Some Good News on Patents

Further evidence of a dramatic slowdown in patent litigation activity in the United States is provided today in data published by Unified Patents, the entity whose business is based on helping SMEs fight frivolous patent suits. According to the research, which covers the third quarter of this year (June to September),  there was a 23% drop in the number of suits filed compared to the second quarter, and a 27% year-on-year reduction.

The findings come just weeks after data released by Lex Machina showed that there had been a 40% fall in patent suits in September 2014 as compared to the same month in the previous year. Commenting to IAM on the reasons for the decline, Lex Machina founder Professor Mark Lemley claimed that much of it could be attributed to the Supreme Court’s Alice v CLS decision. The Alice judgment was handed down towards the end of June.

More here.

Alice has made a difference as well as the cheaper post-grant challenge procedure in the AIA. Once the first software patents make it through, however, there will be an uptick as people learn the new system. Still this is good news especially for software patents but we have a way to go on the larger issue of IP, including copyright.

Here is the updated “Tabarrok curve“.

Tabarrok Curve


I thought your stance was that all patents were bad. But his graph shows increasing returns to innovation up to a certain point. How do we identify the inflection point?

Do you mean the turning point? What use is the inflection point to anyone?

I feel like we'd want to identify the inflection point so that we can slow down our reforms so that we don't overshoot that local optima but I'm also not sure that's what Sam meant.

Economist graphs don't have scales. They don't even have units. What units would you apply to either of these axes? They defy quantification. For economists, that's a feature. Nobody can say you're wrong if you don't say anything testable. Well, they might say you're wrong, but they can't back it up with anything hard because you haven't said anything hard. These people are not physicists.

Astute comment.

"These people are not physicists." Physicists are people who, when a sum comes out wrong, attribute the error to "dark matter".

That's why they call economics: "political economy". Economists are advocates of certain political world views. What AlexT is pushing is a form of IP communism, weakening IP rights since they seem excessive to him. But he's not an IP abolitionist, which is good, since he does feel there's a need for IP. As such, AlexT is like the well known IP gadfly Jeremy Rifkin.

The trouble with these folk is that they are wedded to the past, the status quo. Their argument is simple: civilization got to where it is now with weak or non-existent IP protection. Why rock the boat? If it ain't broke, don't fix it? Nobel prize winners invent and often have no patents nor do they get much money, pre-Nobel prize. It's that simple. However, what AlexT et al. forget is "people respond to incentives". There's no money, relatively speaking, in science today. Who does science these days? Immigrants, poor people, nerds for the fun of it. Anybody else, with a modicum of ambition, becomes a doctor, lawyer, middleman or manager. And for good reason: you don't need that much talent, and you can rest on the coattails of the toilers, and the law protects the jobs and livelihood of such classes (trespass on property, get arrested; trespass on IP, and get a good laugh or slap on the wrist). What's the downside of this strategy for civilization as a whole? It delays technology. If only 1% of your population is innovating, as opposed to say 10%, the rate of growth for the economy will be much less. Consequently, the future is retarded. Today, IMO, had we avoided world wars and had a strong IP system in place (not the present system, which is sloppy, as typically US patent examiners only have a few hours to decide patentability, with no input from more knowledgeable third parties; Europe with their two-team system is only a bit better), we'd have invented flying, antibiotics, nuclear physics and electromagnetic inventions (radio, TV, laser), as well as the PC, back in the days of George Washington; the early 19th century would be the 20th; and the last century and now we'd have stuff like faster than light travel, a cure for cancer, and perpetual youth (related to the cure for cancer). But not only is that difficult to prove, most people don't care since they are content with their lot. It's hard to appreciate something that you can only dream about, if that (many people reading this probably think I'm a crackpot, showing lack of imagination).

Off soapbox now. In the unlikely event this message is preserved 100 years from now, readers will marvel at my insight.

Of course, Alice only applies to software and business method patents, and software, at least, is also covered by copyright. So since you're insistent on lumping all IP together, and Alex isn't calling for getting rid of copyright, I'm pretty sure your rant is misguided at best.

"Of course, Alice only applies to software and business method patents..."

