapons, because the other side did the same.

With a few exceptions, the income from patent royalties does not even come close to covering the huge expense for the legal operations needed to maintain and expand the armory of trivial patents.

The current system is closer to the complex property arrangements seen in Europe during medival period. Whatever the intended purpose, they have become a tool of the well connected to wall off the wealth of society for themselves.

hey Z, reminds me I must visit your blog and 'share my deep thoughts'.

"complex property arrangements seen in Europe during medival [sic] period"- you mean like Fee Simple Absolute, Life Estate, Tenants in Common, Tenants in the Entirety? Those complex property arrangements? Seem to have worked quite well up until now. The 21st century dinged these complex rights a bit tho ("Kelo v. City of New London, 545 U.S. 469 (2005) was a case decided by the Supreme Court of the United States involving the use of eminent domain to transfer {property from Peter the honest taxpayer to Paul the crooked developer}")

I would suppose he means property arrangements like Fee Tail, arrangements which affected a larger percentage of land and had the economic effect he referred to. Sweden only got rid of fee tail (fil a comis) about two decades ago, though it let five such estates remain fee tail to 'preserve' five major estates and their castles. Indeed Swedish law on the division of jointly inherited farm and forest land preserves, de facto, fee tail to the resident (usually eldest) son, hand scraps to the others in place of their supposed 50% inherited interest.

I view most software patents as the patenting of arithmetic problems, and find it scandalous. Copyright would have sufficed to slow down the copying by rivals. As for the idea that patents help the little guy, no. A few lucky little guys get bought by the big guys, rather than dragged into litigation. Most just suffer from the expense of the system.

@ the socialist Marian Kechlibar - "As for your comment re “self-selecting Linux crowd”. - yes, what? Did you want to say something and it got truncated? Happens to me too, since no edits allowed. Or are you really a clueless Linux user and I got your goat? I hit hard! ;-)

As for your Big Software corporate world comment, you are not talking to the right people if you think they are against IP (even and especially the early Bill Gates was for strong copyright), or, your confirmation bias mood affiliation is filtering what you hear so you just think Big Software is anti-IP.

MR: "With a few exceptions, the income from patent royalties does not even come close to covering the huge expense for the legal operations needed to maintain and expand the armory of trivial patents." Not true. IBM gets $2B last I checked from patent royalties, each and every year. You think their in-house lawyers are more expensive than that? Not. Trivial patents? No such thing. Every small advance adds to the armory. Ask Qualcomm about that.

I guess you are a fan of the myth backyard garage inventor who, in a patent-free world, would take on a big engineering shop and beat them at their own game, designing a better widget despite only having "good ideas", obsolete equipment and duct tape is that it? Have I got a movie for you: "Tucker: The Man and His Dream (1988)" I guess you are not a Linux user, where that ad hoc hobbyist approach has been taken, and the hobby coders (like myself BTW) produce mostly buggy junkware that the Linux enthusiasts try an pass off as equal or superior to Microsoft's offerings. That's why you have 1% market share for open-source, IP-free products like LibreOffice (yes I use it, and it's not as good as MS Office) and OpenOffice, and 99% market share for commercial, for-profit, IP-protected software like MS Office. Nuff said.

Flame on dude! That's right, I'm an evil patent attorney arguing for my vested interest, so you can safely ignore me (not). Feel better now? Back to your nursery school worldview then. It's shared by many prominent economists too...

Frankly, this answer is so immature-sounding in general, that I do not feel like responding to the few good points therein. Feel free to consider yourself a hard hitter.

You were trying to have a real conversation? Everyone knows MR is just trolls trolling trolls.

You're living in the past.

Windows owns the desktop. So what. Software is moving to the web and mobile. Do you think those apps are hosted on MS software? Written using MS software?

Let's see...users of 'buggy' Linux....Chicago Merc, NYSE, Amazon, Goog, CERN, NASA,.....

Yeah, this is just wrong and obviously written by someone not in the industry. Software guys are nearly uniformly against software patents.

Software guys are nearly uniformly against software patents

CTOs and product architects are likely to support them.

Not in Germany. But then, software is not patentable here in American terms.

No they aren't.

I mean, I don't know about inside the ranks of Apple or Amazon or whatever. They may have drunk the kool-aid. But every CTO and architect I've ever met in my fifteen year career as a software developer in Silicon Valley has regarded software patents as evil. Some -- but not all -- of them will add that "Okay, they're evil, but that's the world we live in so we're going to patent our stuff too."

>Keep in mind the recipe for concrete was lost and reinvented several times in history just because of trade secrets.

