Monday June 28 was the day the U.S. Supreme Court was to decide the patent case of the century, Bilski v. Kappos, and bring clarity to the debacle of the 1998 State Street Bank decision. In State Street, the Court of Appeals for the Federal Circuit (which hears all patent appeals) had upended centuries of tradition that assumed that patents were for technology and a hundred years of judge-made law that explicitly excluded "methods of doing business." That decision also appeared to abolish all limits on software patents, fueling a land rush in patenting that helped create the backlog of 1,200,000 applications the Patent and Trademark Office faces today.
State Street created an instant constituency for business method patents that wasn't there before. Before State Street, everybody knew that business methods were not patentable. It was understood and accepted. It was rarely litigated.
Under State Street, Bilski would have gotten his patent for a risk-hedging scheme for energy costs -- no questions asked. But it has become clear that business method patents -- which might have seemed like a great idea in those go-go years -- are deeply problematic. They are hard to evaluate for novelty and inventiveness, often sweepingly broad in scope, difficult to interpret, and very controversial. (Would we really want just one airline offering frequent flyer miles?) By abolishing the well-established and uncontroversial business method exclusion, State Street radically extended jurisdiction of the patent system to cover not only business practices such one-click ordering and tax avoidance strategies, but an apparently limitless range of human activities, such as athletic moves and playing with cats. Perhaps the largest professional land grab in modern history.
Ten years after State Street, the Patent and Trademark Office wisely denied Bilski's application -- not by attacking State Street head-on but by pointing to another test that the Supreme Court had used in earlier cases: The principle that a patent for a process must be tied to a particular machine or involve a transformation of matter. Without directly touching State Street, the Federal Circuit agreed, atoning for its reputation as an inveterate booster of patents. Then, to the surprise of many, the Supreme Court took the case -- and now has muddied the waters further.
Justice Kennedy's decision rejected "the machine or transformation" test as determinative while nonetheless praising its probative value, but declined to reinstate the business method exclusion because some processes that could be described as "business methods" may be patentable. Joined by four justices, Kennedy embraced a more amorphous test, the exclusion of "abstract ideas," citing language on algorithms in Benson, Flook, and Diehr, the Supreme Court's last and only word on computer programs, dating back to the 1970s. Kennedy's affirmation of Benson and Flook suggests that software patents remain problematic, but he adds nothing to the old language that he cites, leaving it to further litigation to determine what an "abstract idea" is in different contexts. Maybe it's like obscenity: you know it when you see it.
Only worse: How do you make concrete something characterized by its lack of concreteness?
In Bilski, all the justices rejected the patent -- five on the basis that it was an abstract idea; four would have done so on the basis that it was a business method. Justice Stevens's long and eloquent concurrence shows that patents have historically been limited to technology, as eventually articulated in the business method exclusion. Stevens and three other justices would have reinstated the exclusion, explicitly overruling State Street.
Congress to the Rescue
There is irony to how this came about. State Street's abolition of the business method exclusion was so sudden, unexpected, and retroactive, that it looked like financial firms could see their private inner workings patented out from under them. Congress was already embroiled in a drawn-out battle over patent reform (yes, another one; it seems to happen every decade). To fix this apparent inequity, Congress enacted "prior user rights" for "methods" with "method" limited to "a method of doing or conducting business." But this referred to the traditional exclusion that State Street had just said did not exist. Members of Congress scrambled to read their own definitions into the Congressional Record, some of which included manufacturing processes.
The beauty of the business method exclusion was that it was understood, accepted, and rarely litigated. Nobody was petitioning Congress for patents on business methods. But by suddenly handing out a new competitive weapon, State Street created a stampede. Fatigued by years of contentious and emotional debate over reform, Congress passed a greatly diminished reform package in 1999, but it included a stopgap for the upended expectations that State Street created, naturally without taking on the big question of where to draw the limits of patentability.
This stopgap measure simply opened the door to complete confusion over what business methods were and whether Congress, by mentioning them in this fix, had intended to validate them, whatever they were. In this way, Congress's stopgap measure, designed to remedy one particular risk created by State Street, seems to have breathed eternal life into the underlying problem. Deferring to Congress, Justice Kennedy declined to conclude that all business methods were unpatentable.