Of course, that's not actually true. The Alice decision is so poorly reasoned that it can be applied to any patent at all. All you need to do is distill an invention down to a "gist" and call that "gist" an abstract idea (once you've distilled the claims down to about 5-7 words, you can likely say that your "gist" should be termed abstract with a straight face. Consider, when you doubt what I say, that the Alice decision itself has the gall to state "In any event, we need not labor to delimit the precise contours of the "abstract ideas" category in this case.") Then, you simply say that the rest of the claim elements do not amount to "substantially more" than the abstract "gist" itself. Presto! Claims invalid.

Of course, the argument will not be taken to this logical extreme, but the point that it can be taken there without even changing the mode of reasoning used by the Supreme Court shows the intellectual bankruptcy of the decision.

I am all for bashing economists, but in a lot of fields when quantitative measures first come into being the theoretical concept is often figured out before the data. This is especially common where quantitative measures are hard. Cyclothem data in stratigraphy for example.

This has theoretical interest, and makes sense to me. But what we need to figure out is what should be on the x axis? Is it length of patent, or what I really expect it to be, some as yet unknown metric of patent enforcement.

The problem with economists is that the data is terrible and much vaguer than economists like to think it is. The disaster of the other social sciences is that the data they use isn't actually data at all.

We might all be better off if they just got rid of all business method & software patents.

Probably in the long run you are correct, but in the short run it would probably cause a lot of chaos. Maybe a phased approach where we stop issuing any new patents of that type and let the existing ones decay over time would be a more stable option.

JWatts, in general I agree that getting rid of bad laws should be done slowly. But remember business model and software patents are rather new (in practice, if not in principle). In fact you could say the recent "revolution" was that a a decade or two ago, courts opened floodgates for these patents and that the status quo is the actual post-revolutionary chaos.

What sort of chaos would it cause to get rid of software and business method patents entirely and immediately?

They're of no use to operating companies and innovators; the thicket of patents the PTO has issued in leiu of being able to evaluate such abstract ideas means everyone is infringing thousands every day. Even patent holders -- aside from trolls -- would be better off with immediate abolition.


For vast majority of IT corporations, their collections of patents on everything are a net loss. They need to collect vast sets of trivial patents in order to deter their competition from launching a patent attack. I would guess that 95 per cent of all software patents do not create enough royalties to pay for themselves.

The situation is not that different from the Cold War, when the potential opponents stockpiled nuclear weapons in ridiculous number (enough to kill the entire world ten times over), just to keep the power balance in check. A very costly method of maintaining status quo.

Year-over-year is too short a period to tell anything.

All of these findings could potentially be explained by the cellphone patent war.

+1, though it is worrying that as Lemley says: "We’ve seen in the wake of Alice 15 decisions in the federal courts involving software or business method patents – 13 of those have struck the patents down"

As for AlexT, he is trying to look like an inventor with his "back of the napkin" trope but as we all know nobody from the Dismal Science has ever invented anything worthy of a Nobel Prize...except a Nobel Prize in Economics. Note also the sleight of hand in linking patents (which need to be strengthened) to copyrights (which already are strong enough). Another attempt by a small-time economics professor to make us all into zero profit commodity wheat farmers, like in those Econ 101 textbooks.

"though it is worrying that as Lemley says: “We’ve seen in the wake of Alice 15 decisions in the federal courts involving software or business method patents – 13 of those have struck the patents down”"

If you look at the specific patents involved, it's embarrassing that the PTO even issued them. The most numerous kind of patents, especially software patents -- that get litigated in the USA is the really abusive kind that should never have issued or even been considered eligible. One is for keeping calorie information for food in a database and providing a search operation to select and assemble meals. That's the entirety of the claim; there's no innovation, no specifics, no new technology involved, just a monopoly on food databases. One is for marking timecodes for animated characters' lips to move when speaking; there are no algorithms or rubrics for how to do it, just a pure monopoly on what Disney was doing in the 1920s but with computerized tracking of time codes and no technological guidance on that. One is for trading frequent flyer miles again with no new technology, just a land grab for the miles market.

