Ten years ago, on September 4, 2008, I started the Research Enterprise blog. My idea was to use the blog to document what I had learned about university-based technology transfer over 15 years of licensing practice, and to describe ways to do a better job at helping researchers, entrepreneurs, professionals, and companies get on with working with what they make, create, author, discover, realize, collect, produce, develop, improve, and invent.
Since that first post, I’ve written over 1,300 articles. I have looked at the history of university technology transfer, at the development of university patent policies, the nature of federal research funding, methods of managing intangible assets in a research setting, the dismal monster called the Bayh-Dole Act, and the nature of innovation. Among the most popular articles are these:
A guide to the Bayh-Dole Act. Not like the crazy COGR guide or the mealy-mouthed AUTM descriptions. Not like the misrepresentations of Bayh-Dole offered up by many university technology transfer offices. In the next revision of the Guide, I think I will finally have got all the pieces in place. Core: Bayh-Dole replaced the IPA program. In that, Bayh-Dole was nothing new. Anyone who says university practice with regard to federally funded inventions suddenly changed is full of it. Bayh-Dole was created by the NIH to restore a patent monopoly pipeline from federal funding to the pharma industry, laundered through universities by patent brokers such as those at WARF. Everything else is just fluff to obscure the agenda.
In particular, the public interest apparatus in Bayh-Dole is designed not operate. The public interest apparatus has never operated–not in Bayh-Dole, not in the IPA program, not in the Kennedy patent policy dating from 1963. The public interest apparatus, too, is just fluff. Anyone thinking that the NIH will act on an apparatus that the NIH designed not to operate fails to understand the situation. Finally, Bayh-Dole has been a dismal failure on its own stated objectives. The proxy data put forward by AUTM and advocates of Bayh-Dole is often fake and always intended to mislead. Academics accepting AUTM metrics as generally valid are barking up a tree with no squirrels.
As for Bayh-Dole itself–the law gives no special privileges to contractors to acquire rights in inventions arising in federally supported work. No vesting, no first right of refusal, no mandate, no necessity to “make the law work” or to fix a “mistaken assumption.” Bayh-Dole applies to subject inventions–that is, to inventions *after* they have been acquired by a contractor through conventional means. The basic operation of the law is this: contractors, if they acquire inventions and then disclose them to the federal government, may keep ownership (35 USC 202(a)). Combine that provision with Bayh-Dole’s preemption of statutes on the matter of ownership, and it becomes clear that Bayh-Dole allows any private contractor to preempt the public purpose on which federal funding is provided, including federal agency objectives, university faculty objectives, and the objectives of groups hoping to benefit from federally supported research. Instead, the owner of patent rights on each claimed invention may assert whatever objectives it chooses. The private purposes of institutional patent owners preempt the public purposes of federal agencies, university faculty, and groups hoping for benefit. That’s Bayh-Dole in a nutshell. Yes, there is a public interest apparatus–but remember, it does not operate, was designed not to operate, and thus is just there as fluff to distract easily manipulated folk off the cleverly constructed scheme.
One of the fun dodges in Bayh-Dole practice is substituting assignment of title in patents for assignment of inventions. Though part of federal patent law, Bayh-Dole focuses on inventions (broadened to include plant varieties). For nonprofits, Bayh-Dole forbids assignment of subject inventions unless the nonprofit requirements are also assigned. The nonprofit requirements include restrictions on the use of royalty income and any other income earned with respect to subject inventions–any balance after allowable deductions must be used to support scientific research or education. So far, so good. There are two basic ways to assign an invention. The first is to expressly convey title to the invention in an assignment instrument. The second is to implicitly assign an invention by granting an exclusive license to all substantial rights in the invention–the rights to make, use, and sell–which then gives the “licensee” the right to enforce patent rights in the invention.