What we end up with in Bilski is: All "business methods" are not necessarily unpatentable, the machine-or-transformation test is useful but not fully determinative, and no new guidance on abstract ideas. Instead, Kennedy's opinion pulls back from the disciplined guidance that the Federal Circuit was trying to reinstate, regurgitates language from the 1970s, and says, in effect, "try again."
The Law of Abstraction
In January of 2009, CCIA, Duke Law School, and the Brookings Institution co-sponsored a conference on "abstract patents." Our notice began: "Abstract ideas are not patentable, but what are abstract ideas - and how can judges draw a line around them?" The question reverberates anew after this (non)decision in Bilski. The Supreme Court -- offering "clues" but no guidance -- has just handed this conundrum back to the Federal Circuit, inviting it to develop a concrete law of abstraction that the Supreme Court can then take another shot at.
In their book, Patent Failure, law and economics experts Mike Meurer and Jim Bessen point to the problem of fuzzy boundaries and the attendant failure of disclosure function. Drawing from empirical research by themselves and others, they show that while patents work reasonably well for pharmaceuticals, chemicals, and possibly other very tangible inventions, they work poorly for abstract subject matter such as software and business methods. They attribute this to the nature of the patent claims, which are well-defined for molecules but subject to considerable interpretation else. This makes it risky and costly to define boundaries, whether in litigation or more generally in identifying, evaluating, and navigating patents. (In comparison, consider how easy and inexpensive it is to survey and get title insurance for real estate.) At the same time, software and business method patents do not require the large investment in research and validation that new drugs do.
Justice Kennedy adopts a rhetorical framework that distinguishes the Industrial Age from the Information Age. Of course, it is not possible to separate one age from another. We still have an industrial sector and will continue to have one. Yes, the information sector has been radically expanded. So shouldn't we perhaps investigate whether information sector needs or wants a system designed for industrial use? (And could we perhaps ask those who actually make the technology work, not just the patent lawyers?)
The big problem is that the patent system remains one-size-fits all. We are stuck treating software the same as pharmaceuticals. Once business methods (or computer programs or diagnostic information) are inside the patent system, there is no escape. Middle managers are forced to live with high-priced patent lawyers at their side.
Justice Kennedy zeros in on a core issue of patenting in the Information Age: "This Age puts the possibility of innovation in the hands of more people and raises new difficulties for the patent law. With ever more people trying to innovate and thus seeking patent protections for their inventions, the patent law faces a great challenge in striking the balance between protecting inventors and not granting monopolies over procedures that others would discover by independent, creative application of general principles."
He concludes the paragraph with the ultimate disclaimer: "Nothing in this opinion should be read to take a position on where that balance ought to be struck." At least he acknowledges a balance. State Street did not. Its answer was not to describe a balance or draw a line, but to let it all in.
So the Supreme Court has now charged the Federal Circuit with developing a law of abstraction, a body of law that divides the world between abstract ideas and non-abstract (and so patentable) ideas - based on Bilski and three examples from 1972, 1978, and 1981.
Here is the fuzzy boundary challenge on a grand scale: A vast twilight zone of possibly patentable business methods, software, and diagnostic information. You would be crazy (or at least irresponsible to your shareholders) not to go for as many as you can get. A questionable patent, while not always as good as a solid patent, is a valuable weapon that can be used to threaten and bludgeon competitors, as well as anyone else willing to pay a licensing fee "reasonable" enough to avert the astronomical costs of litigation. The odds of the patent being contested on subject grounds are infinitesimal.
Today we have a huge backlog of patent applications, because the U.S. Patent and Trademark Office devotes scarce resources to managing patents in areas where standards are difficult and costly to apply -- and where patents are controversial and used frequently for ambush and ransom. The USPTO's limited resources -- and the resources of U.S. industry -- should be focused on areas where patents are needed for innovation and the system works by consensus. They should not be used for regulating business practices -- let alone tax avoidance, athletics, and the enjoyment of pets.
Brian Kahin is Senior Fellow at the Computer & Communications Industry Association.