They're all like that with the possible exception of one really bad drug patent that also deserves to be found invalid.

As Ray Lopez likes to say: "People respond to incentives."

Unfortunately for him, this also includes the incentives for patent lawyers to become parasites on a living and healthy industry, the incentives of trolls to do some good trolling, the incentives of unscrupulous CEOs to launch dirty litigations over massive innovations such as "blinking cursor" and the incentives of the employees of the PTO to keep their flock intact (if software patents were abolished, some PTO employees would have to be laid off).

Once Ray Lopez is done with his hooker he's gonna post a screed here...

@ bozo - what is your value add here? Except complaining about the stomach ache you got in the Staples Center Chinese fast food diner?

Sorry, I meant Bozo (TM), since I believe it's technically trademarked. Yes it is.

I didn't see a copyright mark on the picture, nor a trademark on term "Taborrok Curve", so I have filed for both,


You owe me.

No, he doesn't owe you anything.

But I've copyrighted and trademarked the term "Tabarrok Curve". As well as the "napkin assisted design" methodology. Let the undeserved royalties flow. ;)

But, I patented the ink to napkin application process, and therefore my patent blocks the use of yours.

Rock, Scissors, Paper.

I'm afraid the user agreement for the napkin clearly states that it cannot be used for anything other than the use the manufacturer intended.

The user was not in privity with the manufacturer, the user was a bona fide purchaser without notice, and furthermore the restraint was one on alienation, which, because of the 18th century views on this website, is therefore unenforceable as a matter of law.

Ipsi dixit.

Why is less litigation necessarily a good thing? Sometimes you need to protect your intellectual property from misappropriation by way of a lawsuit or the theft will continue.

And why is patent litigation before the Patent Office judges better than litigation in federal court before an Article 3 judge? The former can be influenced by political pressure and threat of demotion or termination, the latter cannot.

How about some other possible explanations:

1. Patents have been devalued by all of the anti-patent Supreme Court (and Fed Circuit) decisions and by the patent reform law, that the deck is stacked against patent owners.

2. Maybe the economy is actually improving and there is more innovation going on - companies focused more on product/service development than litigation.

3. I don't know this, but possibly fewer patent application filings in the U.S. in the past 5 years. More patents, more lawsuits. Fewer patents, fewer lawsuits.

One (in my view more plausible) explanation that gets little press outside the patent-law community are the new, more challenger-friendly patent review proceedings before the US Patent Office: inter partes review and covered business method patent review.

For about 1/5 of the cost of killing a patent in court, an accused infringer can now challenge the patent before a USPTO review panel that, so far, has about a 95% kill rate (by patent claim, as of last March). Notably, the decision to cancel claims/patents is made by technically trained USPTO administrative patent judges, rather than lay people in a jury, and a decision must be made within 18 months of requesting review, faster than most district court cases reach trial.

The results are not patent-friendly. Enough claims have been canceled that patentees are likely now thinking twice before suing, as the economics of a case change dramatically when you are forced to fight validity early in the case. Indeed, some, including the former Chief Judge of the Federal Circuit, have taken to calling these review panels a patent death squad:

Re. explanation number 3, we're at an all time high of patents in force:

Wow, that's good news. The inability to get to the validity issue right away was always the most maddening aspect of the system to a non-lawyer. It's like being accused of trespassing on what is actually a public sidewalk and not being allowed to bring that fact up until after conviction.

Unfortunately Alice did its work in a particularly pernicious way. It may have invalidated a lot of patents, but it did so in an extremely unclear and ambiguous way, so that we have right now very little idea of what the law actually is. This uncertainty acts in the opposite direction, suppressing innovation.

Agreed, Cliff. The court would have been much better off taking the advice of recently former justice Stevens who wanted a simple rule that finance, marketing, and other methods of doing business are not eligible and neither is computer software.

A bright line rule excluding the primary cases of abuse (software, finance, marketing and sales) would be much more effective that the complicated rule in Alice that district courts and even differently composed CAFC and PTAB panels will be interpreting inconsistently for decades.

Should the Bayh-Dole Act (1980) and Federal Technology Transfer Act of 1986 simply be repealed?

