Professor Siebrasse is a Professor of Law at the University of New Brunswick. His patent law blog is Sufficient Description.
The turn to policy in legal analysis that began with the Realists has undoubtedly been salutary in its overall effect on the law, but we have yet to fully come to grips with the evidentiary problems that this insight implies. The problem is illustrated in the debate over the patentability of business methods as reflected in the decisions in Amazon.com in Canada, and Bilski v Kappos in the US.
Two tendencies are evident. One is to assess these decisions on the basis of whether one believes that business method patents will promote or impede innovation. This seems to follow directly from the point that law must be assessed from a policy perspective. The difficulty is that a court is normally required to decide on the basis of evidence that is properly before it, and so has been tested for reliability. A second tendency is to address the question in terms of theory, perhaps because this is seen as the special preserve of legal academics; we have no comparative advantage over practitioners in matters of doctrine, and, for the most part, we are still not empirical social scientists. The difficulty is that few practical policy questions can be resolved on theoretical grounds, and empirical evidence is difficult to assess, even if it were to be brought before the courts in a reliable form.
In respect of business method patents, theory suggests on the one hand that business method patents may impede innovation by needlessly restricting competition and raising prices of basic building blocks, and exposing true innovators to opportunistic actions. On the other hand, the fundamental theoretical justification for patent protection, and indeed intellectual property rights generally, is that innovation is fundamentally knowledge. Knowledge may require substantial investment to develop, and yet by its intangible nature is easily appropriated by free-riders who can then undercut the innovator’s price. Business method innovations fit this basic theoretical paradigm perfectly. This theory is not new. While modern scholars have added some refinements and associated jargon, like “patent thickets” and “opportunism,” the horns of the IP dilemma have been understood for centuries. Nor is this theory particularly helpful. Business method patents may indeed impede innovation; they may also promote innovation. The question is not whether they may impede innovation, but whether they do. This is an empirical question, not a theoretical one.
The empirical question is very difficult. In my forthcoming IPJ article on The Structure of the Law of Patentable Subject Matter, I review the empirical evidence on the role of patents in promoting innovation generally, and in the area of business methods in particular. There is a general consensus that patent protection is crucial to pharmaceutical innovation, and important in chemical products. Beyond that, it is difficult to draw firm conclusions. In this respect, business methods are no different from patents in most traditional fields, where subject matter patentability is uncontroversial. The same arguments made against business method patents may be, and have been made in the aircraft and semiconductor industries, for example, with significant supporting evidence. Empirical studies of the effect of business method patents are accumulating, but the evidence remains inconclusive. In a recent review article which I cite in my IPJ article, Dr. Bronwyn Hall, one of the leading empirical researchers in this field, remarks wryly that “[t]he only conclusion that is certain is that allowing business method patents will cause an increase in the patenting of business methods.”
My review of the empirical evidence is reasonably thorough, but I do not pretend it is exhaustive, and it is not undertaken to make the point that business method patents should, or should not, be patentable. My point was to illustrate the difficulty of the empirical question. This is not necessarily to say that because the question is empirically difficult, the courts should ignore it. The fact that the empirical research is equivocal as to whether business method patents would promote innovation is not the evidentiary problem I referred to at the outset of this post. The problem is that, so far as we can tell from his decision, none of the relevant evidence was even before Phelan J in Amazon.com. It appears that the only empirical evidence offered by the Commissioner on this point was the observation that there have been a great many business innovations without patents. This is true, but almost irrelevant. A great deal of innovation in purely traditional fields, including most of those fundamental to civilization, such as the wheel, the bow, the pulley, the moldboard plow etc. occurred without the incentive of patents. No one supposes that there would be no innovation at all without patents. The argument is that patents may increase the speed of innovation; that business innovations were developed without patents does not address this argument. In the Amazon.com decision Phelan J suggested that policy considerations were irrelevant to whether business methods are patentable. For this, he will no doubt be criticized by some for undue formalism. But what he might equally have said is that even if policy considerations are relevant, the evidence before him was not remotely sufficient to allow him to base his decision on policy. This problem of untested policy evidence is not unique to Canada. In Bilski v Kappos, Stevens J in concurrence directly addressed the policy question of whether business method patents would promote innovation, but the evidence he relied on was presented in a narrow selection of legal articles. The main article relied on by Stevens J – those he cited in the text rather than footnotes – were all theoretical. My own view is that Phelan J was right to recognize the inadequacy of the evidence before him, rather than give in to the temptation to carry out a policy analysis based on speculation.
The answer to this evidentiary problem is not simply for the courts to cite more empirical articles, or for legal academics to be empirical social scientists. Before asking courts to make novel determinations of law based on empirical social science evidence, we should have some reason to believe that the courts will have before them a fair representation of that evidence, and not simply whatever article their law clerk happens to have read most recently. No academic would suggest that a court could make a determination as to obviousness of a new pharmaceutical without extensive expert testimony, tested by cross-examination. The question of whether patents impede or encourage progress in a particular field of endeavour is at least equally complex, and has far more far-reaching consequences than the validity of a single patent. To suggest that a court should decide a difficult empirical question on the basis of speculation and anecdote is just as absurd as suggesting that it should decide, without expert evidence, whether it is obvious to separate a racemate into its isomers using fractional crystallization. Phelan J said that policy is irrelevant, and for this he will no doubt be criticized by some as a formalist who has not learned the Realist lesson. But what he could equally have said is that policy is relevant, but the evidence available on the record before him was grossly inadequate to make any determination as to the policy merits of patenting business methods.
The law typically deals with evidentiary difficulties through presumptions, which is what I suggest should be done in respect of business methods patents. My own view, following Binnie J in Harvard Mouse 2002 SCC 76 [43], is that there should be a presumption in favour of patentability in new fields of endeavour, “because by definition the Patent Act must contemplate the unforeseeable.” There is also a respectable argument to be made that we should not patent business methods now because it has never been done before. This kind of approach will no doubt seem lacking in excitement as compared with grappling with the big question of whether patents are a socially desirable. But the desire for intellectual excitement should never take precedence over a sound policy analysis.