James Boyle: Too hot, too cold and ... just-so stories
Richard A. Epstein: Bombing Goldilocks
For most people, patent law is a subject that conjures up images of high-priced litigation between titans of industry locked in scientific battle. The last thing that they might suppose that these hard-headed lawyers would care about are metaphysical abstractions that baffle even the most skilled philosophers.
Think again. It is just the dollars and cents that drive skilled lawyers to philosophically reflect. One key illustration of this situation was brought home to me last winter when I talked with Robert Merges, one of the world’s pre-eminent patent scholars. The conversation was sparked by my innocent question about the relationship between process and composition of matter patents. My question stemmed from some obvious confusion about why the folks who had isolated various DNA compounds were given subject matter patents for their “isolated and purified” material, instead of a process patent for the particular way in which they isolated and purified their compounds.
The source of my doubt lies in the well-known and oft-iterated proposition that no one can claim a patentable invention in any idea, physical law, or product of nature. The clear instinct behind this categorical rule is that the use of patent protection in these settings would strike a devastating blow to the ability of individuals to communicate with each other, or to continue their own scientific explorations. The protected status of inventions was given therefore, in the words of the US patent act, only to those who invent or discover “any new and useful process, machine, manufacturer, or composition of matter.” So why then is a natural substance some new composition of matter?
The various patent office decisions on DNA took the point so much for granted that they did not go back to the famous Learned Hand decision of 1912, which, as Merges told me, introduced this peculiar patent law deviation from ordinary usage. In 1901 Jokichi Takamine, a Japanese émigré to the United States isolated a variation of adrenaline (originally a trade name) that far exceeded in purity any previous isolate of this key compound. In Parke-Davis Co. v. H. K. Mulford Co. (1911), Takamine had assigned the patent to Parke-Davis - Hand sustained a composition of matter patent , saying: “while it is of course possible logically to call this a purification of the principle, it became for every practical purpose a new thing commercially and therapeutically.”
That practical interpretation is or course at variance with the meaning of the basic test. Yet, from the outset Hand put his finger on a key difficulty in fashioning an intelligent system of intellectual property: call it the Goldilocks principle. It is easy to make the soup too hot, or too cold, but it is very difficult to get the soup just right.
The composition of matter patent is, in a very real sense, too extensive for its own good. It blocks any individual who develops a wholly different technique for, say, synthesis adrenaline. The subject matter patent blocks this mode of attack. But a radically discontinuous approach is not all that likely to emerge during the patent period, especially as the patent nears expiration. And if it does, the two sides could work out a licensing deal that marries the new technology with the older substance.
But the greater risk in this area lies in the other direction. Process patents are in general too weak to protect an inventor who has made, as Takamine did, a huge advance over the prior art. It’s for this reason that inventors of special processes often protect them as trade secrets - where they are subject to the risks of reverse engineering or independent discovery - rather than by some porous patent. Generally, process patents protect a precise recipe so that the infringement action is likely to fail if the new entrant circumvents the patent by a small variation easily derivable from the basic technique. Learned Hand made a rough, but sound calculation that better a patent system too strong than too weak.
His view was not uniformly followed in the ensuing years. Before the 1952 Patent Act, the requirements for a patent stiffened, and useful discoveries about particular bacterial interactions in plants, for example, were treated as laws of nature, even though their patentable would do nothing to block the orderly progression of knowledge. But those cases have been distinguished away in more recent times. The key decision in this regard was the 1980 case of Diamond v. Chakabartry, which allowed the patenting of man-made bacteria. We can truly say that much of this huge boom in biotechnology derives from the fact that we give too much patent protection, instead of too little, in those cases where we, like Goldilocks, can never get it exactly right.
Richard A. Epstein is the James Parker Hall Distinguished Service Professor of Law at the University of Chicago, and the Peter and Kirsten Senior Fellow at the Hoover Institution
James Boyle: Too hot, too cold and ... just-so stories
Richard Epstein is absolutely right that the key goal in patent law is hitting the Goldilocks spot - not too hot, and not too cold. But the analogy both illuminates and understates the problem. Goldilocks knew what she wanted. Porridge. And there was only one dimension to fail and to succeed. Heat.
Unfortunately, patent law is not so easy. What happens with too little patent protection? Insufficient investment in innovation? Perhaps. Or perhaps too much investment in innovation as competing firms, relying on trade secrets instead of patent rights, all spend money to solve the same R&D problems. What happens with too much protection? Too much investment in innovation? It is, after all, only one social good among many. We can overspend on innovation, just as we can overspend on sports equipment, aesthetics, beer, or safety.
