We are working through the Faster Cures FAQ on Bayh-Dole.
2. What does Bayh-Dole say about the ownership of inventions and technologies?
Pursuant to Bayh-Dole, universities and other nonprofit organizations that receive federal funding, may “elect to retain title to any subject invention.”
Bayh-Dole addresses “patentable inventions” (and plant varieties), not “technologies.” In a sense, Bayh-Dole breaks up technologies into newly patentable fragments. As a federal agency distributes research work to many nonprofits, Bayh-Dole’s effect is to fragment any emerging technology into patentable bits and unpatentable bits. The patentable bits then end up being owned by many different nonprofits, making it virtually impossible for any company to acquire all the rights necessary to practice any particular bit of the technology–and it’s an even greater problem if a company desires to develop a product for sale. Can you think of any slower way to create a “cure”?
Thus, we might say that Bayh-Dole “says” that ownership of bits and pieces of emerging technologies should be fragmented among competing nonprofits, giving preference to each nonprofit pursuing its own financial interest in favor of combining all such ownership positions in a single entity. In prior executive branch patent policy, outside of procurement from commercial firms and the IPA program, the effort was to consolidate patent ownership claims in the federal government and then release for public use the consolidated package of rights–everyone got the same package deal to research, use, develop, compete over, obtain patents on improvements and applications, and develop standards.
If a nonprofit acquires title–making a given invention become a subject invention– Bayh-Dole requires federal agencies to allow the nonprofit to “elect to retain title,” subject to the public covenant that runs with subject inventions as expressed in the standard patent rights clause.
In 2011, the Supreme Court clarified in Stanford v. Roche that title in a patented invention vests first in the inventor.
Let’s be clear. The Supreme Court ruled that Bayh-Dole did not vest title of inventions with federal contractors. That was the question before the Court. The Court observed that Congress had not intended to alter longstanding patent law regarding ownership of inventions. Thus, it is not that title vests “first” in inventors–it is that title “vests” in inventors. There’s no “first” about it. The Supreme Court did not “rule” this–it ruled that Bayh-Dole did not address it. The Supreme Court stipulated that Bayh-Dole applies only to subject inventions–inventions that a contractor has come to own.
In that case, Stanford University sued Roche for patent infringement.
Stanford claimed that an HIV test kit invented by a Stanford professor
No. The Stanford invention had three co-inventors. One of the co-inventors was a post doc, not a professor. The post doc had been sent by Stanford to work for nine months at Cetus, a company with a new technology, PCR, that could be used for such things as evaluation of AIDS therapies. The post doc signed (with Stanford’s knowledge and consent) a patent agreement with Cetus that stipulated that the post doc would assign to Cetus any inventions made as result of the post doc’s access to PCR and Cetus research. The post doc’s faculty supervision also served on Cetus’s technical advisory board. Roche acquired Cetus’s PCR business and the post doc’s patent agreement.
belonged to the university and not to Roche, even though Roche had signed a contract with the professor that granted the company ownership of any of his inventions.
Stanford’s patent policy at the time provided that inventions would be owned by their inventors unless Stanford had a legal obligation to take ownership. The post doc was entirely within Stanford’s policy to agree that inventions done with Cetus information would be assigned to Cetus. The post doc’s patent agreement was specific to inventions arising from access to Cetus’s proprietary research, not all of his future inventions. The post doc returned from his nine months at Cetus and promptly reported an invention (co-invented with two others) that the CAFC on appeal determined to be clearly within the scope of what the post doc had agreed to assign to Cetus.
The post doc had no obligation to assign this invention to Stanford, because Stanford’s invention agreement operated only if Stanford had a legal obligation to take title. The CAFC ruled that when the invention was made, the post doc’s agreement with Cetus operated to assign his interest in the invention to Cetus. Stanford filed a patent application and then required the post doc and the other inventors to assign the invention to Stanford. The CAFC ruled that the post doc had no rights remaining to assign to Stanford. Thus Stanford could not perfect ownership of the invention, and therefore did not have standing to sue Roche for infringement.
