In 1983, Charles Freeny Jr., an electrical engineer in Fort Worth, Texas, applied for a patent on an “information manufacturing machine” for producing music tapes, and sheet music and other products in retail stores using digital information. Two years later, Patent No. 4,528,643 “propertized” his claim.
But what claim exactly? Freeny never built the system. He sold the patent in 1989 for around $100,000. The shrewd investor who bought must have noticed that its title was strangely general: “System for Reproducing Information in Material Objects at a Point of Sale Location” – in other words, a process that we have since learned to call “down-loading.”
Just music tapes? How about CDs? MP-3 files? Digital photographs? E-books? Software? Only at the cash register, within the four walls of a retail establishment? What about, say, over the Internet, anywhere there is a personal computer to instantiate the information and to transmit payment in return? Freeney had imagined some of that, too, though not in any detail. His “information manufacturing machines” would send encrypted digital signals over telephone lines to sites equipped to make digital payments in return. The 36-page application mentioned “tapes, videodisks, handheld calculators, handheld electronic games, greeting cards, maps, and sheet music.”
In 1996, a third owner of the patent, E-Data (nee Interactive Gift Express), a tiny firm in Port Washington, NY, sent out packages to 25,000 companies involved in what gradually was becoming known as e-commerce, offering what the firm called “amnesty” in exchange for a licensing fee, as opposed to a presumably more expensive patent-infringement suit. IBM, Adobe and others quickly negotiated agreements, which were reported to range from a few thousand dollars to $40,000 a year. But many other businesses, led by Compuserve (not yet a part of AOL), chose to counter-sue.
Two years later a US District Court judge upheld the claim for retail establishments only. Three years after that the Court of Appeals of the Federal Circuit inexplicably broadened it to any and all locations. The relatively small stream of payments that the owners of the Freeney patent has managed to divert from what was becoming a mighty river of commerce didn’t matter much in the overall scheme of things The US patent expired after seventeen years in 2002, a European court invalidated it a couple of years later. But by then E-Data’s investment in the patent, said to be around $300,000, had paid off many times over. The story would make a fine movie, in the spirit of “The Wheeler Dealers” or “Five Golden Hours.”
It didn’t have to be that way, of course. A century before, Thomas Edison had beat back a challenge to his light-bulb patent, precisely because he had been specific, testing 6000 possibilities before settling on bamboo filament. The rival claimants, William Sawyer and Albion Man, had tested precisely one – carbonized paper. And forty years before that, Supreme Court Justice Joseph Story had dismissed an application for a patent on “the art of cutting ice, by means of any power, other than human power,” as “utterly unmaintainable.”
These stories, and others like them, are central to the message of Patent Failure: How Judges, Bureaucrats and Lawyers Put Innovators at Risk, a new book, by James Bessen, a software developer turned economic savant, and Michael J. Meurer, a professor at Boston University School of Law. They illuminate the murky landscape of US patent law today, wherein some 5,500 relatively low-paid patent examiners take an average of eighteen hours to read and understand each new voluminous claim, attorneys bombard them with claims that can be read narrowly in examination and broadly in litigation, good faith infringers risk having to pay heavy damages for trespassing boundaries they had little opportunity to know existed, and “submarine” patents may remain concealed for decades, until enabling technologies are enabled and commercialized.
Patent Failure is not, however, a book of war stories, or a standard economic analysis, but rather something in between. The authors draw on a wide range of literature – research on the Industrial Revolution, cross-country transnational comparisons, various natural experiments and case studies – in order to argue that the case for a strong patent system is far from proven. Distinguish the real operation of the system from the theoretical abstractions that are ordinarily employed to justify it, they say. They reason the way a lawyer might talk to a friendly judge.
Bessen and Meurer pick up roughly where the authors of an earlier book left off. Innovation and Its Discontents: How Our Broken Patent System is Endangering Innovation and Progress, and What to Do about It, by Adam Jaffe, of Brandeis University, and Josh Lerner, of Harvard Business School, documented how a series of measures undertaken in the 1980s, aimed at improving US competitiveness and accelerating growth, had profoundly altered the patent system, and perhaps even backfired.
Patent cases typically begin in a Federal District Court, whichever plaintiffs expect to give the most favorable ruling. Beleaguered by complaints about various injustices arising from such jurisdiction-shopping — especially the then-popular complaint that it was enabling Japanese manufacturers to steal a march on their American counterparts – the US Congress in 1982 decided to establish a new twelve-member centralized Court of Appeals for the Federal Circuit to hear all patent appeals. The CAFC went into business and rapidly became the patent bar’s best friend, dramatically broadening rights and remedies, extending patent protections to software and business methods, all but inviting increased litigation.