Because of trade secrets? If only the Romans had possessed a central patent office, then the German barbarians would have carefully recreated the use of concrete!

I suspect the Byzantine Empire would have been using those ideas if their had been a central patent office. In almost every case, disbursing knowledge is a good thing. And the Byzantium empire might well have benefited from stronger walls.

"After 2,000 years, a long-lost secret behind the creation of one of the world’s most durable man-made creations ever—Roman concrete—has finally been discovered by an international team of scientists, and it may have a significant impact on how we build cities of the future."

If I recall there was a previous post here on MR about the concrete, and some in the field indicated the older mix was not used because it took a long time to cure - not conducive to constructive schedules of today.

One of the resaons Roman Concrete was "lost" was that it was not really reproducible. It required the use of volcanic ash, something that was unavailable at the surface in much of Europe. Roman cement continued to be used in various locations where the pyroclastic material is available. During the Empire the Romans never used the stuff outsidd of Italy anyway. When the empire fell it stopped being available. The reason the Byzantines did use it was that they weren't in Italy.

As to being a lost secret, Vitruvius described its manufacture in considerable detail by Vitruvius. Though he was lost in the west these books were recovered by way of the Byzantines pretty early in the 1400s.

The rest of the formula continued to be used and improved on throughout the Dark Ages and into recent times, it is just very inferior. Portland cement, what we call concrete, was invented by people before we could chemically explain why Roman, or Portland, cement actually worked.

One of the resaons Roman Concrete was "lost" was that it was not really reproducible. It required the use of volcanic ash, something that was unavailable at the surface in much of Europe. Roman cement continued to be used in various locations where the pyroclastic material is available. During the Empire the Romans apparently shipped the stuff from the Naples area. When the empire fell it stopped being available. As to being a lost secret, Vitruvius described its manufacture in considerable detail by Vitruvius. Though he was lost in the west these books were recovered by way of the Byzantines pretty early in the 1400s.

The rest of the formula continued to be used and improved on throughout the Dark Ages and into recent times, it is just very inferior. Portland cement, what we call concrete, was invented by people before we could chemically explain why Roman, or Portland, cement actually worked.

"As liberalarts correctly noted, the CAFC (1982 court) has been a boon to technology, as computers have exploded and helped increase productivity."

Yes, because literally nothing in the entire space of computer technology has changed since 1982. Everyone knows that iPhones easily could have been build circa 1980, if not for weak patent protection.

And the Newton clearly was so successful because of the stronger patent protection in the 90s

@Doug- Your point (sarcasm with double negatives, too clever by half) is not made or well taken. It seems you don't follow the mobile phone industry: try Googling on Bing: "samsung vs Apple patent war" (7.3m hits) and get back to us when you educate yourself.

@Doug and @Somebody--indeed, if we had stronger patents, perhaps we would have had today's technology sooner, if that's your point. As for the Newton--who owns the patents on Newton? Have you researched that? Did Google spend $12.5 billion on Motorola patents in vain? You are smarter than Google, right? Not.

Ray, Google's purchase of Motorola patents makes no argument. Google, at that moment, was playing within the patent system here called into question.

"Everyone knows that iPhones easily could have been build circa 1980, if not for weak patent protection."

Hopefully you say the above with tongue in cheek (meaning sarcastically).

You forgot however, to mention that recently archeologists "dug" up some old Roman cement roads and found mixed in with the volcanic ash filler, some ancient 8 Terabyte flash memory drives. And on one of those drives there were recorded some ancient Rome patent applications for a new type of cold fusion engine. Others of the drives had music recordings by the Roman music diva, Caveat Buy-ouncie. (/end sarcasm)

I blame Carter completely for this -

'The RSA algorithm was publicly described in 1977 by Ron Rivest, Adi Shamir, and Leonard Adleman at MIT; the letters RSA are the initials of their surnames, listed in the same order as on the paper.[3]

MIT was granted U.S. Patent 4,405,829 for a "Cryptographic communications system and method" that used the algorithm in 1983. The patent would have expired on September 21, 2000 (the term of patent was 17 years at the time), but the algorithm was released to the public domain by RSA Security on September 6, 2000, two weeks earlier.[4] Since a paper describing the algorithm had been published in August 1977,[3] prior to the December 1977 filing date of the patent application, regulations in much of the rest of the world precluded patents elsewhere and only the US patent was granted. Had Cocks' work been publicly known, a patent in the US might not have been possible, either.'