Universities routinely grant exclusive patent licenses that also assign the subject invention. They withhold assignment of the patents, but they implicitly assign the subject inventions, without complying with the standard patent rights clause requirement that the assignee agrees to the nonprofit restrictions. University administrators are in general denial about all this–they want the law to be whatever it is they do, a do WTF sort of law.
While this distinction between assignment of a patent and assignment of a subject invention might seem arcane, it makes a huge difference in practice. Were universities to comply with the law, they would have to break up their patent monopolies in any licensing activity where the licensees did not dedicate any income they earned to the specified public purposes. If pharma companies wanted exclusive licenses that were also assignments, they would have to set aside their earnings on those inventions for scientific research or education, not for dividend payments to shareholders, not to salaries for executives, not to new construction, not to acquiring other companies, not to development of other commercial products.
The upshot, of course, is that universities would not grant such exclusive licenses that are really assignments–instead, they would break up their patent monopolies and spread the opportunity around. That is what Bayh-Dole requires. But since Bayh-Dole is not enforced and its public interest apparatus does not operate, Bayh-Dole becomes a pipeline for patent monopolies, held by pharma companies as if their own money, not federal money, had supported the research that led to the patent positions. Federal money becomes a subsidy for the creation of pharma monopolies, using a massive creation of university monopolies in research assets–mostly unlicensed and when licensed mostly undeveloped–as the means. If you think that this is a really good thing, then Bayh-Dole is your wildly successful dream come true.
This article works at the problem of scope in Bayh-Dole. Bayh-Dole uses a now pretty much obsolete distinction made in patent practice having to do with the priority date for first to invent–both the date of conceiving an invention and the date of reduction to practice mattered for determining the priority for first to invent. Bayh-Dole, however, follows a pattern of practice in federal contracting that uses “conceived or first actually reduced to practice” to establish the scope of federal claims on inventions made in projects receiving federal support. Again, this is arcane stuff, but the technical details make a huge difference in what Bayh-Dole defines as a subject invention.
Again, university administrators, generally not up to the mental requirements of dealing with such a poorly drafted statute, replace the law with their own patch-thinking on the matter. They treat subject invention as whatever they assert to be a subject invention, and federal agencies for the most part of let them do WTF they want. Because university administrators were under the delusion that Bayh-Dole vested invention ownership with universities that hosted federally supported work, administrators wanted everything to be a subject invention–because then “federal law” required inventors to give up any claims in their inventions and administrators did not have rely on their own badly drafted patent policies and employment agreements as a basis to demand assignment.
I am guessing, however, that the part that makes this article popular is the account of Richard Feynman’s inventions while at Los Alamos. Anyone on a cutting edge of science or engineering or medicine can invent pretty much at will. The problem is not whether such inventions are patentable but rather whether they come within the scope of Bayh-Dole after they are assigned to a federal contractor. And it’s here that university administrators (and COGR and AUTM and even NIST and most everyone else) have Bayh-Dole bass-ackwards. They think Bayh-Dole’s “subject invention” is any invention that gets made with any remote connection to federal funding. But that is not the case, as the Supreme Court made clear in Stanford v Roche (2011). Thus, the crucial element in Bayh-Dole’s definition of subject invention is that a contractor owns it. If a contractor does not own such an invention, then those laws that Bayh-Dole otherwise preempts continue to operate. Essentially, things are just as they were under the Kennedy patent policy, except where Bayh-Dole preempts–and for research contracting Bayh-Dole preempts only when a contractor acquires, not when an invention is made.
Sure, it’s complicated–but the consequences are hugely different depending on whether one follows the rule of law or the do WTF university administrators lusting to deal in patents for profits, especially a share of pharma and biotech profits, or at least a share of what speculators make trading on the future value of patent monopolies in pharma and biotech. Bayh-Dole actually creates a narrow definition of subject invention. The “first actually reduced to practice” means that somewhere in the statement of work, there has to be something that expressly specifies actual reduction to practice–that is demonstration that an invention already owned by a contractor operates in every part of what is claimed. Yes–a contractor would have to propose actually reducing to practice an invention that the contractor owned for the invention to come within the scope of Bayh-Dole’s requirements as a subject invention. The invention may well come within the broader regulatory requirements that Bayh-Dole otherwise preempts when a contractor owns–but at that point, it is no longer a Bayh-Dole issue.