Seems to me that after they were passed, innovation began declining.

They certainly failed to deliver the surge in innovation promised by President Carter in 1980:

I have today signed H.R. 6933, a bill which makes several major reforms in the patent and trademark system.

One of my administration's major concerns has been the role of industrial innovation in promoting this Nation's economic health. More than 2 1/2 years ago, I initiated a review of industrial innovation to identify ways in which the Federal Government could improve the innovation process. We were assisted by hundreds of individuals from private industry, organized labor, the universities, and public interest groups.

As a result of these efforts, in October of 1979 I sent to the Congress an industrial innovation message proposing initiatives in nine critical areas relevant to innovation, including legislation to strengthen our patent system. I am pleased to sign this bill, which embodies many of my proposals.

We have already taken other steps to revitalize our patent system, including the application of modern methods of management and computer technology in the Patent and Trademark Office.

The legislation I am signing today does not eliminate all the problems we identified. It does not establish a comprehensive governmentwide policy for the allocation of rights in inventions made with Federal support. This legislation leaves in place the existing, often inconsistent array of statutory and nonstatutory patent policies governing individual agencies. Also left uncovered are large business contractors who perform more than 90 percent of the government's research and development work.

I am persuaded, however, that the present package of reforms goes far toward strengthening the effectiveness of the patent incentive in stimulating innovation in the United States.

The patent reexamination procedures established by this legislation constitute the most significant improvement in our patent laws in more than a century. Under these procedures, during the life of an issued patent any interested person—for example, a patent owner, a potential licensee, or a competitor—may obtain a prompt and relatively inexpensive reevaluation of its validity by the Patent and Trademark Office. Patent reexamination will make it possible to focus extra attention on the most commercially significant patents. This legislation will improve the reliability of reexamined patents, thereby reducing the costs and uncertainties of testing patent validity in the courts. The provisions of this legislation will result in less cost to the public for patent reexamination.

This legislation also authorizes the Commissioner of Patents and Trademarks to set fees at levels that recover a greater portion of the Office's operating costs. A new system of fees for maintaining patents in force will further shift some of the burden of supporting the patent system away from the public to the commercial users of inventions. This feature will be helpful to small businesses and individual inventors by keeping initial patent fees low.

In the area of Government patent policy, this legislation enables small businesses and nonprofit organizations to obtain title to inventions made with Federal support. I hope that this measure will benefit the public in some measure by making the fruits of federally supported research and development more widely available. While I regret that it was not possible to enact comprehensive legislation, I am pleased with the progress we have made.

The enactment of this legislation promises some real benefits to the Nation's economic health by stimulating our people's innovative activity. It now becomes the responsibility of the Government and industry to work together to realize that promise.

The best 1980s legislation to repeal would be the 1982 creation of the Court Of Appeals for the Federal Circuit with appellate jurisdiction over all patent cases. The patent bar immediately captured the CAFC and the patent system has been getting more corrupt and abusive ever since. Single subject courts are a danger to justice and what we see at the CAFC has been a race to empower clever attorneys to blackmail engineers with vague, broad, and obvious patents that the CAFC will not invalidate regardless of the letter of the law. The Supreme Court has had to overturn CAFC decisions 9-0 again and again in recent years just to move the needle partway back toward sanity.

So we should abolish the CAFC and return jurisdiction to the district courts of appeal where it had always been. But Bayh-Dole could stand to be mostly reversed, too.

Weaken software patents and you get more SaaS business models.

Weaken business method "patents" and best practices for one firm become best practices across industries.

I fail to see a problem with either.

Well here we now have Alex Tabarrok coming out against Private Property rights

My economic policy test for when IP has gone too far: If inventors are accidentally reinventing patented stuff 99% of the time or even going out of their way not to look at the patent literature to preserve their legal position, then too much obvious stuff is getting patented. It's supposed to be a trade of disclosure in return for temporary monopoly, but if nothing valuable is disclosed (as evinced by other inventors not even trying to look at the patents to see how to proceed) then the patents in question ought not to have been issued.

Comments for this post are closed