Stanford’s main claim to ownership was that Bayh-Dole trumped any contract signed by a professor with an external company.
Yes, pretty much. The argument was that Bayh-Dole required Stanford ownership, and that should void any prior assignment of the invention by the post doc. At the appeals level, however, Stanford’s argument was slightly different. There, Stanford argued that its patent agreement promise to assign to Stanford should take priority over the post doc’s later patent agreement to assign to Cetus. The CAFC did not consider the conditional nature of Stanford’s patent agreement and instead ruled that since Stanford’s agreement was a promise to assign in the future and the post doc’s agreement with Cetus included the wording “hereby assigns,” that the post doc’s agreement with Cetus operated immediately while the agreement with Stanford required a separate step to act on the promise to assign. Thus, when Stanford got around to requesting assignment, assignment had already been made. Had the lower courts had access to the Supreme Court’s ruling on Bayh-Dole from the very first, the analysis would have been different. Justice Breyer dissented with the ruling because it allowed to stand the CAFC’s distinction between a promise to assign and a present assignment of a future interest (“hereby assigns”). Breyer argued that there should not be such a magical difference between such similar wording aiming for the same outcome.
However, the Supreme Court ruled in favor of Roche, noting that the act “does not have any language automatically vesting ownership with Universities.”iii
Yes. Nailed this part.
Assignment of title is a common condition of employment at universities and research institutions, so while this case provided important clarification of existing law, it had minimal practical impact. iv
This is nonsense, citing an op-ed by a law professor. University treatment of faculty obligations to assign inventions varies, is often ambiguous, and the operation of Bayh-Dole’s patent rights clauses–if university administrators complied with them–would limit the claims that universities could make in inventions made in faculty-led federally supported research. While “employment” might include a requirement to assign, for most research, including federally supported research, faculty are not employed. They are compensated–but by the federal government providing the funds for their salary time while working on federal projects. The law professor fixates on the question of whether a promise to assign works differently from a present assignment–which wasn’t what the Supreme Court was asked to rule on, and so didn’t.
The Stanford v Roche decision did not merely “clarify” existing law. It overturned three decades of university misrepresentation of Bayh-Dole. The “practical effect” of the ruling is that university administrators, their legal advisors, and a horde of law professors ignored the decision, ignored university non-compliance, and fixated on changing university patent policies and assignment wording so that universities might claim all inventions made by in research they hosted (and even research they didn’t host). Even Stanford, which had a liberal policy allowing inventors to own, changed its policy to be the same predatory, grasping policy that other universities had adopted, apparently to the delight of law professors everywhere. The practical impact of the decision then was for university administrators to extend (even overreach) their demand to own inventions.
Given the Supreme Court ruling on Bayh-Dole, even if Stanford had had a grasping patent policy rather than a liberal one–and had used “hereby assigns”–it would not have made any difference in the case, because Stanford subsequently permitted its post doc to enter into the later agreement. The later agreement would operate because Stanford, by permitting the agreement with Cetus, had waived the post doc’s obligations to Stanford. If the lower courts had understood that Bayh-Dole did not require Stanford to hold title, then they would have had no reason to think that the post doc’s conditional agreement to assign to Stanford (only if Stanford had a legal obligation to own) would have had any bearing on the situation. Stanford had no legal obligation to own because of the federal funding. Under Stanford’s own patent policy, it had no authority to demand ownership.
As Hermocrates pointed out in the eighth year of the Peloponnesian War, the problem is not so much people who desire to rule as it is people too ready to serve them.
3. Does Bayh-Dole apply to all research conducted at universities?
The act applies to any invention “conceived or first actually reduced to practice” in the course of research funded with federal dollars.