A guide on research management published by an “information service” provides disinformation on Bayh-Dole. This article works through the article and demonstrates the bad guidance. It’s just one of many guides to Bayh-Dole that get a lot of things–crucial things–wrong. At some point after I worked through the guide, it was put behind a paywall at Thomson, so now people have to pay to get bad advice that previously they could get for free, assuming that the guide hasn’t been revised with Research Enterprise’s helpful (and free) comments.
There are many details that the guide gets wrong or omits. Read the article to work through them. The bigger issue is that the guide, like almost everything out there on Bayh-Dole, is written from the perspective of advising university administrators how to abuse the law to get what the guide assumes they want–ownership of everything they can get their mitts on. Bayh-Dole then becomes a form of cover for administrative claims–“you must assign because federal law requires it,” or “federal law requires us to require you to assign,” or something. That perspective is one in which bureaucrats serve the public interest by stripping inventors of their rights, citing a federal law that does no such thing. It’s not just that the guide assumes that stripping inventors of their rights is in the public interest–it’s that it is a felony to do so under the color of law.
Luckily for the authors of the guide and university administrators who love these sorts of documents backing up their claims, no one cares to enforce Bayh-Dole or to discipline university administrators for breaking the law. University innovation must be too important for laws to apply–or too unimportant for anyone to seriously care what university bureaucrats do. Perhaps the high cost of prescription drugs combined with the amazing lack of progress in finding cures or causes might fuel a reaction to Bayh-Dole that gets the law enforced or repealed or modified to hold back university bureaucrats from putting their thumbs into every innovation pie.
University administrators are fixated on a version of the “linear model” of innovation. In this mindset, basic research leads to applied research, leads to development of commercial products, and that’s how the public will benefit from federally supported work. According to this mindset, commercial products are necessary and commercial products are only developed if a company holds a monopoly position–has full control over a patent that covers not only the product that the company intends to make, but that also covers all other products that a company does not intend to make, all functional equivalents of all products, and all uses of the invention in any of its variations that do not require a commercial product at all.
People trace this linear model back to Vannevar Bush, and especially Science the Endless Frontier. But it’s just not there, as Benoît Godin has made clear. If one works through Bush’s published work, it’s clear that Bush realized a substantially different approach to innovation that has been conflated to justify the linear model. Instead, Bush advocated for expansion of basic science “frontiers” to open up new territory for technological development. The actual process of innovation Bush used involved gathering scientists (from universities), engineers (from industry), and gadgeteers (from the trades). See Modern Arms and Free Men. These teams, then, could develop new technologies for use by the military. In effect, Bush found a way to develop things that an established order couldn’t imagine or if they could imagine they wouldn’t agree to fund, and even if they wanted to fund the work they wouldn’t have the slightest idea how to go about it.
One might say that Bush operated skunk works outside the established military order and engaged that order at its highest levels in order to introduce practice and technology change into that order. An innovation engine. Bush and Roosevelt aimed to adapt this innovation to civilian uses, starting with medicine in place of the military, and with communications apparently lined up next (for which see “As We May Think”). Basic research then expanded the scientific resources available to teams working outside established orders of practice. A team working skunk relative to medicine might draw from chemistry, or from physics, or from (these days, computer science). A discovery might jump fields, might be applied in unexpected ways, might sit for years before the right context comes along, and for all that, might be just a doorway into a phenomenon to explore and exploit for a given purpose.