Meanwhile, under the pressure of budget stringency, Congress was converting the US Patent Office from a tax-supported agency to a profit center, one subsisting entirely on fees charged for the patents it issues and returning the overage to the Treasury. The result, not surprisingly, was to generate a flood of dubious patents and an explosion of subsequent litigation. Together, argued Jaffe and Lerner, the “reforms” had achieved a perverse effect, turning a once-powerful goad to innovation into an arena for speculators and litigators, generating uncertainty that now threatens the innovation it was meant to stimulate.
Patent Failure goes further, to argue that the problems plaguing the patent system rest on the flawed analogy between technology and real property.
In the case of land, rights and responsibilities are ordinarily quite clear. Before buying a parcel of land, a purchaser typically researches the history of the title, conducts a survey of the land itself, and takes out insurance on the deed. Before erecting a structure on his property, the owner hires a lawyer, checks the boundaries and the zoning laws, contacts abutters and proceeds with caution. Hardly ever does one person unintentionally build something on another’s land.
Yet with intellectual property, the authors note, it happens all the time. Boundaries are fuzzy and unpredictable. Would-be title-holders aren’t required to give clear notice; claims of ownership often are withheld for many years. The actual possession of an invention may not be required; often having patented the idea is enough. Search costs are high and climbing. “If you can’t tell the boundaries, then it ain’t property,” they write. And indeed most “intellectual property” really isn’t property. Instead it is an income stream, sometimes legally vouchsafed by a patent, trademark or copyright; or, far more often, secured by secrecy, skill and adaptability.
The problem stems from a key difference between technology and physical property, according to Bessen and Meurer. Real property – a parcel of land, a tractor, a bushel of grain — is inherently a rival good, meaning that only one person at a time can possess the rights to own or use it. But many people simultaneously can employ an invention; more than one person can claim to have invented it; indeed many persons can invent it independently of one another. The new ideas that we call inventions are nonrival goods, meaning that they can be employed by any number of persons at the same time with no diminution of their usefulness.
And precisely because of this nonrival aspect of technology, patents, unlike titles, don’t confer a simple affirmative right to use a particular technological idea. In order to create an artificial monopoly, they permit their owner to exclude others from its use. And that is where the present trouble has its origins. “The power to block innovation is especially troublesome when property boundaries are not well identified,” write Bessen and Meurer.
In a post on the Research on Innovation site that he runs, Bessen explains how Kenneth Arrow in a famous paper in 1962 described how property rights in ideas might help solve some incentive problems in innovation, but says that Arrow was also keenly aware of the limits of this approach. Almost all of the subsequent theoretical literature and much of the empirical literature cites Arrow, says Bessen, and then plainly assumes that patents work effectively as property, that they have well-defined boundaries that are costless to search, and any infringer is an intentional copyist.
“The book challenges this whole line of thinking. Economics seems to have a way of enshrining assumptions. Initially, a theoretical economist will make a casual assumption, often because it is convenient for mathematical analysis. Over time this assumption gets repeated and somehow becomes accepted as an empirical reality, without ever being subjected to any serious verification. And then policy recommendations are made on this basis.”
The authors are interesting participants in their own right. In addition to teaching at BU, Meurer has an economics PhD. As an MIT undergraduate, his adviser was Robert Solow and his debate coach was Lawrence Summers. Bessen wrote the first WYSIWYG (what-you-see-is-what-you-get) PC publishing software in 1983, and the company he founded sold desktop publishing solutions to commercial publishers from Sears’ catalogue to TV Guide, before he sold it to Intergraph Corp. He has worked on the economics of technological innovation ever since. (Joint research on software patents by Bessen, his college roommate Eric Maskin, of the Institute for Advanced Study, and Robert Hunt, of the Federal Reserve Bank of Philadelphia, has been influential in European policy deliberations.) Patent Failure is the subject of a symposium next week at the University of Georgia Law School. And when the brouhaha over intellectual property reform finally comes to a boil in the next Congress, Bessen and Meurer will have a seat at the table along side the law school grandees – Lawrence Lessig, James Boyle, Yochai Benkler – and the various economic dons.
Almost certainly it is too late to abandon the analogy. Intellectual property is too well-established in the public mind. Besides, no suitable alternative conceptual apparatus is available. All the more reason, then, to do as Bessen and Meurer suggest: make patents more like land claims, to the extent it is possible, by buttressing the institutions that deal with them. Make claims more transparent, sharpen their boundaries and improve the feasibility of clearance searches. Otherwise, remain mindful of the differences.
For the central problem of patent law has always been to balance the incentives to innovation against the imperative to make the benefits of new knowledge available to the widest possible number of people. That’s why patents, unlike land titles, expire after a limited period of time. It was fourteen years in 1790, when Congress passed the Patent Act; seventeen years from date of issue, from 1861 until 1995; and today, twenty years from date of application.