Perhaps we will get too much investment in innovation. Or perhaps we will get too little because venture capitalists fear that patents have been granted so broadly and vaguely that the slightest misstep by their new company could set off the minefield. In fact, perhaps minesweeping is a better metaphor than the search for the perfectly heated bowl of mush.
Professor Epstein and I both make a living training the legal minefield sweepers. We both know that every $500 billable hour charged by lawyers to do patent searches takes money away from paying engineers or bench scientists. Are the decisions that Professor Epstein mildly praises doing a good job in minimising that cost - the cost of legal uncertainty? Or are they scattering overbroad patents of uncertain dimension across the landscape and leaving it to scientists and entrepreneurs to find them - with their feet if necessary?
Empirical studies about the dysfunctions of the patent system do not paint quite the same picture as Professor Epstein does. In some areas the porridge seems to be too cold, in others too hot and in many it seems to be randomly smeared across the walls. In one recent column in these pages, Professor Epstein quoted a key empirical study that indicated that patents were not in fact slowing down research as scientists struggled to negotiated the minefield. Because licensing markets worked well? No. Because academic researchers and others were engaged in massive infringement. One has to worry any time someone defends the operation of a property system by saying “do not be concerned, many people do not obey the rules anyway, so the illogicalities don’t hurt.” The worry becomes particularly acute because the courts have recently closed the “loophole” the academics believed immunized their conduct.
If the main patent court in the United States, the Court of Appeals for the Federal Circuit, had followed Professor Epstein’s logic, then at least we could all agree that we all have the same goals, are all using the same metrics. But sadly, that court uses an idiosyncratic mixture of formalism (“a patent is a patent is a patent”) and functionalism (“let us maximise the socially desirable investment in innovation”) in ways that end up having the predictability of neither. It is that curious mixture that produced the decisions Professor Epstein describes. Were those decisions as bad as their critics claimed? No. If the courts had applied the doctrine of “non-obviousness” to curtail overbroad patents they might indeed have had the beneficial effects Professor Epstein imagines, without the negative effects the critics feared. Unfortunately the Court chose to broaden subject matter while also relaxing some of the other limitations on patentability.
The author of the American patent system, Thomas Jefferson, was humble about his ability to hit Professor Epstein’s sweet spot. “I know well the difficulty of drawing a line between the things which are worth to the public the embarrassment of an exclusive patent, and those which are not.” Unlike Professor Epstein, if Jefferson did not believe there was clear evidence that a class of patent was necessary, he would not grant it. The burden of proof was a high one.
That burden of proof, and that humility, seem to have been forgotten by the judges of the Federal Circuit. It would take more than Goldilocks to see anything truly positive in the lurches of patent jurisprudence over the last fifteen years - whether the subject is gene patents, or software or business method patents. It would take Polyanna. And a minesweeper.
James Boyle is William Neal Reynolds Professor of Law at Duke Law School, founder of the Center for the Study of the Public Domain and a board member of Creative Commons and Science Commons
Richard A. Epstein: Bombing Goldilocks
In my original column for this posting I offered a short account of how difficult it was to make appropriate trade-offs in determining the optimal scope of patent protection for one area, research into natural biological compounds that were isolated and purified by a given firm.
On balance, I think that Learned Hand got it right, and that the huge influx of funds into this area could not have taken place. Until I read James Boyle’s entertaining broadside, I did not think that I was thereby defending every decision that had been handed down by the Federal Circuit (which has a thankless task) since its inception over 20 years ago. In response I shall make two quick observations. First, I don’t think that with regard to medical patents one should infer that the cases of infringements (which do occur) lead to a breakdown of the system. Often these incursions are tolerated because the basic research done by others will increase the commercial yield of the patent down the road. Why incur transactions costs for no particular good? But when the infringers become competitors, as with the BRAC genes, then the property rights become more critical.
Second, the point of my little column was to urge a go-slow approach on patent reform. I don’t regard myself as an apologist for the current system. Quite the opposite, I am quite happy to learn about any specific areas in which the system blows too or too cold, and to work intelligent reform. But I don’t think that we should just say right off the bat that because Thomas Jefferson couldn’t locate the sweet spot on the tennis racket, that we should necessarily raise the bar for patentability. That would only aggravate the situation for those areas in which the cost of patenting is already too high. So I stand ready to learn from Boyle which decisions he thinks wrong and how they ought to be corrected. Given his vast knowledge of the field, I’d like to hear one or two specifics first.