I call this the “unexpected” model of innovation. While basic research can be encouraged, managing the results–even the inventive results–runs against the approach. The last thing that the results need is “management” by being owned as a monopoly for two decades, held on behalf of some future company that requires an exclusive position even to consider attempting to create a commercial product from 1% of the claimed invention. Instead, the expansion of scientific frontiers is decoupled with innovation engines created outside established orders–outside government, outside corporations and industries controlled by corporations, outside even “established” science. And this approach has worked amazingly on the information technology side, with the digital computer and the internet both coming about as skunk works that engaged industry suppliers–and did so without either the linear model or university-laundered patent monopolies. Instead, first there was skunk development with industry supplying parts, and then standards, and then crazy development with all sorts of proprietary positions and rounds of new and additional standards. The commercial opportunities and the commons ran in parallel. Not either communist-free or communist-controlled or capitalist-free or capitalist-controlled. Not only “open” and not only “proprietary.” That’s hard to grasp for binary thinkers, but there it is.
If university administrators did their jobs and also followed Bayh-Dole, they could help university research fuel skunk works to develop really new stuff–stuff that obsolesces commercial products rather than attempts to get onto existing development roadmaps, but with a patent cum roadside bomb. Instead of pandering to the pharma industry, the skunk works would aim to get beyond what the pharma industry could specify for itself in its own wisdom and self-interest. We would be working beyond the clinical trial and all of its statistical fallacy and misdirection. We would be working on causes and cures, not on making acute conditions chronic. We would do that regardless of whether there is the prospect of vast wealth on offer for any particular bit of science or discovery. Hard to believe that anyone these days would work other than for money on offer from wealthy sources. Perhaps it is over the top to imagine university faculty motivated other than by having an expensive car to drive, paid for by a patent bought up by pharma. But that’s how fuddy-duddy Vannevar Bush thought things might work. Once Truman and government lawyers and university administrators got done with things, we had a linear model and every federal agency sponsoring “basic research” using a watered-down procurement contract pleasing to university administrators.
This article works through five areas in which university administrators generally do not comply with Bayh-Dole. There are companies out there giving workshops and offering audit services with regard to Bayh-Dole compliance, but they focus on the most mundane paperwork requirements of Bayh-Dole and even get these wrong (as far as I can tell). What gets audited in Bayh-Dole has next to nothing to do with research innovation.
Here are the five areas:
- Misrepresentations of Bayh-Dole and its implementing regulations.
- Implementation of the (f)(2) written agreement requirement (not in Bayh-Dole).
- Determination of what constitutes a subject invention.
- Assignments mislabeled as exclusive licenses.
- Use of subject invention income.
For each of these areas, I give examples of university policy statements and practice documents that show the failure to comply. The difference between these areas and the practices that do receive audit attention is that these areas, were compliant practice enforced, would materially change how university administrators managed inventions. Universities would make fewer claims to ownership, and would be constrained with how they managed what they did own. Many more inventions–perhaps the bulk of them–would be governed by the laws and regulations that aren’t preempted by Bayh-Dole when a contractor acquires ownership of an invention made in work with federal support. Perhaps that regime could be improved as well–but the convenience of university-affiliated patent brokers would not have anything to do with discussion.
At the heart of Bayh-Dole–as the law is written and implemented–is a simple fact. University inventors are small business contractors everywhere that they are not assigned and controlled by the university as their employer. Thus, university faculty and students and even staff working under the direction of university faculty are, generally, small business contractors in their own right. They have the benefit of Bayh-Dole’s basic disposition of ownership at 35 USC 202(a) just as universities do, protection from university predatory ownership claims under 37 CFR 401.14(g), and they have the further benefit of their own patent rights clause at 37 CFR 401.9. Under that patent rights clause, university inventors are not required to use the patent system, and are not required to assign their inventions to anyone, and if they assign their inventions, they are not restricted to the requirements placed on nonprofit assignments. That heart of Bayh-Dole is pretty interesting stuff. But university administrators, and AUTM, and other university front groups such as APLU, AAU, and COGR, have worked like good aztecs to slice the heart out of Bayh-Dole in favor of institutional control of research results–early, broadly, often, and without accountability.