Note the key point - most of the world, even back then, does not recognize the American patent process in the area of software, though for a technical reason in this case, as compared to the actual impossibility of patenting algorithms or business methods in most of the world.

So, anything related to jimmy carter must be bad. Even if all he did was approve the court -- which is and was a huge improvement.

I should point out, a much larger event in terms of software occurred in 1989 - but Stallman is not the sort of person that this blog is likely to celebrate. If only, in part, because he is such a demonstrable refutation of much of what the people behind this blog hold dear.

The apple OS is based on a free unix variant. The Safari browser is based on an engine written by a bunch of german guys for free. The Microsoft monopoly was threatened and broken by the power of free software. Remember the Asus EEEpc an inexpensive small laptop that forced Microsoft to continue producing XP because the new windows was too resource demanding? Asus brought out something that changed the market by not having to kowtow to the Microsoft base platforms that had defined computing up till then. They released it with an open source variant of Linux, soon afterwards Microsoft had a pared down version of XP for them.

I understand the desire for patent protection, but in reality, the way that it works is to protect monopolies, to defend incumbents from new entrants into the market place, and to enrich patent lawyers. It is working as designed, and the utopian images of potential wealth from invention are simply a pretty drawing that maintains the political support for monopolies.

I don't see how the success of GNU would be something inconvenient to Alex's narrative at all. If anything, it shows that with the right alignment of incentives, you don't need money to have markets, and that you don't have to charge for something to create value.

Many companies out there, including many not in software as their core competency, are open sourcing a bunch of their code, because it is more valuable to them as a project in github than as a guarded secret. For instance, if you were working on gene sequencing, you might still want to keep your highest level analysis techniques secret, but you will want to open source whatever you are using as a core technology to do computations along your computing grid.

Markets without money and where many people benefit of the sharing of non scarce resources. It's something that every economist should celebrate.

Given Jimmy Carter bought the "the private sector is better than government at innovating so all the inventions developed with government spending need to be patented and turned over to corporations who will speed them to market driven by the monopoly profit incentive" argument of conservatives, why is this a Jimmy Carter miscalculation?

Is Jimmy Carter's miscalculation the assumptions conservatives are correct on economic growth policies?

The idea that government is better at picking winners is a liberal or progressive point of view. The 80s were driven by the "keep government from picking the winners and let the private sector do it by patent monopoly creating the superior solutions".

As someone in the computer industry, the only progress came outside that model, like inside the government and university labs. RISC was not patented. CISC was not patented. The Internet protocols were not patented. The process of developing industry standard communication protocols required cutting through the patent thicket which was time consuming. The driver of a lot of the efforts was the lure of government contracts under COTS terms which was to level the playing field.

In the end, government picked the winner: the Internet - over the industry standard solution and all the proprietary solutions. Only a few vendors were happy at the choice of Internet. But five years after the Internet was picked, the "world ha changed".

Jimmy Carter was wrong to listen to conservatives. Ronald Reagan could easily have reversed the Carter policy proposal, but instead they pushed it to the extreme in claiming universities would soon no longer need government funding because of all the profits from all the patents they would be claiming from doing research.

@mulp: " RISC was not patented. CISC was not patented. The Internet protocols were not patented" - LOL! Thanks for the humor. The jokes on you however, if you think RISC/CISC are not patented or copyright: they are, in the x86 (Intel uP). Here's a link: Internet protocols are patented all the time, by Cisco and others.

Further, if you design a chip that does what Intel's chip does, you will be sued. If you looked at their chip microinstructions you'll be sued for copyright, or if you did not look but your stuff acts the same as theirs you can be sued for patent infringement. Source: AMD vs Intel patent wars. ARM and Intel should get into a patent war in the next 10 years I would imagine. Background:

Flame on dude! Ignorance is bliss in the blogsphere.

Internet protocols are patented all the time, by Cisco and others.

Indeed. And if they were not patentable, then Cisco etc. would have a strong motivation to make them trade secrets - perhaps by transmitting them only over encrypted links to other Cisco devices - and vendors would not be able to interoperate. It would be like going back to the days of SNA, DECnet, and Appletalk.

"...which Congress created in 1982..."

Doesn't that mean that Saint Ronnie signed the legislation? Or did I miss the outcome of the 1980 elections?

Alex is what you call a "useful idiot" on this topic to all the oligarchic firms (Google, Microsoft, Facebook, etc.) that want to maintain their monopolies in perpetuity. The only thing they fear are strong patents for software inventions that could drive them out of the market.