Not quite so. The Supreme Court in Stanford v Roche made clear that Bayh-Dole applies only to subject inventions (in discussing Bayh-Dole’s preemption in 35 USC 210 of other statutes but not federal patent law itself):
But because the Bayh-Dole Act, including §210(a), applies only to “subject inventions”—“inventions of the contractor”—it does not displace an inventor’s antecedent title to his invention. Only when an invention belongs to the contractor does the Bayh-Dole Act come into play.
The Supreme Court discounted the claim that if even one dollar of federal funding is used, Bayh-Dole comes into play. The Court was clear that Bayh-Dole applies only to subject inventions–inventions owned by a contractor. Not to all inventions. Bayh-Dole’s implementing regulations (37 CFR 4o1.1) make clear that federal funding is not the criteria. It is work performed in a project that has received at least some of its funding from the federal government. Separate accounting of funds is not a determining factor. If it were, then there would be an easy work-around to federal interest in inventions. Do federally supported research, but anywhere one encountered a problem to be solved by inventing, shift that work to a non-federal account and woohee–the invention would not “use” federal dollars. But that’s not how Bayh-Dole’s scope works.
Bayh-Dole in its contracting requirements applies to subject inventions. Subject inventions are (i) patentable inventions (or plant varieties) (ii) that have been acquired by a contractor (iii) made in performance of work (iv) funded at least in part by the federal government. Here, look (35 USC 201(e)):
The term “subject invention” means any invention of the contractor conceived or first actually reduced to practice in the performance of work under a funding agreement
“Conceived or first actually reduced to practice” are obsolete terms used by the courts to establish who was first to invent where there were competing claims to having invented. An invention was “made” when it was conceived and reduced to practice. But reduction to practice could be “constructive,” in the form of a patent application, rather than “actual” in the form of creating and testing a prototype. But the wording that follows these obsolete terms is what matters–“in the performance of work” “under a funding agreement.” “Funding agreement” is also defined:
The term “funding agreement” means any contract, grant, or cooperative agreement . . . for the performance of experimental, developmental, or research work funded in whole or in part by the Federal Government.
That is, work “under a funding agreement” must be (i) experimental, developmental, or research work; and (ii) funded in whole or in part by the Federal government. It’s that “in part” that matters. “Work under a funding agreement” then ends up being as broad as whatever project the federal government provides support for at least some part of.
Let’s look again at the Faster Cures answer:
The act applies to any invention “conceived or first actually reduced to practice” in the course of research funded with federal dollars.v
Recast: The act applies to any patentable invention owned by a contractor that has been made in the performance of a project funded at least in part by the federal government.
Implementing regulations make clear that where a non-government sponsor establishes a project that falls outside the “planned and committed activities” of a government-funded project, inventions made in the performance of that work would not be subject to Bayh-Dole.vi
Yes, this is the gist of 37 CFR 401.1. But the non-government sponsor’s project is outside the planned and committed activities of the project that the government has provided at least some funding for, and must not “diminish or distract” from the government supported project. The implementing regulations give an example: a government-funded project to “expand scientific knowledge” and a sponsor’s project to “apply” that expanded scientific knowledge. The essential point is that one has to consider the entire project, not only the bit of the project that has federal funding. Similarly, that federal funds happen to be used outside of a planned and committed project does not mean that an invention made outside that project (when acquired) becomes a subject invention.
The guidance in the implementing regulations is anything but clearly written, even if the meaning can be established with some work. The focus is on the project. Is a part of the project federally funded? Is that federal funding for experimental, developmental, or research work? Is there a patentable invention made in the project’s planned and committed activities? Has that invention been acquired by a contractor? Then it is a subject invention.
Examples of research that may be exempt from Bayh-Dole include non-government funded fellowships and industry-sponsored research projects.
The “may be exempt” means that these examples “may not be” examples. Any research that is not in a project receiving federal support is exempt from Bayh-Dole’s standard patent rights clauses. Any federal awards “primarily for educational purposes” are exempt from Bayh-Dole’s standard patent rights clauses. Any federal awards used for construction or administration are exempt from Bayh-Dole–since they are not for experimental, developmental, or research work.