Patent examiners misrepresenting their work hours

In my last post, I discussed a recent study that suggested that time constraints cause patent examiners to grant more patents, particularly by decreasing the rate for time-intensive tests like obviousness. On a related note, the patent office recently released an investigative report on patent examiners which alleges systemic abuse of work hours.

The report names four abuses. Summarized, they are:

1. Time fraud: misrepresentations of time and attendance
2. End-loading: sudden increase in productivity/work at end of the quarter
3. Mortgaging of work: claiming work before actually done
4. Performance plan issues

A Washington Post article reports:

Some of the 8,300 patent examiners, about half of whom work from home full time, repeatedly lied about the hours they were putting in, and many were receiving bonuses for work they didn’t do...More than 70 percent of the 80 managers interviewed also told investigators that a "significant" number of examiners did not work for long periods, then rushed to get their reviews done at the end of each quarter...[this] practice "negatively affects" the quality of the work.

Patent examiners, unlike most federal employees, are allowed to work from home, even full-time. I think work flexibility is good, but not at the cost of abuse of course.

In my last post the study suggested that time-constrained higher-ranking examiners performed worse at obviousness rejections. This report of abuse could demonstrate another problem. With end-loading abuse, is there a sudden rush to grant patents at the end of the quarter? If the application backlog was shorter, would it be beneficial to time my application for less scrutiny?

Could this abuse lead to "stupid patents" being granted or inconsistent examinations that are unfair to patent applicants? With the high cost of patent litigation and risk, and concerns over the quality of granted patents, abuse should be taken seriously.

From the report:

...approximately 20 percent of examiners (roughly 1,600 examiners) completed 50 percent or more of their total annual balanced disposals in the last two pay periods (or four weeks) of the quarter (typically six or seven bi-weeks), and approximately five percent of examiners (roughly 450 examiners) completed 70 percent or more of their total balanced disposals in the last two pay periods of the quarter.

End-loading work will create a quality control problem. It only takes one granted "stupid patent" to fall into the hands of a troll to impose large economic and legal costs. According to a Boston University study, the lost U.S. capital attributable to patent trolling was $29 billion in 2011.

I have sympathy for patent examiners, however. They have a difficult job that requires both tedious work and intellectual judgment. It's not their fault that courts change interpretations of patent-eligible subject matter, for example with software patents. It's unfair to use them as a scapegoat for patent abuses that reform (through laws or court precedent) could solve.

Pressed for time, examiners grant more patents

A recent study from the University of Illinois (UI) suggests that time constraints imposed on patent examiners is allowing more "on the margin" patents to be be granted, likely because they receive less scrutiny. This makes sense. If patent examiners are alloted less time, they are unable to reject patents under Section 102 (novel) and especially Section 103 (non-obviousness), which require tedious search for prior art and judgment, as I discussed in another blog post (and in class).

It's no surprise that faster work means poorer results, especially for patents. I mentioned in a previous blog post that during David Kappos's tenure as director of the patent office, the backlog of patent applications decreased but the approval rate for patent applications increased by 20 percent, suggesting that the patent office may have lowered its standards to process the backlog.

Moreover, the study claims that this trend impedes the performance of promoted examiners, who are expected to spend less time reviewing and work faster. Thus, higher-ranking examiners are more lenient at granting patents than are lower-ranking examiners.

From the UI press release:

The researchers found that an examiner’s promotion to each subsequent step-up in pay equated to a 10- to 15-percent decrease in the number of hours allocated to review a patent application...as an examiner is given less time to review an application, they become less inclined to search for prior art, which, in turn, makes it less likely that the examiner makes a prior art-based rejection.

On the plots of examiner rank level and grants, the grant rate consistently increases with increasing rank, from GS-7 (mid-level) to GS-14 (top-level). Likewise, the share of obviousness rejections consistently decrease with increasing rank.

While I suspect that the grant rate increases could be due to another factor, like greater confidence, the consistency with the obviousness rejections is particularly damning. All other things equal, I would assume that an experienced examiner would be more adept at creating a claim to obviousness from prior art, and in fact make more obviousness rejections. Thus, I think that the decrease in obvious rejections is more likely caused by time constraints.

I assume the examiner could change during the application process. Does the study account for patent applications that were originally objected to by one examiner and then accepted after changes by another examiner? Is it possible that the patent office would assign these responses to higher-ranking examiners? If so, the higher-ranking examiner might appear to be more lenient while the persistence of the applicant is the underlying cause. However, I think the patent examiner doesn't change frequently, and assume that the study would have otherwise controlled for it (hopefully).

Regardless, the inconsistency alone between grant rates and obviousness rejections among different examiner levels suggests that patent examination is unfair. If my patent application is assigned to a lower-ranking examiner, according to the study, I will be less likely to be granted a patent than a hypothetical competitor whose patent application is assigned to a high-ranking examiner.

Scanner troll sued by states and FTC

In a previous blog post, I mentioned an (in)famous "long tail" patent assertion entity called Project Paperless. Project Paperless, along with its shell companies, targeted business offices with scanners on their network, claiming what appears to be patents on scanning paper documents and viewing images.

