Rochelle Dreyfuss: Whither Experimental Use?

Since the Federal Circuit's formation in 1982, Rochelle Dreyfuss has witnessed patent law go through some significant changes, and the climate for university research and biotechnology innovation today is markedly different from what it was thirty years ago. However, Dreyfuss stated at the beginning of remarks she gave at Akron Law on Friday November 4, "sometimes facts change; but one's views on a subject do not." Patent law's common law "experimental use" defense–or lack thereof–is, she believes, a case in point. When she wrote on this topic over ten years ago in her paper Protecting the Public Domain of Science: Has the Time for an Experimental Use Defense Arrived?, she felt largely the same as she does now: universities need the benefit of "some version of an invigorated experimental use defense."  In her new paper, "Reconsidering Experimental Use," to be published in the Akron Law Review this spring., Dreyfuss returns to the important question of whether, and to what extent, patented inventions can be used for experimental purposes during the term of a patent. Read more at the jump.

There are two forms of experimental use defense in patent law: the statutory defense and the common law defense. The statutory defense under 35 U.S.C. § 271(e) exempts certain activities "reasonably related" to development and submission of information to the FDA. The idea is to allow pharmaceutical drug companies to get a head-start in obtaining regulatory approval, in order to provide consumers with quicker access to cheap generic drugs without undermining patentee incentives.

But Dreyfuss' interest is the common law experimental use defense–or what remains of it. First crafted in two cases decided in 1813 by Justice Story, the common law experimental use defense broadly permitted use of patented innovations during the patent term for mere "philosophical experiments" as opposed to for-profit activities. In the modern era, this defense has been whittled down significantly. In Madey v. Duke Univ., 307 F.3d 1351 (Fed. Cir. 2002), the Federal Circuit virtually eliminated the defense, at least in the context of university research, holding that non-profit motives and the pursuit of science do not create an exception to infringement for universities using patented innovation.

To be fair, the Federal Circuit noted at the time that it seems an anomaly to allow universities to benefit from a defense to infringement when they themselves profited so extensively from commercializing their patents. Duke’s ongoing research projects involving a patented laser, the court wrote in Madey, “unmistakably further [Duke’s] legitimate business objectives, including educating and enlightening students and faculty participating in these projects.” Moreover, thanks to sovereign immunity, state universities cannot be sued for patent damages when they infringe others' patents.

But Dreyfuss viewed the Federal Circuit's holding with apprehension. Now that "universities are themselves deep in the intellectual property business[,]" she wrote in 2004, "there is a kind of vicious cycle taking hold.  Once universities acquired incentives to push for patents, they began to look like commercial actors. Once they began to look like commercial actors, the Duke court decided to treat them like commercial actors. Since commercial actors do not enjoy the experimental use defense, universities lost its benefits too."

As Peter Lee has observed, the common law experimental use exception has indeed come a long way since its formation if even traditional nonprofit actors like universities cannot in the majority of cases rely on it–regardless of their "philosophical" motives and regardless of the social importance of their work. As Lee puts it, "the Federal Circuit’s conception of the modern research university diverges sharply from 'Justice Story’s early-19th-century picture of a gentleman scientist driven by idle curiosity.' " Lee, Patents and the University) (quoting Eisenberg's 1989 article, Patents and the Progress of Science: Exclusive Rights and Experimental Use).

Dreyfuss is both concerned and frustrated by this situation. In 2011, it appeared Congress might act to broaden experimental use to cover certain kinds of university research when it passed the America Invents Act (AIA). But it did not do so. Now, Dreyfuss reflects with disappointment, U.S. patent law is crippled in comparison to the laws of other countries. The problem is on two ends of patent doctrine: patentable subject matter, on the one hand, and experimental use, on the other. Rather than broadly allowing patents on biotechnology research and simultaneously permitting a narrowly crafted defense for non-profit biotechnology experiments in order to stimulate further innovation, instead we have effectively no common law experimental use exception and a patentable subject matter bar that limits patents for "products of nature" and "abstract ideas." Needless to say the doctrinal hurdles put in place by Alice and Mayo aggravate this double-whammy.

Dreyfuss sees several problems in experimental use's demise, most of them economic. (Others such as Eisenberg and Arti Rai have also focused on the effects that patents-without-exceptions can have on the "norms of science.") The first is the problem of "brain drain." If scientists in this country cannot rely on an experimental use defense, Dreyfuss stated, they may migrate in protest to institutions in countries whose patent laws do allow the defense, such as Japan, Germany, France, the United Kingdom, Italy, the Netherlands, Belgium, Luxembourg, Ireland, Greece or Spain. This would potentially deplete U.S. research clusters down to those willing to perform a more limited range of research. Second is a related problem she calls "arbitrage": where a company wishing to do otherwise prohibited research shifts its research to other places where U.S. patents are not enforced. Lastly is the broader problem of patents' effect on domestic innovation. Dreyfuss summarized the results of a variety of research by economists such as Fiona Murray and Heidi Williams suggesting that the impact of patents on scientific research may be negative, reducing the extent of the use of knowledge to create more knowledge.

What's the solution? Dreyfuss discussed several options proposed by herself and others, ranging from private-ordering solutions such as a voluntary waiver-of-infringement for non-profit experiments, to a "fair-use-style" defense that would permit uses of patents that are to some degree "transformative" and so limited that they do not impair patentees' markets for their inventions. Such limitations on patentees' exclusive rights would obviously not be unprecedented. Justice Story's comparatively broad exception for "philosophical" experiments is just one historical example.  As I will discuss in a paper I am writing for the "IP and federalism" panel at AALS this year, U.S. states were once permitted to regulate assertions of patents–for instance, by taxing patentees who wished to license in the state or by requiring patentees to declare "oaths of genuineness"–but only to the extent such restrictions were not "unreasonably" burdensome on patentees. (For those of you attending AALS this year, it is panel #7050, to be held on Saturday, January 7, at 8:30 - 10:15 am.)

Which institution should be making these kinds of exceptions?  The states are obviously not the first institution we think of.  The Supreme Court, Dreyfuss noted, certainly has an opportunity here, should the "right case" come up on appeal. (Why did Myriad have to be a PSM case anyway, attendees of Dreyfuss' talk pondered?)  On the other hand, this may be an area better left to Congress. Arguably, the drafters of the patent statute should be involved in any re-formulation of the scope of an experimental use defense, especially given the complexity of the AIA's new "first to file" rule. For example, as Akron Law's Dan Brean noted, adoption of a broader experimental use defense for infringement purposes might also appear to justify an expansion of the corresponding experimental use exception for prior art purposes. In other words, if defendants can use patented inventions more liberally in order to experiment, then perhaps patentees should likewise be able to use their own inventions more liberally for experimental purposes before filing for a patent. Dreyfuss agreed that some symmetry would be appropriate, but did not specifically comment on whether or how much such a result would offset the benefits of her proposals to limit infringement liability by removing potential patentability hurdles for experimental activities.

In sum, it was a fascinating talk and a reminder that in patent law, much of what seems new and unprecedented has actually happened, in some iteration, before. Look out for Dreyfuss' paper in the Akron Law Review in the spring.