Isn't the situation exactly the opposite? Didn't IBM sue SUN and try to tie them up to keep them out of their markets? Isn't Apple suing Samsung because they dared to make something that competes with their golden goose?

Those big oligarch firms can easily pay the 'Patent tax', and spend enough of their money on patenting everything that they can. They could use their patent portfolios to crush startups, patent troll style, but it'd be about as smart as starting a nuclear war.

The reason they don't want any patents is that, without the tax, they think they can outhire anyone. Act like patent trolls, and then their advantage disappears.

Its moments like this you wish more people were economists. The moments you start thinking of consumers as the servants of firms and not vice versa is the moment welfare starts suffering. Can someone honestly give an example of a product or service that adds or added real value to people that has disappeared because it was not protected?

The majority of Pharma's costs are not in R&D and they are more profitable than the average industry, is this completely unrelated to the patent premium? I don't think so.

@Diomides: " this completely unrelated to the patent premium? I don’t think so"

Modern pharma is able to cure diseases that once required surgery, and keep people alive that would otherwise die, and do miraculous stuff just by swallowing a handful of pills. " this completely unrelated to the patent premium? I don’t think so"

I would recommend everyone Paul Graham's essay on the subject: .

TL;DR: Software patents are not that different from any other patents. Start-up companies don't really benefit from the patents that much except the protection they get against patent litigation from large companies and patent trolls.

As a user of many national & international standards, I am annoyed by "submarine" patents that are not declared during the standardization process (today that generally means the owner of the patent did not participate in the standardization process) suddenly show up years later as a threatened lawsuit. For example, for video codecs such as MPEG-2. This was a serious drag on the adoption of H.264 and remains a threat to HEVC video coding.

I would like to see a situation where if ANSI decided a standard should be elevated to a national standard, there should be a "put up or shut up" phase (perhaps lasting one year) during which all IP holders must declare that they own IP essential to a standard and will license it RAND (reasonable & non-discriminatory), or they lose that IP. That way, users of the standard can feel safe that all IP in the standard has been declared and licensed RAND.

"As a user of many national & international standards, I am annoyed by “submarine” patents ..."

As a user of much technology, I am annoyed by the fact that somebody invented each little piece of it and that some of those inventors demand to be "compensated" (ha, everyone knows technology wants to be "free) for the work they did (/end sarcasm)

It's encouraging to see that people are focusing better on the core issue of scope. The system itself is not as broken as it might seem from some of the abuses in the software area made possible by the lack of granularity or resolution in defining the scope of a patent claim.

But I still don't see enough discussion of an even more fundamental problem of private information. Patent licenses and sales aren't public record like real estate. Real estate is unique with idiosyncratic valuations too yet there's a market there. Comaprables give people comfort that they're not being cheated.

At the end of the day, there is value in the work of coming up with the right idea independent of the value in the work of building it. Abolishing patents simply puts the inventors even more at the mercy of the builders. They are already. But even more would they be were patents abolished. How would getting rid of the abuses compensate for that?

The arguments against the current patent system are clothed more in emotion and outrage over some fringe cases of abuse than any sound logic. First, getting a patent issued by the USPTO is a god awful pain in the * process. By default PTO rejects every damn patent applied for. 99.99% are rejected as part of due process. You need to go through repeated appeals and long drawn frustrating and very expensive refilings, amendments, continuations etc. to get one issued. Which of course favors the big firms over the small guy. Which by itself, would be cause to make it easier to get a patent, not harder.

Second, having a patent issued means nothing by itself. For it to be prosecuted, you will need to get over discovery and fight off many challenges based on prior art, which in vast majority of cases makes it almost impossible to prosecute the patent. So what we are left with is the small few cases that go to prosecution. Most of these are genuine infringement of truly innovative art, which is rightfully protected -- not just as an economic incentive but one can argue, also as a matter of justice. But some, very few are truly ridiculous cases, that get massive media coverage. These cases, are rare and a product of less than a fraction of a percent of the patents actually issued. And solving the problems is extremely simple. Grant protection to companies below a certain size (revenue, employee, age) threshold from prosecution for x number of years, so they can take advantage of any IP until they have enough revenue to pay royalties, or find a way to get around the infringements. As for the big guys, let them sue each other to death -- its a free market, there's no reason we should be worried about what competitive tactics companies use to compete. If anything, a strong second market in patents, which has developed recently as a result, only encourages more out of the box thinking purely for filing patents, before the technology matures to make them possible. Which in turn provides incentives and strong demand indicators to the research people. Just look at the explosion in graphene patents of late.

Comments for this post are closed