Project Paperless later sold its patents to another troll, MPHJ Technology, owned by a Texas attorney Mac Rust (no surprise there). Apparently, Rust's job is to calm down "really irate" letter recipients to get them to pay from $900 to $1,200 per worker. MPHJ typically sends notices to small businesses demanding payment for each employee or risk an impending lawsuit. As I understand, the notices are harsh but indiscriminate. MPHJ does not investigate infringement. It probably assumes that every business office owns a scanner with an infringing scan-to-email feature.

With such tactics, it's no surprise that politicians are reacting. Congressman Peter DeFazio (from Oregon) mentioned MPHJ when advocating for his SHIELD bill, which would require some losers of patent litigation to cover the other side's costs. (The SHIELD act might create problems of its own.) Legal action was also taken against the troll in several states, including Vermont, Minnesota, New York, and Nebraska. Minnesota's attorney general was able to convince a judge to force MPHJ out of the state entirely. New York's attorney general reached a settlement, requiring MPHJ to return money it had received from businesses in the state, but allowing it to send out "toned down" demand letters in the future.

MPHJ is trying to strike back. During an investigation by the Federal Trade Commission (FTC), the U.S. regulator for consumer protection and business practices, MPHJ unexpectedly sued the FTC. In turn, the FTC filed a motion to dismiss the lawsuit, claiming:

[MPHJ] improperly seeks to impede an ongoing investigation by the FTC...The FTC's investigation centers around whether Plaintiff [MPHJ] violated, or is violating [the FTC Act]...by falsely threatening thousands of small businesses with imminent patent infringement litigation when, in truth, Plaintiff did not intend to take and did not take such action.

In Nebraska, MPHJ claimed that its demand letters were constitutionally protected free speech. The judge agreed, but the state attorney general has appealed.

Like with other states, Vermont's attorney general sued claiming that MPHJ violated consumer protection laws. However, MPHJ called the case "sham litigation" that sought to illegally restrict free speech and "usurp federal patent law". It requested that the state attorney general be sanctioned and that state pay its legal fees.

Thankfully, it appears that saner heads will prevail. MPHJ attempted to get the case heard in a federal district court and then appealed to the Federal Circuit. Both times the judges pushed the case back to state court. A Vermont venue makes a victory for MPHJ unlikely.

I'm glad to see that startups and small businesses are not alone in the battle against trolls. But it's unfortunate that an abusive business with a patent (that I assume/hope has prior art) is able to run a legal circus. One wonders how much lost time/money is imposed on small businesses, bureaucrats, and the judiciary alike.

Tesla's true intentions?

In my last blog post, I discussed Tesla Motors opening up its patent portfolio. I mentioned that this generosity is not completely unusual in the tech industry. However, it's still surprising for a public-traded company (i.e., answerable to shareholders) with research in not just software, but also more tangible items like vehicles and their batteries.

Akshay wrote a blog post asking whether Tesla had other motives? He questioned the meaning of CEO Elon Musk's statement that Tesla would "not initiate lawsuits against anyone who, in good faith, wants to use our technology". In his words:

With other, larger companies getting in on the electric car market, Elon Musk hopes they will help him in the building of these super-chargers...Tesla is currently by far the leader...they don't see a significant threat to sharing their technology with other companies because they already have a huge head start, both in brand-recognition in the EV market as well as in quality of product.

I agree. I also agree that Tesla would benefit from competitors using compatible "supercharger" charging station technology, which is currently restricted by the most important Tesla patents.

Later on, Musk himself writes:

We believe that Tesla, other companies making electric cars, and the world would all benefit from a common, rapidly-evolving technology platform

I'd also, however, like to also point out that Tesla probably views their patents defensively (deterrent to opposing patent lawsuits from potential competitors), not commercially. They don't see patent licensing as an important source of revenue, as Nokia does, or a competitive advantage, as Apple does. So, the disadvantages to opening up their patents are minimal. As it stands, they don't really have competitors.

Tesla actually tried and failed to monetize their patents. In 2011, Tesla partnered with Toyota to supply its battery, gearbox, power train, and battery-management software for Toyota's RAV4 EV. At the time, the deal was projected to be worth $100 million. But the deal ended sourly in May 2014. (In fairness, Tesla was more successful in a similar contracts with Daimler's Mercedes-Benz.)

Furthermore, according to Blattberg of VentureBeat, Musk's "good faith" qualifier and lack of legally-enforceable patent license means that nothing legally has changed:

Tesla's patents are not available under some automative [sic] equivalent of an Apache license...It doesn't matter how nonchalant Musk acts -- Toyota, Ford, Honda, and the rest would never knowingly infringe on Tesla's patents without legal documentation that everything is cool.

I suspect this leaves Tesla with flexibility to change its mind or deny the "license" to a particular competitor. In short, Tesla would benefit from encouraging competition that expands the market and charging infrastructure, but isn't completely committed.

Tesla is making a positive step for electric vehicles, though not solely out of goodwill.

Tesla opens patent portfolio

Tesla Motors, the electric vehicle manufacturer based here in the Bay Area, known for its Roadster and Model S cars, announced that it "will not initiate patent lawsuits against anyone who, in good faith, wants to use our technology." In the announcement, CEO Elon Musk writes:

Technology leadership is not defined by patents, which history has repeatedly shown to be small protection indeed against a determined competitor, but rather by the ability of a company to attract and motivate the world’s most talented engineers. We believe that applying the open source philosophy to our patents will strengthen rather than diminish Tesla’s position in this regard.

The exact details haven't yet been announced, but it certainly looks generous. Tesla has 203 patents on its batteries and electric vehicles. Another 280 patent applications have been filed. The earliest patent expires in 2026.

I think this is a surprise to a tech world with patent trolls and smartphone patent wars. However, it's not completely unusual.

I suspect that by "good faith", Musk means that the patents will be used for defensive purposes. In other words, the patents would be a deterrent. If an entity sues Tesla, they open themselves to a counter-lawsuit by Tesla. Tesla, like many other tech companies, might not see significant value in their patents "offensively", but rather as protection against lawsuits from competitors or trolls.

Along similar lines, Twitter, through its Innovator's Patent Agreement, promises its employees not to use patents they develop for the company offensively without their permission.

Or perhaps Tesla is trying a strategy similar to Google with Android. By giving away Android, Android gained market share faster, and Google expanded the market for its core product, advertising. Tesla would benefit from more charging stations, for example. Aaron Levie, CEO of cloud storage company Box, observed:

By opening its patents, Tesla rightly realizes it’s better to be the best product in a large industry than the only product in a niche one.

Section 101 vs. Section 103

Norman wrote a blog post that discussed patent validity tests using Section 101 or Section 103 (from Title 35 of the U.S. Code, which is about patents). He mentioned that groups Google and Netflix prefer a Section 101 test, while groups like IBM prefer a Section 103 test.

These different sections correspond to the requirements for patent-eligible subject matter:

• Section 101: not abstract or law of nature
• Section 102: novel
• Section 103: non-obvious

So what are the differences between a test that emphasizes section 101 as compared to one that emphasizes 102 or 103? Section 101 is the easy way to kill a patent. To invalidate a patent, Section 102 requires an expensive search for prior art. Similarly, Section 103 requires expensive expert opinions, which can be contested by the patent applicant.

This case at hand, Alice v. CLS Bank, concerns the patentabilty of software patents. I discussed the case and the larger debate about software patents in an earlier blog post.

I also believe that a stronger Section 101 test would specifically target many software patents, which can be seen as abstract. As I discussed in a blog post, in a number of Supreme Court cases, from Gottschalk v. Benson (1972) to Bilski v. Kappos (2010), the Court has consistently held that an abstract process or transformation cannot become patent-eligible subject matter simply by including a computer.

I think your preferred test reflects your opinion on software patents. If you oppose software patents, you will want a strong Section 101 test that makes it easy to invalidate patents. If you are a proponent of software patents, you will want more flexibility with a Section 102 or Section 103 test that makes it harder to invalidate patents. It's hardly surprising that the briefs filed by software companies, which usually oppose software patents (or seek to restrict them), advocate for a stronger Section 101 test. Conversely, IBM, the largest patent holder, and AIPLA, an IP lawyers' group, advocate for Section 103 and against Section 101, respectively.

The Supreme Court ultimately invalidated the patents under Section 101.

EFF stupid patent of the month

The Electronic Frontier Foundation (EFF), a technology-related civil liberties group that also advocates for patent reform and "eliminating stupid patents" is running a new monthly blog series:

[W]e're introducing a new blog series, Stupid Patent of the Month, featuring spectacularly dumb patents that have been recently issued or asserted. With this series, we hope to illustrate by example just how badly reform is needed -- at the Patent Office, in court, and in Congress.

Stupid patents are nothing new, of course. I previously discussed a bizzare Amazon patent on taking pictures with particular photo settings. EFF is known for its Patent Busting Project, which seeks to invalidate or narrow high-priority "bogus software patents". Still, I like their method of advocating for (stalling) patent reform by example.

I like how they address their concerns in the blog post by specifically discussing the first claim of the patent. The blog post includes the first claim in verbatim and in a more human-readable form (by claim element), to support the argument that this is simply a patent on a "doctor's computer-secretary".

I am also intrigued that, according to the file history, the claim was originally rejected as abstract. The claim was, however, accepted after an element "g" was appended to elements "a"-"f":

(g) providing a computer, the computer performing steps “a” through “f”. 

In other words, by simply adding a computer to the claim, it became eligible for a patent!

As I understand it, in a number of Supreme Court cases, from Gottschalk v. Benson (1972) to the recent Alice v. CLS (2014), the Court has consistently held that an abstract process or transformation cannot become patent-eligible subject matter simply by including a computer.

If so, this patent appears to blatantly violate that ruling. It looks too obvious to be overlooked.